start 2026-10-05T05:54:15Z
Seed set to 42
GPU available: True (mps), used: True
TPU available: False, using: 0 TPU cores
💡 Tip: For seamless cloud logging and experiment tracking, try installing [litlogger](https://pypi.org/project/litlogger/) to enable LitLogger, which logs metrics and artifacts automatically to the Lightning Experiments platform.

  | Name         | Type         | Params | Mode  | FLOPs
--------------------------------------------------------------
0 | fc_ms        | AlignmentMLP | 12.7 M | train | 0    
1 | fc_mol       | AlignmentMLP | 1.0 M  | train | 0    
  | other params | n/a          | 1      | n/a   | n/a  
--------------------------------------------------------------
13.7 M    Trainable params
0         Non-trainable params
13.7 M    Total params
54.835    Total estimated model params size (MB)
18        Modules in train mode
0         Modules in eval mode
0         Total Flops

Sanity Checking: |          | 0/? [00:00<?, ?it/s]/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/.venv/lib/python3.11/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:485: Your `val_dataloader`'s sampler has shuffling enabled, it is strongly recommended that you turn shuffling off for val/test dataloaders.
/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/.venv/lib/python3.11/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:429: Consider setting `persistent_workers=True` in 'val_dataloader' to speed up the dataloader worker initialization.

Sanity Checking: |          | 0/? [00:00<?, ?it/s]
Sanity Checking DataLoader 0:   0%|          | 0/2 [00:00<?, ?it/s]
Sanity Checking DataLoader 0:  50%|█████     | 1/2 [00:00<00:00,  6.79it/s]
Sanity Checking DataLoader 0: 100%|██████████| 2/2 [00:00<00:00, 12.40it/s]
                                                                           

Training: |          | 0/? [00:00<?, ?it/s]
Training: |          | 0/? [00:00<?, ?it/s]
Epoch 0:   0%|          | 0/3295 [00:00<?, ?it/s]
Epoch 0:   0%|          | 1/3295 [00:02<2:09:39,  0.42it/s]
Epoch 0:   0%|          | 1/3295 [00:02<2:09:40,  0.42it/s]
Epoch 0:   0%|          | 2/3295 [00:02<1:05:41,  0.84it/s]
Epoch 0:   0%|          | 2/3295 [00:02<1:05:41,  0.84it/s]
Epoch 0:   0%|          | 3/3295 [00:02<44:04,  1.24it/s]  
Epoch 0:   0%|          | 3/3295 [00:02<44:04,  1.24it/s]
Epoch 0:   0%|          | 4/3295 [00:03<54:14,  1.01it/s]
Epoch 0:   0%|          | 4/3295 [00:03<54:14,  1.01it/s]
Epoch 0:   0%|          | 5/3295 [00:03<43:42,  1.25it/s]
Epoch 0:   0%|          | 5/3295 [00:03<43:42,  1.25it/s]
Epoch 0:   0%|          | 6/3295 [00:04<36:39,  1.50it/s]
Epoch 0:   0%|          | 6/3295 [00:04<36:39,  1.50it/s]
Epoch 0:   0%|          | 7/3295 [00:05<44:23,  1.23it/s]
Epoch 0:   0%|          | 7/3295 [00:05<44:23,  1.23it/s]
Epoch 0:   0%|          | 8/3295 [00:05<39:31,  1.39it/s]
Epoch 0:   0%|          | 8/3295 [00:05<39:31,  1.39it/s]
Epoch 0:   0%|          | 9/3295 [00:05<35:39,  1.54it/s]
Epoch 0:   0%|          | 9/3295 [00:05<35:39,  1.54it/s]
Epoch 0:   0%|          | 10/3295 [00:07<40:44,  1.34it/s]
Epoch 0:   0%|          | 10/3295 [00:07<40:44,  1.34it/s]
Epoch 0:   0%|          | 11/3295 [00:07<37:12,  1.47it/s]
Epoch 0:   0%|          | 11/3295 [00:07<37:12,  1.47it/s]
Epoch 0:   0%|          | 12/3295 [00:07<35:28,  1.54it/s]
Epoch 0:   0%|          | 12/3295 [00:07<35:28,  1.54it/s]
Epoch 0:   0%|          | 13/3295 [00:09<38:54,  1.41it/s]
Epoch 0:   0%|          | 13/3295 [00:09<38:54,  1.41it/s]
Epoch 0:   0%|          | 14/3295 [00:09<36:26,  1.50it/s]
Epoch 0:   0%|          | 14/3295 [00:09<36:26,  1.50it/s]
Epoch 0:   0%|          | 15/3295 [00:09<34:45,  1.57it/s]
Epoch 0:   0%|          | 15/3295 [00:09<34:45,  1.57it/s]
Epoch 0:   0%|          | 16/3295 [00:11<38:28,  1.42it/s]
Epoch 0:   0%|          | 16/3295 [00:11<38:28,  1.42it/s]
Epoch 0:   1%|          | 17/3295 [00:11<36:51,  1.48it/s]
Epoch 0:   1%|          | 17/3295 [00:11<36:51,  1.48it/s]
Epoch 0:   1%|          | 18/3295 [00:11<34:52,  1.57it/s]
Epoch 0:   1%|          | 18/3295 [00:11<34:52,  1.57it/s]
Epoch 0:   1%|          | 19/3295 [00:13<37:27,  1.46it/s]
Epoch 0:   1%|          | 19/3295 [00:13<37:27,  1.46it/s]
Epoch 0:   1%|          | 20/3295 [00:13<36:24,  1.50it/s]
Epoch 0:   1%|          | 20/3295 [00:13<36:25,  1.50it/s]
Epoch 0:   1%|          | 21/3295 [00:13<34:44,  1.57it/s]
Epoch 0:   1%|          | 21/3295 [00:13<34:44,  1.57it/s]
Epoch 0:   1%|          | 22/3295 [00:14<36:54,  1.48it/s]
Epoch 0:   1%|          | 22/3295 [00:14<36:54,  1.48it/s]
Epoch 0:   1%|          | 23/3295 [00:15<35:52,  1.52it/s]
Epoch 0:   1%|          | 23/3295 [00:15<35:52,  1.52it/s]
Epoch 0:   1%|          | 24/3295 [00:15<34:25,  1.58it/s]
Epoch 0:   1%|          | 24/3295 [00:15<34:25,  1.58it/s]
Epoch 0:   1%|          | 25/3295 [00:16<36:30,  1.49it/s]
Epoch 0:   1%|          | 25/3295 [00:16<36:31,  1.49it/s]
Epoch 0:   1%|          | 26/3295 [00:16<35:36,  1.53it/s]
Epoch 0:   1%|          | 26/3295 [00:16<35:36,  1.53it/s]
Epoch 0:   1%|          | 27/3295 [00:17<34:26,  1.58it/s]
Epoch 0:   1%|          | 27/3295 [00:17<34:26,  1.58it/s]
Epoch 0:   1%|          | 28/3295 [00:18<36:21,  1.50it/s]
Epoch 0:   1%|          | 28/3295 [00:18<36:21,  1.50it/s]
Epoch 0:   1%|          | 29/3295 [00:18<35:19,  1.54it/s]
Epoch 0:   1%|          | 29/3295 [00:18<35:19,  1.54it/s]
Epoch 0:   1%|          | 30/3295 [00:18<34:24,  1.58it/s]
Epoch 0:   1%|          | 30/3295 [00:18<34:24,  1.58it/s]
Epoch 0:   1%|          | 31/3295 [00:20<36:07,  1.51it/s]
Epoch 0:   1%|          | 31/3295 [00:20<36:07,  1.51it/s]
Epoch 0:   1%|          | 32/3295 [00:20<35:18,  1.54it/s]
Epoch 0:   1%|          | 32/3295 [00:20<35:18,  1.54it/s]
Epoch 0:   1%|          | 33/3295 [00:20<34:16,  1.59it/s]
Epoch 0:   1%|          | 33/3295 [00:20<34:16,  1.59it/s]
Epoch 0:   1%|          | 34/3295 [00:22<35:34,  1.53it/s]
Epoch 0:   1%|          | 34/3295 [00:22<35:34,  1.53it/s]
Epoch 0:   1%|          | 35/3295 [00:22<34:55,  1.56it/s]
Epoch 0:   1%|          | 35/3295 [00:22<34:55,  1.56it/s]
Epoch 0:   1%|          | 36/3295 [00:22<34:16,  1.58it/s]
Epoch 0:   1%|          | 36/3295 [00:22<34:16,  1.58it/s]
Epoch 0:   1%|          | 37/3295 [00:24<35:32,  1.53it/s]
Epoch 0:   1%|          | 37/3295 [00:24<35:32,  1.53it/s]
Epoch 0:   1%|          | 38/3295 [00:24<34:44,  1.56it/s]
Epoch 0:   1%|          | 38/3295 [00:24<34:44,  1.56it/s]
Epoch 0:   1%|          | 39/3295 [00:24<34:13,  1.59it/s]
Epoch 0:   1%|          | 39/3295 [00:24<34:13,  1.59it/s]
Epoch 0:   1%|          | 40/3295 [00:26<35:29,  1.53it/s]
Epoch 0:   1%|          | 40/3295 [00:26<35:29,  1.53it/s]
Epoch 0:   1%|          | 41/3295 [00:26<34:39,  1.56it/s]
Epoch 0:   1%|          | 41/3295 [00:26<34:39,  1.56it/s]
Epoch 0:   1%|▏         | 42/3295 [00:26<34:04,  1.59it/s]
Epoch 0:   1%|▏         | 42/3295 [00:26<34:04,  1.59it/s]
Epoch 0:   1%|▏         | 43/3295 [00:27<35:14,  1.54it/s]
Epoch 0:   1%|▏         | 43/3295 [00:27<35:14,  1.54it/s]
Epoch 0:   1%|▏         | 44/3295 [00:28<34:34,  1.57it/s]
Epoch 0:   1%|▏         | 44/3295 [00:28<34:34,  1.57it/s]
Epoch 0:   1%|▏         | 45/3295 [00:28<33:52,  1.60it/s]
Epoch 0:   1%|▏         | 45/3295 [00:28<33:52,  1.60it/s]
Epoch 0:   1%|▏         | 46/3295 [00:29<35:07,  1.54it/s]
Epoch 0:   1%|▏         | 46/3295 [00:29<35:07,  1.54it/s]
Epoch 0:   1%|▏         | 47/3295 [00:29<34:32,  1.57it/s]
Epoch 0:   1%|▏         | 47/3295 [00:29<34:33,  1.57it/s]
Epoch 0:   1%|▏         | 48/3295 [00:30<33:52,  1.60it/s]
Epoch 0:   1%|▏         | 48/3295 [00:30<33:52,  1.60it/s]
Epoch 0:   1%|▏         | 49/3295 [00:31<34:56,  1.55it/s]
Epoch 0:   1%|▏         | 49/3295 [00:31<34:56,  1.55it/s]
Epoch 0:   2%|▏         | 50/3295 [00:31<34:29,  1.57it/s]
Epoch 0:   2%|▏         | 50/3295 [00:31<34:29,  1.57it/s]
Epoch 0:   2%|▏         | 51/3295 [00:31<33:50,  1.60it/s]
Epoch 0:   2%|▏         | 51/3295 [00:31<33:50,  1.60it/s]
Epoch 0:   2%|▏         | 52/3295 [00:33<34:44,  1.56it/s]
Epoch 0:   2%|▏         | 52/3295 [00:33<34:44,  1.56it/s]
Epoch 0:   2%|▏         | 53/3295 [00:33<34:20,  1.57it/s]
Epoch 0:   2%|▏         | 53/3295 [00:33<34:20,  1.57it/s]
Epoch 0:   2%|▏         | 54/3295 [00:33<33:43,  1.60it/s]
Epoch 0:   2%|▏         | 54/3295 [00:33<33:43,  1.60it/s]
Epoch 0:   2%|▏         | 55/3295 [00:35<34:36,  1.56it/s]
Epoch 0:   2%|▏         | 55/3295 [00:35<34:36,  1.56it/s]
Epoch 0:   2%|▏         | 56/3295 [00:35<34:20,  1.57it/s]
Epoch 0:   2%|▏         | 56/3295 [00:35<34:20,  1.57it/s]
Epoch 0:   2%|▏         | 57/3295 [00:35<33:45,  1.60it/s]
Epoch 0:   2%|▏         | 57/3295 [00:35<33:45,  1.60it/s]
Epoch 0:   2%|▏         | 58/3295 [00:37<34:29,  1.56it/s]
Epoch 0:   2%|▏         | 58/3295 [00:37<34:29,  1.56it/s]
Epoch 0:   2%|▏         | 59/3295 [00:37<34:03,  1.58it/s]
Epoch 0:   2%|▏         | 59/3295 [00:37<34:03,  1.58it/s]
Epoch 0:   2%|▏         | 60/3295 [00:37<33:34,  1.61it/s]
Epoch 0:   2%|▏         | 60/3295 [00:37<33:34,  1.61it/s]
Epoch 0:   2%|▏         | 61/3295 [00:38<34:18,  1.57it/s]
Epoch 0:   2%|▏         | 61/3295 [00:38<34:18,  1.57it/s]
Epoch 0:   2%|▏         | 62/3295 [00:39<33:58,  1.59it/s]
Epoch 0:   2%|▏         | 62/3295 [00:39<33:58,  1.59it/s]
Epoch 0:   2%|▏         | 63/3295 [00:39<33:30,  1.61it/s]
Epoch 0:   2%|▏         | 63/3295 [00:39<33:30,  1.61it/s]
Epoch 0:   2%|▏         | 64/3295 [00:40<34:13,  1.57it/s]
Epoch 0:   2%|▏         | 64/3295 [00:40<34:13,  1.57it/s]
Epoch 0:   2%|▏         | 65/3295 [00:40<33:57,  1.59it/s]
Epoch 0:   2%|▏         | 65/3295 [00:40<33:57,  1.59it/s]
Epoch 0:   2%|▏         | 66/3295 [00:41<33:26,  1.61it/s]
Epoch 0:   2%|▏         | 66/3295 [00:41<33:26,  1.61it/s]
Epoch 0:   2%|▏         | 67/3295 [00:42<33:58,  1.58it/s]
Epoch 0:   2%|▏         | 67/3295 [00:42<33:58,  1.58it/s]
Epoch 0:   2%|▏         | 68/3295 [00:42<33:51,  1.59it/s]
Epoch 0:   2%|▏         | 68/3295 [00:42<33:51,  1.59it/s]
Epoch 0:   2%|▏         | 69/3295 [00:43<33:30,  1.60it/s]
Epoch 0:   2%|▏         | 69/3295 [00:43<33:30,  1.60it/s]
Epoch 0:   2%|▏         | 70/3295 [00:44<33:48,  1.59it/s]
Epoch 0:   2%|▏         | 70/3295 [00:44<33:48,  1.59it/s]
Epoch 0:   2%|▏         | 71/3295 [00:44<33:51,  1.59it/s]
Epoch 0:   2%|▏         | 71/3295 [00:44<33:51,  1.59it/s]
Epoch 0:   2%|▏         | 72/3295 [00:44<33:26,  1.61it/s]
Epoch 0:   2%|▏         | 72/3295 [00:44<33:26,  1.61it/s]
Epoch 0:   2%|▏         | 73/3295 [00:45<33:44,  1.59it/s]
Epoch 0:   2%|▏         | 73/3295 [00:45<33:44,  1.59it/s]
Epoch 0:   2%|▏         | 74/3295 [00:46<33:48,  1.59it/s]
Epoch 0:   2%|▏         | 74/3295 [00:46<33:48,  1.59it/s]
Epoch 0:   2%|▏         | 75/3295 [00:46<33:27,  1.60it/s]
Epoch 0:   2%|▏         | 75/3295 [00:46<33:27,  1.60it/s]
Epoch 0:   2%|▏         | 76/3295 [00:47<33:45,  1.59it/s]
Epoch 0:   2%|▏         | 76/3295 [00:47<33:45,  1.59it/s]
Epoch 0:   2%|▏         | 77/3295 [00:48<33:46,  1.59it/s]
Epoch 0:   2%|▏         | 77/3295 [00:48<33:46,  1.59it/s]
Epoch 0:   2%|▏         | 78/3295 [00:48<33:28,  1.60it/s]
Epoch 0:   2%|▏         | 78/3295 [00:48<33:28,  1.60it/s]
Epoch 0:   2%|▏         | 79/3295 [00:49<33:45,  1.59it/s]
Epoch 0:   2%|▏         | 79/3295 [00:49<33:45,  1.59it/s]
Epoch 0:   2%|▏         | 80/3295 [00:50<33:47,  1.59it/s]
Epoch 0:   2%|▏         | 80/3295 [00:50<33:47,  1.59it/s]
Epoch 0:   2%|▏         | 81/3295 [00:50<33:26,  1.60it/s]
Epoch 0:   2%|▏         | 81/3295 [00:50<33:26,  1.60it/s]
Epoch 0:   2%|▏         | 82/3295 [00:51<33:43,  1.59it/s]
Epoch 0:   2%|▏         | 82/3295 [00:51<33:43,  1.59it/s]
Epoch 0:   3%|▎         | 83/3295 [00:52<33:42,  1.59it/s]
Epoch 0:   3%|▎         | 83/3295 [00:52<33:42,  1.59it/s]
Epoch 0:   3%|▎         | 84/3295 [00:52<33:23,  1.60it/s]
Epoch 0:   3%|▎         | 84/3295 [00:52<33:23,  1.60it/s]
Epoch 0:   3%|▎         | 85/3295 [00:53<33:45,  1.58it/s]
Epoch 0:   3%|▎         | 85/3295 [00:53<33:45,  1.58it/s]
Epoch 0:   3%|▎         | 86/3295 [00:54<33:39,  1.59it/s]
Epoch 0:   3%|▎         | 86/3295 [00:54<33:39,  1.59it/s]
Epoch 0:   3%|▎         | 87/3295 [00:54<33:19,  1.60it/s]
Epoch 0:   3%|▎         | 87/3295 [00:54<33:19,  1.60it/s]
Epoch 0:   3%|▎         | 88/3295 [00:55<33:45,  1.58it/s]
Epoch 0:   3%|▎         | 88/3295 [00:55<33:45,  1.58it/s]
Epoch 0:   3%|▎         | 89/3295 [00:55<33:35,  1.59it/s]
Epoch 0:   3%|▎         | 89/3295 [00:55<33:35,  1.59it/s]
Epoch 0:   3%|▎         | 90/3295 [00:56<33:16,  1.61it/s]
Epoch 0:   3%|▎         | 90/3295 [00:56<33:16,  1.61it/s]
Epoch 0:   3%|▎         | 91/3295 [00:57<33:39,  1.59it/s]
Epoch 0:   3%|▎         | 91/3295 [00:57<33:39,  1.59it/s]
Epoch 0:   3%|▎         | 92/3295 [00:57<33:35,  1.59it/s]
Epoch 0:   3%|▎         | 92/3295 [00:57<33:35,  1.59it/s]
Epoch 0:   3%|▎         | 93/3295 [00:57<33:14,  1.61it/s]
Epoch 0:   3%|▎         | 93/3295 [00:57<33:14,  1.61it/s]
Epoch 0:   3%|▎         | 94/3295 [00:59<33:38,  1.59it/s]
Epoch 0:   3%|▎         | 94/3295 [00:59<33:38,  1.59it/s]
Epoch 0:   3%|▎         | 95/3295 [00:59<33:31,  1.59it/s]
Epoch 0:   3%|▎         | 95/3295 [00:59<33:31,  1.59it/s]
Epoch 0:   3%|▎         | 96/3295 [00:59<33:10,  1.61it/s]
Epoch 0:   3%|▎         | 96/3295 [00:59<33:10,  1.61it/s]
Epoch 0:   3%|▎         | 97/3295 [01:01<33:37,  1.58it/s]
Epoch 0:   3%|▎         | 97/3295 [01:01<33:37,  1.58it/s]
Epoch 0:   3%|▎         | 98/3295 [01:01<33:27,  1.59it/s]
Epoch 0:   3%|▎         | 98/3295 [01:01<33:27,  1.59it/s]
Epoch 0:   3%|▎         | 99/3295 [01:01<33:07,  1.61it/s]
Epoch 0:   3%|▎         | 99/3295 [01:01<33:07,  1.61it/s]
Epoch 0:   3%|▎         | 100/3295 [01:03<33:32,  1.59it/s]
Epoch 0:   3%|▎         | 100/3295 [01:03<33:32,  1.59it/s]
Epoch 0:   3%|▎         | 101/3295 [01:03<33:26,  1.59it/s]
Epoch 0:   3%|▎         | 101/3295 [01:03<33:26,  1.59it/s]
Epoch 0:   3%|▎         | 102/3295 [01:03<33:06,  1.61it/s]
Epoch 0:   3%|▎         | 102/3295 [01:03<33:06,  1.61it/s]
Epoch 0:   3%|▎         | 103/3295 [01:04<33:28,  1.59it/s]
Epoch 0:   3%|▎         | 103/3295 [01:04<33:28,  1.59it/s]
Epoch 0:   3%|▎         | 104/3295 [01:05<33:23,  1.59it/s]
Epoch 0:   3%|▎         | 104/3295 [01:05<33:23,  1.59it/s]
Epoch 0:   3%|▎         | 105/3295 [01:05<33:04,  1.61it/s]
Epoch 0:   3%|▎         | 105/3295 [01:05<33:04,  1.61it/s]
Epoch 0:   3%|▎         | 106/3295 [01:06<33:24,  1.59it/s]
Epoch 0:   3%|▎         | 106/3295 [01:06<33:24,  1.59it/s]
Epoch 0:   3%|▎         | 107/3295 [01:07<33:20,  1.59it/s]
Epoch 0:   3%|▎         | 107/3295 [01:07<33:20,  1.59it/s]
Epoch 0:   3%|▎         | 108/3295 [01:07<33:01,  1.61it/s]
Epoch 0:   3%|▎         | 108/3295 [01:07<33:01,  1.61it/s]
Epoch 0:   3%|▎         | 109/3295 [01:08<33:26,  1.59it/s]
Epoch 0:   3%|▎         | 109/3295 [01:08<33:26,  1.59it/s]
Epoch 0:   3%|▎         | 110/3295 [01:09<33:24,  1.59it/s]
Epoch 0:   3%|▎         | 110/3295 [01:09<33:24,  1.59it/s]
Epoch 0:   3%|▎         | 111/3295 [01:09<33:06,  1.60it/s]
Epoch 0:   3%|▎         | 111/3295 [01:09<33:06,  1.60it/s]
Epoch 0:   3%|▎         | 112/3295 [01:10<33:20,  1.59it/s]
Epoch 0:   3%|▎         | 112/3295 [01:10<33:20,  1.59it/s]
Epoch 0:   3%|▎         | 113/3295 [01:11<33:20,  1.59it/s]
Epoch 0:   3%|▎         | 113/3295 [01:11<33:20,  1.59it/s]
Epoch 0:   3%|▎         | 114/3295 [01:11<33:02,  1.60it/s]
Epoch 0:   3%|▎         | 114/3295 [01:11<33:02,  1.60it/s]
Epoch 0:   3%|▎         | 115/3295 [01:12<33:15,  1.59it/s]
Epoch 0:   3%|▎         | 115/3295 [01:12<33:15,  1.59it/s]
Epoch 0:   4%|▎         | 116/3295 [01:12<33:18,  1.59it/s]
Epoch 0:   4%|▎         | 116/3295 [01:12<33:18,  1.59it/s]
Epoch 0:   4%|▎         | 117/3295 [01:12<33:01,  1.60it/s]
Epoch 0:   4%|▎         | 117/3295 [01:12<33:01,  1.60it/s]
Epoch 0:   4%|▎         | 118/3295 [01:14<33:12,  1.59it/s]
Epoch 0:   4%|▎         | 118/3295 [01:14<33:12,  1.59it/s]
Epoch 0:   4%|▎         | 119/3295 [01:14<33:14,  1.59it/s]
Epoch 0:   4%|▎         | 119/3295 [01:14<33:14,  1.59it/s]
Epoch 0:   4%|▎         | 120/3295 [01:14<32:57,  1.61it/s]
Epoch 0:   4%|▎         | 120/3295 [01:14<32:57,  1.61it/s]
Epoch 0:   4%|▎         | 121/3295 [01:15<33:07,  1.60it/s]
Epoch 0:   4%|▎         | 121/3295 [01:15<33:07,  1.60it/s]
Epoch 0:   4%|▎         | 122/3295 [01:16<33:08,  1.60it/s]
Epoch 0:   4%|▎         | 122/3295 [01:16<33:08,  1.60it/s]
Epoch 0:   4%|▎         | 123/3295 [01:16<32:52,  1.61it/s]
Epoch 0:   4%|▎         | 123/3295 [01:16<32:52,  1.61it/s]
Epoch 0:   4%|▍         | 124/3295 [01:17<33:06,  1.60it/s]
Epoch 0:   4%|▍         | 124/3295 [01:17<33:06,  1.60it/s]
Epoch 0:   4%|▍         | 125/3295 [01:18<33:09,  1.59it/s]
Epoch 0:   4%|▍         | 125/3295 [01:18<33:09,  1.59it/s]
Epoch 0:   4%|▍         | 126/3295 [01:18<32:53,  1.61it/s]
Epoch 0:   4%|▍         | 126/3295 [01:18<32:53,  1.61it/s]
Epoch 0:   4%|▍         | 127/3295 [01:19<33:05,  1.60it/s]
Epoch 0:   4%|▍         | 127/3295 [01:19<33:05,  1.60it/s]
Epoch 0:   4%|▍         | 128/3295 [01:20<33:05,  1.60it/s]
Epoch 0:   4%|▍         | 128/3295 [01:20<33:05,  1.60it/s]
Epoch 0:   4%|▍         | 129/3295 [01:20<32:50,  1.61it/s]
Epoch 0:   4%|▍         | 129/3295 [01:20<32:50,  1.61it/s]
Epoch 0:   4%|▍         | 130/3295 [01:21<32:59,  1.60it/s]
Epoch 0:   4%|▍         | 130/3295 [01:21<32:59,  1.60it/s]
Epoch 0:   4%|▍         | 131/3295 [01:22<33:02,  1.60it/s]
Epoch 0:   4%|▍         | 131/3295 [01:22<33:02,  1.60it/s]
Epoch 0:   4%|▍         | 132/3295 [01:22<32:47,  1.61it/s]
Epoch 0:   4%|▍         | 132/3295 [01:22<32:47,  1.61it/s]
Epoch 0:   4%|▍         | 133/3295 [01:23<32:56,  1.60it/s]
Epoch 0:   4%|▍         | 133/3295 [01:23<32:56,  1.60it/s]
Epoch 0:   4%|▍         | 134/3295 [01:23<32:59,  1.60it/s]
Epoch 0:   4%|▍         | 134/3295 [01:23<32:59,  1.60it/s]
Epoch 0:   4%|▍         | 135/3295 [01:23<32:45,  1.61it/s]
Epoch 0:   4%|▍         | 135/3295 [01:23<32:45,  1.61it/s]
Epoch 0:   4%|▍         | 136/3295 [01:25<32:56,  1.60it/s]
Epoch 0:   4%|▍         | 136/3295 [01:25<32:56,  1.60it/s]
Epoch 0:   4%|▍         | 137/3295 [01:25<32:58,  1.60it/s]
Epoch 0:   4%|▍         | 137/3295 [01:25<32:58,  1.60it/s]
Epoch 0:   4%|▍         | 138/3295 [01:25<32:44,  1.61it/s]
Epoch 0:   4%|▍         | 138/3295 [01:25<32:44,  1.61it/s]
Epoch 0:   4%|▍         | 139/3295 [01:26<32:52,  1.60it/s]
Epoch 0:   4%|▍         | 139/3295 [01:26<32:52,  1.60it/s]
Epoch 0:   4%|▍         | 140/3295 [01:27<32:57,  1.60it/s]
Epoch 0:   4%|▍         | 140/3295 [01:27<32:57,  1.60it/s]
Epoch 0:   4%|▍         | 141/3295 [01:27<32:43,  1.61it/s]
Epoch 0:   4%|▍         | 141/3295 [01:27<32:43,  1.61it/s]
Epoch 0:   4%|▍         | 142/3295 [01:28<32:48,  1.60it/s]
Epoch 0:   4%|▍         | 142/3295 [01:28<32:48,  1.60it/s]
Epoch 0:   4%|▍         | 143/3295 [01:29<32:56,  1.59it/s]
Epoch 0:   4%|▍         | 143/3295 [01:29<32:56,  1.59it/s]
Epoch 0:   4%|▍         | 144/3295 [01:29<32:42,  1.61it/s]
Epoch 0:   4%|▍         | 144/3295 [01:29<32:42,  1.61it/s]
Epoch 0:   4%|▍         | 145/3295 [01:30<32:44,  1.60it/s]
Epoch 0:   4%|▍         | 145/3295 [01:30<32:44,  1.60it/s]
Epoch 0:   4%|▍         | 146/3295 [01:31<32:56,  1.59it/s]
Epoch 0:   4%|▍         | 146/3295 [01:31<32:56,  1.59it/s]
Epoch 0:   4%|▍         | 147/3295 [01:31<32:42,  1.60it/s]
Epoch 0:   4%|▍         | 147/3295 [01:31<32:42,  1.60it/s]
Epoch 0:   4%|▍         | 148/3295 [01:32<32:43,  1.60it/s]
Epoch 0:   4%|▍         | 148/3295 [01:32<32:43,  1.60it/s]
Epoch 0:   5%|▍         | 149/3295 [01:33<32:58,  1.59it/s]
Epoch 0:   5%|▍         | 149/3295 [01:33<32:58,  1.59it/s]
Epoch 0:   5%|▍         | 150/3295 [01:33<32:44,  1.60it/s]
Epoch 0:   5%|▍         | 150/3295 [01:33<32:44,  1.60it/s]
Epoch 0:   5%|▍         | 151/3295 [01:34<32:42,  1.60it/s]
Epoch 0:   5%|▍         | 151/3295 [01:34<32:42,  1.60it/s]
Epoch 0:   5%|▍         | 152/3295 [01:35<32:53,  1.59it/s]
Epoch 0:   5%|▍         | 152/3295 [01:35<32:53,  1.59it/s]
Epoch 0:   5%|▍         | 153/3295 [01:35<32:41,  1.60it/s]
Epoch 0:   5%|▍         | 153/3295 [01:35<32:41,  1.60it/s]
Epoch 0:   5%|▍         | 154/3295 [01:36<32:38,  1.60it/s]
Epoch 0:   5%|▍         | 154/3295 [01:36<32:38,  1.60it/s]
Epoch 0:   5%|▍         | 155/3295 [01:37<32:49,  1.59it/s]
Epoch 0:   5%|▍         | 155/3295 [01:37<32:49,  1.59it/s]
Epoch 0:   5%|▍         | 156/3295 [01:37<32:36,  1.60it/s]
Epoch 0:   5%|▍         | 156/3295 [01:37<32:36,  1.60it/s]
Epoch 0:   5%|▍         | 157/3295 [01:37<32:36,  1.60it/s]
Epoch 0:   5%|▍         | 157/3295 [01:37<32:36,  1.60it/s]
Epoch 0:   5%|▍         | 158/3295 [01:39<32:49,  1.59it/s]
Epoch 0:   5%|▍         | 158/3295 [01:39<32:49,  1.59it/s]
Epoch 0:   5%|▍         | 159/3295 [01:39<32:36,  1.60it/s]
Epoch 0:   5%|▍         | 159/3295 [01:39<32:36,  1.60it/s]
Epoch 0:   5%|▍         | 160/3295 [01:39<32:35,  1.60it/s]
Epoch 0:   5%|▍         | 160/3295 [01:39<32:35,  1.60it/s]
Epoch 0:   5%|▍         | 161/3295 [01:41<32:46,  1.59it/s]
Epoch 0:   5%|▍         | 161/3295 [01:41<32:46,  1.59it/s]
Epoch 0:   5%|▍         | 162/3295 [01:41<32:34,  1.60it/s]
Epoch 0:   5%|▍         | 162/3295 [01:41<32:34,  1.60it/s]
Epoch 0:   5%|▍         | 163/3295 [01:41<32:33,  1.60it/s]
Epoch 0:   5%|▍         | 163/3295 [01:41<32:33,  1.60it/s]
Epoch 0:   5%|▍         | 164/3295 [01:42<32:44,  1.59it/s]
Epoch 0:   5%|▍         | 164/3295 [01:42<32:44,  1.59it/s]
Epoch 0:   5%|▌         | 165/3295 [01:42<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 165/3295 [01:42<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 166/3295 [01:43<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 166/3295 [01:43<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 167/3295 [01:44<32:44,  1.59it/s]
Epoch 0:   5%|▌         | 167/3295 [01:44<32:44,  1.59it/s]
Epoch 0:   5%|▌         | 168/3295 [01:44<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 168/3295 [01:44<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 169/3295 [01:45<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 169/3295 [01:45<32:32,  1.60it/s]
Epoch 0:   5%|▌         | 170/3295 [01:46<32:41,  1.59it/s]
Epoch 0:   5%|▌         | 170/3295 [01:46<32:41,  1.59it/s]
Epoch 0:   5%|▌         | 171/3295 [01:46<32:29,  1.60it/s]
Epoch 0:   5%|▌         | 171/3295 [01:46<32:29,  1.60it/s]
Epoch 0:   5%|▌         | 172/3295 [01:47<32:29,  1.60it/s]
Epoch 0:   5%|▌         | 172/3295 [01:47<32:29,  1.60it/s]
Epoch 0:   5%|▌         | 173/3295 [01:48<32:38,  1.59it/s]
Epoch 0:   5%|▌         | 173/3295 [01:48<32:38,  1.59it/s]
Epoch 0:   5%|▌         | 174/3295 [01:48<32:26,  1.60it/s]
Epoch 0:   5%|▌         | 174/3295 [01:48<32:26,  1.60it/s]
Epoch 0:   5%|▌         | 175/3295 [01:49<32:24,  1.60it/s]
Epoch 0:   5%|▌         | 175/3295 [01:49<32:24,  1.60it/s]
Epoch 0:   5%|▌         | 176/3295 [01:50<32:37,  1.59it/s]
Epoch 0:   5%|▌         | 176/3295 [01:50<32:37,  1.59it/s]
Epoch 0:   5%|▌         | 177/3295 [01:50<32:25,  1.60it/s]
Epoch 0:   5%|▌         | 177/3295 [01:50<32:25,  1.60it/s]
Epoch 0:   5%|▌         | 178/3295 [01:50<32:21,  1.61it/s]
Epoch 0:   5%|▌         | 178/3295 [01:50<32:21,  1.61it/s]
Epoch 0:   5%|▌         | 179/3295 [01:52<32:34,  1.59it/s]
Epoch 0:   5%|▌         | 179/3295 [01:52<32:34,  1.59it/s]
Epoch 0:   5%|▌         | 180/3295 [01:52<32:23,  1.60it/s]
Epoch 0:   5%|▌         | 180/3295 [01:52<32:23,  1.60it/s]
Epoch 0:   5%|▌         | 181/3295 [01:52<32:16,  1.61it/s]
Epoch 0:   5%|▌         | 181/3295 [01:52<32:16,  1.61it/s]
Epoch 0:   6%|▌         | 182/3295 [01:54<32:32,  1.59it/s]
Epoch 0:   6%|▌         | 182/3295 [01:54<32:32,  1.59it/s]
Epoch 0:   6%|▌         | 183/3295 [01:54<32:21,  1.60it/s]
Epoch 0:   6%|▌         | 183/3295 [01:54<32:21,  1.60it/s]
Epoch 0:   6%|▌         | 184/3295 [01:54<32:14,  1.61it/s]
Epoch 0:   6%|▌         | 184/3295 [01:54<32:14,  1.61it/s]
Epoch 0:   6%|▌         | 185/3295 [01:55<32:29,  1.60it/s]
Epoch 0:   6%|▌         | 185/3295 [01:55<32:29,  1.60it/s]
Epoch 0:   6%|▌         | 186/3295 [01:56<32:18,  1.60it/s]
Epoch 0:   6%|▌         | 186/3295 [01:56<32:18,  1.60it/s]
Epoch 0:   6%|▌         | 187/3295 [01:56<32:14,  1.61it/s]
Epoch 0:   6%|▌         | 187/3295 [01:56<32:14,  1.61it/s]
Epoch 0:   6%|▌         | 188/3295 [01:57<32:25,  1.60it/s]
Epoch 0:   6%|▌         | 188/3295 [01:57<32:25,  1.60it/s]
Epoch 0:   6%|▌         | 189/3295 [01:57<32:14,  1.61it/s]
Epoch 0:   6%|▌         | 189/3295 [01:57<32:14,  1.61it/s]
Epoch 0:   6%|▌         | 190/3295 [01:58<32:12,  1.61it/s]
Epoch 0:   6%|▌         | 190/3295 [01:58<32:12,  1.61it/s]
Epoch 0:   6%|▌         | 191/3295 [01:59<32:22,  1.60it/s]
Epoch 0:   6%|▌         | 191/3295 [01:59<32:22,  1.60it/s]
Epoch 0:   6%|▌         | 192/3295 [01:59<32:12,  1.61it/s]
Epoch 0:   6%|▌         | 192/3295 [01:59<32:12,  1.61it/s]
Epoch 0:   6%|▌         | 193/3295 [02:00<32:11,  1.61it/s]
Epoch 0:   6%|▌         | 193/3295 [02:00<32:11,  1.61it/s]
Epoch 0:   6%|▌         | 194/3295 [02:01<32:21,  1.60it/s]
Epoch 0:   6%|▌         | 194/3295 [02:01<32:21,  1.60it/s]
Epoch 0:   6%|▌         | 195/3295 [02:01<32:11,  1.60it/s]
Epoch 0:   6%|▌         | 195/3295 [02:01<32:11,  1.60it/s]
Epoch 0:   6%|▌         | 196/3295 [02:01<32:06,  1.61it/s]
Epoch 0:   6%|▌         | 196/3295 [02:01<32:06,  1.61it/s]
Epoch 0:   6%|▌         | 197/3295 [02:03<32:17,  1.60it/s]
Epoch 0:   6%|▌         | 197/3295 [02:03<32:17,  1.60it/s]
Epoch 0:   6%|▌         | 198/3295 [02:03<32:07,  1.61it/s]
Epoch 0:   6%|▌         | 198/3295 [02:03<32:07,  1.61it/s]
Epoch 0:   6%|▌         | 199/3295 [02:03<32:01,  1.61it/s]
Epoch 0:   6%|▌         | 199/3295 [02:03<32:01,  1.61it/s]
Epoch 0:   6%|▌         | 200/3295 [02:05<32:15,  1.60it/s]
Epoch 0:   6%|▌         | 200/3295 [02:05<32:15,  1.60it/s]
Epoch 0:   6%|▌         | 201/3295 [02:05<32:06,  1.61it/s]
Epoch 0:   6%|▌         | 201/3295 [02:05<32:06,  1.61it/s]
Epoch 0:   6%|▌         | 202/3295 [02:05<31:58,  1.61it/s]
Epoch 0:   6%|▌         | 202/3295 [02:05<31:58,  1.61it/s]
Epoch 0:   6%|▌         | 203/3295 [02:06<32:13,  1.60it/s]
Epoch 0:   6%|▌         | 203/3295 [02:06<32:13,  1.60it/s]
Epoch 0:   6%|▌         | 204/3295 [02:06<32:03,  1.61it/s]
Epoch 0:   6%|▌         | 204/3295 [02:06<32:03,  1.61it/s]
Epoch 0:   6%|▌         | 205/3295 [02:07<31:55,  1.61it/s]
Epoch 0:   6%|▌         | 205/3295 [02:07<31:55,  1.61it/s]
Epoch 0:   6%|▋         | 206/3295 [02:08<32:09,  1.60it/s]
Epoch 0:   6%|▋         | 206/3295 [02:08<32:09,  1.60it/s]
Epoch 0:   6%|▋         | 207/3295 [02:08<31:59,  1.61it/s]
Epoch 0:   6%|▋         | 207/3295 [02:08<31:59,  1.61it/s]
Epoch 0:   6%|▋         | 208/3295 [02:09<31:55,  1.61it/s]
Epoch 0:   6%|▋         | 208/3295 [02:09<31:55,  1.61it/s]
Epoch 0:   6%|▋         | 209/3295 [02:10<32:06,  1.60it/s]
Epoch 0:   6%|▋         | 209/3295 [02:10<32:06,  1.60it/s]
Epoch 0:   6%|▋         | 210/3295 [02:10<31:57,  1.61it/s]
Epoch 0:   6%|▋         | 210/3295 [02:10<31:57,  1.61it/s]
Epoch 0:   6%|▋         | 211/3295 [02:10<31:51,  1.61it/s]
Epoch 0:   6%|▋         | 211/3295 [02:10<31:51,  1.61it/s]
Epoch 0:   6%|▋         | 212/3295 [02:12<32:04,  1.60it/s]
Epoch 0:   6%|▋         | 212/3295 [02:12<32:04,  1.60it/s]
Epoch 0:   6%|▋         | 213/3295 [02:12<31:55,  1.61it/s]
Epoch 0:   6%|▋         | 213/3295 [02:12<31:55,  1.61it/s]
Epoch 0:   6%|▋         | 214/3295 [02:12<31:50,  1.61it/s]
Epoch 0:   6%|▋         | 214/3295 [02:12<31:50,  1.61it/s]
Epoch 0:   7%|▋         | 215/3295 [02:14<32:02,  1.60it/s]
Epoch 0:   7%|▋         | 215/3295 [02:14<32:02,  1.60it/s]
Epoch 0:   7%|▋         | 216/3295 [02:14<31:53,  1.61it/s]
Epoch 0:   7%|▋         | 216/3295 [02:14<31:53,  1.61it/s]
Epoch 0:   7%|▋         | 217/3295 [02:14<31:49,  1.61it/s]
Epoch 0:   7%|▋         | 217/3295 [02:14<31:49,  1.61it/s]
Epoch 0:   7%|▋         | 218/3295 [02:16<32:01,  1.60it/s]
Epoch 0:   7%|▋         | 218/3295 [02:16<32:01,  1.60it/s]
Epoch 0:   7%|▋         | 219/3295 [02:16<31:52,  1.61it/s]
Epoch 0:   7%|▋         | 219/3295 [02:16<31:52,  1.61it/s]
Epoch 0:   7%|▋         | 220/3295 [02:16<31:43,  1.62it/s]
Epoch 0:   7%|▋         | 220/3295 [02:16<31:43,  1.62it/s]
Epoch 0:   7%|▋         | 221/3295 [02:17<31:56,  1.60it/s]
Epoch 0:   7%|▋         | 221/3295 [02:17<31:56,  1.60it/s]
Epoch 0:   7%|▋         | 222/3295 [02:17<31:47,  1.61it/s]
Epoch 0:   7%|▋         | 222/3295 [02:17<31:47,  1.61it/s]
Epoch 0:   7%|▋         | 223/3295 [02:18<31:43,  1.61it/s]
Epoch 0:   7%|▋         | 223/3295 [02:18<31:43,  1.61it/s]
Epoch 0:   7%|▋         | 224/3295 [02:19<31:56,  1.60it/s]
Epoch 0:   7%|▋         | 224/3295 [02:19<31:56,  1.60it/s]
Epoch 0:   7%|▋         | 225/3295 [02:19<31:47,  1.61it/s]
Epoch 0:   7%|▋         | 225/3295 [02:19<31:47,  1.61it/s]
Epoch 0:   7%|▋         | 226/3295 [02:20<31:43,  1.61it/s]
Epoch 0:   7%|▋         | 226/3295 [02:20<31:43,  1.61it/s]
Epoch 0:   7%|▋         | 227/3295 [02:21<31:53,  1.60it/s]
Epoch 0:   7%|▋         | 227/3295 [02:21<31:53,  1.60it/s]
Epoch 0:   7%|▋         | 228/3295 [02:21<31:44,  1.61it/s]
Epoch 0:   7%|▋         | 228/3295 [02:21<31:44,  1.61it/s]
Epoch 0:   7%|▋         | 229/3295 [02:22<31:41,  1.61it/s]
Epoch 0:   7%|▋         | 229/3295 [02:22<31:41,  1.61it/s]
Epoch 0:   7%|▋         | 230/3295 [02:23<31:49,  1.60it/s]
Epoch 0:   7%|▋         | 230/3295 [02:23<31:49,  1.60it/s]
Epoch 0:   7%|▋         | 231/3295 [02:23<31:41,  1.61it/s]
Epoch 0:   7%|▋         | 231/3295 [02:23<31:41,  1.61it/s]
Epoch 0:   7%|▋         | 232/3295 [02:23<31:40,  1.61it/s]
Epoch 0:   7%|▋         | 232/3295 [02:23<31:40,  1.61it/s]
Epoch 0:   7%|▋         | 233/3295 [02:25<31:46,  1.61it/s]
Epoch 0:   7%|▋         | 233/3295 [02:25<31:46,  1.61it/s]
Epoch 0:   7%|▋         | 234/3295 [02:25<31:38,  1.61it/s]
Epoch 0:   7%|▋         | 234/3295 [02:25<31:38,  1.61it/s]
Epoch 0:   7%|▋         | 235/3295 [02:25<31:38,  1.61it/s]
Epoch 0:   7%|▋         | 235/3295 [02:25<31:38,  1.61it/s]
Epoch 0:   7%|▋         | 236/3295 [02:27<31:45,  1.61it/s]
Epoch 0:   7%|▋         | 236/3295 [02:27<31:45,  1.61it/s]
Epoch 0:   7%|▋         | 237/3295 [02:27<31:37,  1.61it/s]
Epoch 0:   7%|▋         | 237/3295 [02:27<31:37,  1.61it/s]
Epoch 0:   7%|▋         | 238/3295 [02:27<31:36,  1.61it/s]
Epoch 0:   7%|▋         | 238/3295 [02:27<31:36,  1.61it/s]
Epoch 0:   7%|▋         | 239/3295 [02:28<31:44,  1.60it/s]
Epoch 0:   7%|▋         | 239/3295 [02:28<31:44,  1.60it/s]
Epoch 0:   7%|▋         | 240/3295 [02:28<31:36,  1.61it/s]
Epoch 0:   7%|▋         | 240/3295 [02:28<31:36,  1.61it/s]
Epoch 0:   7%|▋         | 241/3295 [02:29<31:34,  1.61it/s]
Epoch 0:   7%|▋         | 241/3295 [02:29<31:34,  1.61it/s]
Epoch 0:   7%|▋         | 242/3295 [02:30<31:41,  1.61it/s]
Epoch 0:   7%|▋         | 242/3295 [02:30<31:41,  1.61it/s]
Epoch 0:   7%|▋         | 243/3295 [02:30<31:33,  1.61it/s]
Epoch 0:   7%|▋         | 243/3295 [02:30<31:33,  1.61it/s]
Epoch 0:   7%|▋         | 244/3295 [02:31<31:32,  1.61it/s]
Epoch 0:   7%|▋         | 244/3295 [02:31<31:32,  1.61it/s]
Epoch 0:   7%|▋         | 245/3295 [02:32<31:39,  1.61it/s]
Epoch 0:   7%|▋         | 245/3295 [02:32<31:39,  1.61it/s]
Epoch 0:   7%|▋         | 246/3295 [02:32<31:31,  1.61it/s]
Epoch 0:   7%|▋         | 246/3295 [02:32<31:31,  1.61it/s]
Epoch 0:   7%|▋         | 247/3295 [02:33<31:29,  1.61it/s]
Epoch 0:   7%|▋         | 247/3295 [02:33<31:29,  1.61it/s]
Epoch 0:   8%|▊         | 248/3295 [02:34<31:39,  1.60it/s]
Epoch 0:   8%|▊         | 248/3295 [02:34<31:39,  1.60it/s]
Epoch 0:   8%|▊         | 249/3295 [02:34<31:31,  1.61it/s]
Epoch 0:   8%|▊         | 249/3295 [02:34<31:31,  1.61it/s]
Epoch 0:   8%|▊         | 250/3295 [02:34<31:27,  1.61it/s]
Epoch 0:   8%|▊         | 250/3295 [02:34<31:27,  1.61it/s]
Epoch 0:   8%|▊         | 251/3295 [02:36<31:36,  1.60it/s]
Epoch 0:   8%|▊         | 251/3295 [02:36<31:36,  1.60it/s]
Epoch 0:   8%|▊         | 252/3295 [02:36<31:29,  1.61it/s]
Epoch 0:   8%|▊         | 252/3295 [02:36<31:29,  1.61it/s]
Epoch 0:   8%|▊         | 253/3295 [02:36<31:24,  1.61it/s]
Epoch 0:   8%|▊         | 253/3295 [02:36<31:24,  1.61it/s]
Epoch 0:   8%|▊         | 254/3295 [02:38<31:34,  1.61it/s]
Epoch 0:   8%|▊         | 254/3295 [02:38<31:34,  1.61it/s]
Epoch 0:   8%|▊         | 255/3295 [02:38<31:26,  1.61it/s]
Epoch 0:   8%|▊         | 255/3295 [02:38<31:26,  1.61it/s]
Epoch 0:   8%|▊         | 256/3295 [02:38<31:20,  1.62it/s]
Epoch 0:   8%|▊         | 256/3295 [02:38<31:20,  1.62it/s]
Epoch 0:   8%|▊         | 257/3295 [02:40<31:34,  1.60it/s]
Epoch 0:   8%|▊         | 257/3295 [02:40<31:34,  1.60it/s]
Epoch 0:   8%|▊         | 258/3295 [02:40<31:27,  1.61it/s]
Epoch 0:   8%|▊         | 258/3295 [02:40<31:27,  1.61it/s]
Epoch 0:   8%|▊         | 259/3295 [02:40<31:20,  1.61it/s]
Epoch 0:   8%|▊         | 259/3295 [02:40<31:20,  1.61it/s]
Epoch 0:   8%|▊         | 260/3295 [02:42<31:32,  1.60it/s]
Epoch 0:   8%|▊         | 260/3295 [02:42<31:32,  1.60it/s]
Epoch 0:   8%|▊         | 261/3295 [02:42<31:24,  1.61it/s]
Epoch 0:   8%|▊         | 261/3295 [02:42<31:24,  1.61it/s]
Epoch 0:   8%|▊         | 262/3295 [02:42<31:17,  1.62it/s]
Epoch 0:   8%|▊         | 262/3295 [02:42<31:17,  1.62it/s]
Epoch 0:   8%|▊         | 263/3295 [02:44<31:31,  1.60it/s]
Epoch 0:   8%|▊         | 263/3295 [02:44<31:31,  1.60it/s]
Epoch 0:   8%|▊         | 264/3295 [02:44<31:23,  1.61it/s]
Epoch 0:   8%|▊         | 264/3295 [02:44<31:23,  1.61it/s]
Epoch 0:   8%|▊         | 265/3295 [02:44<31:16,  1.61it/s]
Epoch 0:   8%|▊         | 265/3295 [02:44<31:16,  1.61it/s]
Epoch 0:   8%|▊         | 266/3295 [02:45<31:29,  1.60it/s]
Epoch 0:   8%|▊         | 266/3295 [02:45<31:29,  1.60it/s]
Epoch 0:   8%|▊         | 267/3295 [02:45<31:22,  1.61it/s]
Epoch 0:   8%|▊         | 267/3295 [02:45<31:22,  1.61it/s]
Epoch 0:   8%|▊         | 268/3295 [02:46<31:15,  1.61it/s]
Epoch 0:   8%|▊         | 268/3295 [02:46<31:15,  1.61it/s]
Epoch 0:   8%|▊         | 269/3295 [02:47<31:27,  1.60it/s]
Epoch 0:   8%|▊         | 269/3295 [02:47<31:27,  1.60it/s]
Epoch 0:   8%|▊         | 270/3295 [02:47<31:19,  1.61it/s]
Epoch 0:   8%|▊         | 270/3295 [02:47<31:19,  1.61it/s]
Epoch 0:   8%|▊         | 271/3295 [02:47<31:12,  1.61it/s]
Epoch 0:   8%|▊         | 271/3295 [02:47<31:12,  1.61it/s]
Epoch 0:   8%|▊         | 272/3295 [02:49<31:25,  1.60it/s]
Epoch 0:   8%|▊         | 272/3295 [02:49<31:25,  1.60it/s]
Epoch 0:   8%|▊         | 273/3295 [02:49<31:18,  1.61it/s]
Epoch 0:   8%|▊         | 273/3295 [02:49<31:18,  1.61it/s]
Epoch 0:   8%|▊         | 274/3295 [02:49<31:11,  1.61it/s]
Epoch 0:   8%|▊         | 274/3295 [02:49<31:11,  1.61it/s]
Epoch 0:   8%|▊         | 275/3295 [02:51<31:24,  1.60it/s]
Epoch 0:   8%|▊         | 275/3295 [02:51<31:24,  1.60it/s]
Epoch 0:   8%|▊         | 276/3295 [02:51<31:16,  1.61it/s]
Epoch 0:   8%|▊         | 276/3295 [02:51<31:16,  1.61it/s]
Epoch 0:   8%|▊         | 277/3295 [02:51<31:09,  1.61it/s]
Epoch 0:   8%|▊         | 277/3295 [02:51<31:09,  1.61it/s]
Epoch 0:   8%|▊         | 278/3295 [02:53<31:21,  1.60it/s]
Epoch 0:   8%|▊         | 278/3295 [02:53<31:21,  1.60it/s]
Epoch 0:   8%|▊         | 279/3295 [02:53<31:14,  1.61it/s]
Epoch 0:   8%|▊         | 279/3295 [02:53<31:14,  1.61it/s]
Epoch 0:   8%|▊         | 280/3295 [02:53<31:07,  1.61it/s]
Epoch 0:   8%|▊         | 280/3295 [02:53<31:07,  1.61it/s]
Epoch 0:   9%|▊         | 281/3295 [02:55<31:19,  1.60it/s]
Epoch 0:   9%|▊         | 281/3295 [02:55<31:19,  1.60it/s]
Epoch 0:   9%|▊         | 282/3295 [02:55<31:12,  1.61it/s]
Epoch 0:   9%|▊         | 282/3295 [02:55<31:12,  1.61it/s]
Epoch 0:   9%|▊         | 283/3295 [02:55<31:05,  1.61it/s]
Epoch 0:   9%|▊         | 283/3295 [02:55<31:05,  1.61it/s]
Epoch 0:   9%|▊         | 284/3295 [02:57<31:19,  1.60it/s]
Epoch 0:   9%|▊         | 284/3295 [02:57<31:19,  1.60it/s]
Epoch 0:   9%|▊         | 285/3295 [02:57<31:12,  1.61it/s]
Epoch 0:   9%|▊         | 285/3295 [02:57<31:12,  1.61it/s]
Epoch 0:   9%|▊         | 286/3295 [02:57<31:06,  1.61it/s]
Epoch 0:   9%|▊         | 286/3295 [02:57<31:06,  1.61it/s]
Epoch 0:   9%|▊         | 287/3295 [02:59<31:18,  1.60it/s]
Epoch 0:   9%|▊         | 287/3295 [02:59<31:18,  1.60it/s]
Epoch 0:   9%|▊         | 288/3295 [02:59<31:11,  1.61it/s]
Epoch 0:   9%|▊         | 288/3295 [02:59<31:11,  1.61it/s]
Epoch 0:   9%|▉         | 289/3295 [02:59<31:04,  1.61it/s]
Epoch 0:   9%|▉         | 289/3295 [02:59<31:04,  1.61it/s]
Epoch 0:   9%|▉         | 290/3295 [03:01<31:15,  1.60it/s]
Epoch 0:   9%|▉         | 290/3295 [03:01<31:15,  1.60it/s]
Epoch 0:   9%|▉         | 291/3295 [03:01<31:08,  1.61it/s]
Epoch 0:   9%|▉         | 291/3295 [03:01<31:08,  1.61it/s]
Epoch 0:   9%|▉         | 292/3295 [03:01<31:02,  1.61it/s]
Epoch 0:   9%|▉         | 292/3295 [03:01<31:02,  1.61it/s]
Epoch 0:   9%|▉         | 293/3295 [03:02<31:14,  1.60it/s]
Epoch 0:   9%|▉         | 293/3295 [03:02<31:14,  1.60it/s]
Epoch 0:   9%|▉         | 294/3295 [03:02<31:07,  1.61it/s]
Epoch 0:   9%|▉         | 294/3295 [03:02<31:07,  1.61it/s]
Epoch 0:   9%|▉         | 295/3295 [03:03<31:01,  1.61it/s]
Epoch 0:   9%|▉         | 295/3295 [03:03<31:01,  1.61it/s]
Epoch 0:   9%|▉         | 296/3295 [03:04<31:12,  1.60it/s]
Epoch 0:   9%|▉         | 296/3295 [03:04<31:12,  1.60it/s]
Epoch 0:   9%|▉         | 297/3295 [03:04<31:05,  1.61it/s]
Epoch 0:   9%|▉         | 297/3295 [03:04<31:05,  1.61it/s]
Epoch 0:   9%|▉         | 298/3295 [03:04<30:58,  1.61it/s]
Epoch 0:   9%|▉         | 298/3295 [03:04<30:58,  1.61it/s]
Epoch 0:   9%|▉         | 299/3295 [03:06<31:09,  1.60it/s]
Epoch 0:   9%|▉         | 299/3295 [03:06<31:09,  1.60it/s]
Epoch 0:   9%|▉         | 300/3295 [03:06<31:03,  1.61it/s]
Epoch 0:   9%|▉         | 300/3295 [03:06<31:03,  1.61it/s]
Epoch 0:   9%|▉         | 301/3295 [03:06<30:56,  1.61it/s]
Epoch 0:   9%|▉         | 301/3295 [03:06<30:56,  1.61it/s]
Epoch 0:   9%|▉         | 302/3295 [03:08<31:09,  1.60it/s]
Epoch 0:   9%|▉         | 302/3295 [03:08<31:09,  1.60it/s]
Epoch 0:   9%|▉         | 303/3295 [03:08<31:02,  1.61it/s]
Epoch 0:   9%|▉         | 303/3295 [03:08<31:02,  1.61it/s]
Epoch 0:   9%|▉         | 304/3295 [03:08<30:56,  1.61it/s]
Epoch 0:   9%|▉         | 304/3295 [03:08<30:56,  1.61it/s]
Epoch 0:   9%|▉         | 305/3295 [03:10<31:08,  1.60it/s]
Epoch 0:   9%|▉         | 305/3295 [03:10<31:08,  1.60it/s]
Epoch 0:   9%|▉         | 306/3295 [03:10<31:02,  1.61it/s]
Epoch 0:   9%|▉         | 306/3295 [03:10<31:02,  1.61it/s]
Epoch 0:   9%|▉         | 307/3295 [03:10<30:55,  1.61it/s]
Epoch 0:   9%|▉         | 307/3295 [03:10<30:55,  1.61it/s]
Epoch 0:   9%|▉         | 308/3295 [03:12<31:05,  1.60it/s]
Epoch 0:   9%|▉         | 308/3295 [03:12<31:05,  1.60it/s]
Epoch 0:   9%|▉         | 309/3295 [03:12<30:58,  1.61it/s]
Epoch 0:   9%|▉         | 309/3295 [03:12<30:58,  1.61it/s]
Epoch 0:   9%|▉         | 310/3295 [03:12<30:52,  1.61it/s]
Epoch 0:   9%|▉         | 310/3295 [03:12<30:52,  1.61it/s]
Epoch 0:   9%|▉         | 311/3295 [03:14<31:03,  1.60it/s]
Epoch 0:   9%|▉         | 311/3295 [03:14<31:03,  1.60it/s]
Epoch 0:   9%|▉         | 312/3295 [03:14<30:57,  1.61it/s]
Epoch 0:   9%|▉         | 312/3295 [03:14<30:57,  1.61it/s]
Epoch 0:   9%|▉         | 313/3295 [03:14<30:51,  1.61it/s]
Epoch 0:   9%|▉         | 313/3295 [03:14<30:51,  1.61it/s]
Epoch 0:  10%|▉         | 314/3295 [03:16<31:00,  1.60it/s]
Epoch 0:  10%|▉         | 314/3295 [03:16<31:00,  1.60it/s]
Epoch 0:  10%|▉         | 315/3295 [03:16<30:54,  1.61it/s]
Epoch 0:  10%|▉         | 315/3295 [03:16<30:54,  1.61it/s]
Epoch 0:  10%|▉         | 316/3295 [03:16<30:48,  1.61it/s]
Epoch 0:  10%|▉         | 316/3295 [03:16<30:48,  1.61it/s]
Epoch 0:  10%|▉         | 317/3295 [03:17<30:59,  1.60it/s]
Epoch 0:  10%|▉         | 317/3295 [03:17<30:59,  1.60it/s]
Epoch 0:  10%|▉         | 318/3295 [03:17<30:52,  1.61it/s]
Epoch 0:  10%|▉         | 318/3295 [03:17<30:52,  1.61it/s]
Epoch 0:  10%|▉         | 319/3295 [03:17<30:46,  1.61it/s]
Epoch 0:  10%|▉         | 319/3295 [03:17<30:46,  1.61it/s]
Epoch 0:  10%|▉         | 320/3295 [03:19<30:56,  1.60it/s]
Epoch 0:  10%|▉         | 320/3295 [03:19<30:56,  1.60it/s]
Epoch 0:  10%|▉         | 321/3295 [03:19<30:50,  1.61it/s]
Epoch 0:  10%|▉         | 321/3295 [03:19<30:50,  1.61it/s]
Epoch 0:  10%|▉         | 322/3295 [03:19<30:44,  1.61it/s]
Epoch 0:  10%|▉         | 322/3295 [03:19<30:44,  1.61it/s]
Epoch 0:  10%|▉         | 323/3295 [03:21<30:55,  1.60it/s]
Epoch 0:  10%|▉         | 323/3295 [03:21<30:55,  1.60it/s]
Epoch 0:  10%|▉         | 324/3295 [03:21<30:49,  1.61it/s]
Epoch 0:  10%|▉         | 324/3295 [03:21<30:49,  1.61it/s]
Epoch 0:  10%|▉         | 325/3295 [03:21<30:43,  1.61it/s]
Epoch 0:  10%|▉         | 325/3295 [03:21<30:43,  1.61it/s]
Epoch 0:  10%|▉         | 326/3295 [03:23<30:53,  1.60it/s]
Epoch 0:  10%|▉         | 326/3295 [03:23<30:53,  1.60it/s]
Epoch 0:  10%|▉         | 327/3295 [03:23<30:47,  1.61it/s]
Epoch 0:  10%|▉         | 327/3295 [03:23<30:47,  1.61it/s]
Epoch 0:  10%|▉         | 328/3295 [03:23<30:41,  1.61it/s]
Epoch 0:  10%|▉         | 328/3295 [03:23<30:41,  1.61it/s]
Epoch 0:  10%|▉         | 329/3295 [03:25<30:51,  1.60it/s]
Epoch 0:  10%|▉         | 329/3295 [03:25<30:51,  1.60it/s]
Epoch 0:  10%|█         | 330/3295 [03:25<30:45,  1.61it/s]
Epoch 0:  10%|█         | 330/3295 [03:25<30:45,  1.61it/s]
Epoch 0:  10%|█         | 331/3295 [03:25<30:39,  1.61it/s]
Epoch 0:  10%|█         | 331/3295 [03:25<30:39,  1.61it/s]
Epoch 0:  10%|█         | 332/3295 [03:27<30:49,  1.60it/s]
Epoch 0:  10%|█         | 332/3295 [03:27<30:49,  1.60it/s]
Epoch 0:  10%|█         | 333/3295 [03:27<30:43,  1.61it/s]
Epoch 0:  10%|█         | 333/3295 [03:27<30:43,  1.61it/s]
Epoch 0:  10%|█         | 334/3295 [03:27<30:37,  1.61it/s]
Epoch 0:  10%|█         | 334/3295 [03:27<30:37,  1.61it/s]
Epoch 0:  10%|█         | 335/3295 [03:29<30:47,  1.60it/s]
Epoch 0:  10%|█         | 335/3295 [03:29<30:47,  1.60it/s]
Epoch 0:  10%|█         | 336/3295 [03:29<30:41,  1.61it/s]
Epoch 0:  10%|█         | 336/3295 [03:29<30:41,  1.61it/s]
Epoch 0:  10%|█         | 337/3295 [03:29<30:35,  1.61it/s]
Epoch 0:  10%|█         | 337/3295 [03:29<30:35,  1.61it/s]
Epoch 0:  10%|█         | 338/3295 [03:30<30:45,  1.60it/s]
Epoch 0:  10%|█         | 338/3295 [03:30<30:45,  1.60it/s]
Epoch 0:  10%|█         | 339/3295 [03:30<30:39,  1.61it/s]
Epoch 0:  10%|█         | 339/3295 [03:30<30:39,  1.61it/s]
Epoch 0:  10%|█         | 340/3295 [03:30<30:33,  1.61it/s]
Epoch 0:  10%|█         | 340/3295 [03:30<30:33,  1.61it/s]
Epoch 0:  10%|█         | 341/3295 [03:32<30:43,  1.60it/s]
Epoch 0:  10%|█         | 341/3295 [03:32<30:43,  1.60it/s]
Epoch 0:  10%|█         | 342/3295 [03:32<30:37,  1.61it/s]
Epoch 0:  10%|█         | 342/3295 [03:32<30:37,  1.61it/s]
Epoch 0:  10%|█         | 343/3295 [03:32<30:32,  1.61it/s]
Epoch 0:  10%|█         | 343/3295 [03:32<30:32,  1.61it/s]
Epoch 0:  10%|█         | 344/3295 [03:34<30:41,  1.60it/s]
Epoch 0:  10%|█         | 344/3295 [03:34<30:41,  1.60it/s]
Epoch 0:  10%|█         | 345/3295 [03:34<30:36,  1.61it/s]
Epoch 0:  10%|█         | 345/3295 [03:34<30:36,  1.61it/s]
Epoch 0:  11%|█         | 346/3295 [03:34<30:30,  1.61it/s]
Epoch 0:  11%|█         | 346/3295 [03:34<30:30,  1.61it/s]
Epoch 0:  11%|█         | 347/3295 [03:36<30:38,  1.60it/s]
Epoch 0:  11%|█         | 347/3295 [03:36<30:38,  1.60it/s]
Epoch 0:  11%|█         | 348/3295 [03:36<30:32,  1.61it/s]
Epoch 0:  11%|█         | 348/3295 [03:36<30:32,  1.61it/s]
Epoch 0:  11%|█         | 349/3295 [03:36<30:26,  1.61it/s]
Epoch 0:  11%|█         | 349/3295 [03:36<30:26,  1.61it/s]
Epoch 0:  11%|█         | 350/3295 [03:38<30:36,  1.60it/s]
Epoch 0:  11%|█         | 350/3295 [03:38<30:36,  1.60it/s]
Epoch 0:  11%|█         | 351/3295 [03:38<30:30,  1.61it/s]
Epoch 0:  11%|█         | 351/3295 [03:38<30:30,  1.61it/s]
Epoch 0:  11%|█         | 352/3295 [03:38<30:25,  1.61it/s]
Epoch 0:  11%|█         | 352/3295 [03:38<30:25,  1.61it/s]
Epoch 0:  11%|█         | 353/3295 [03:39<30:32,  1.61it/s]
Epoch 0:  11%|█         | 353/3295 [03:39<30:32,  1.61it/s]
Epoch 0:  11%|█         | 354/3295 [03:39<30:27,  1.61it/s]
Epoch 0:  11%|█         | 354/3295 [03:39<30:27,  1.61it/s]
Epoch 0:  11%|█         | 355/3295 [03:39<30:21,  1.61it/s]
Epoch 0:  11%|█         | 355/3295 [03:39<30:21,  1.61it/s]
Epoch 0:  11%|█         | 356/3295 [03:41<30:29,  1.61it/s]
Epoch 0:  11%|█         | 356/3295 [03:41<30:29,  1.61it/s]
Epoch 0:  11%|█         | 357/3295 [03:41<30:24,  1.61it/s]
Epoch 0:  11%|█         | 357/3295 [03:41<30:24,  1.61it/s]
Epoch 0:  11%|█         | 358/3295 [03:41<30:18,  1.61it/s]
Epoch 0:  11%|█         | 358/3295 [03:41<30:18,  1.61it/s]
Epoch 0:  11%|█         | 359/3295 [03:43<30:27,  1.61it/s]
Epoch 0:  11%|█         | 359/3295 [03:43<30:27,  1.61it/s]
Epoch 0:  11%|█         | 360/3295 [03:43<30:22,  1.61it/s]
Epoch 0:  11%|█         | 360/3295 [03:43<30:22,  1.61it/s]
Epoch 0:  11%|█         | 361/3295 [03:43<30:16,  1.61it/s]
Epoch 0:  11%|█         | 361/3295 [03:43<30:16,  1.61it/s]
Epoch 0:  11%|█         | 362/3295 [03:45<30:25,  1.61it/s]
Epoch 0:  11%|█         | 362/3295 [03:45<30:25,  1.61it/s]
Epoch 0:  11%|█         | 363/3295 [03:45<30:19,  1.61it/s]
Epoch 0:  11%|█         | 363/3295 [03:45<30:19,  1.61it/s]
Epoch 0:  11%|█         | 364/3295 [03:45<30:14,  1.62it/s]
Epoch 0:  11%|█         | 364/3295 [03:45<30:14,  1.62it/s]
Epoch 0:  11%|█         | 365/3295 [03:47<30:23,  1.61it/s]
Epoch 0:  11%|█         | 365/3295 [03:47<30:23,  1.61it/s]
Epoch 0:  11%|█         | 366/3295 [03:47<30:18,  1.61it/s]
Epoch 0:  11%|█         | 366/3295 [03:47<30:18,  1.61it/s]
Epoch 0:  11%|█         | 367/3295 [03:47<30:13,  1.61it/s]
Epoch 0:  11%|█         | 367/3295 [03:47<30:13,  1.61it/s]
Epoch 0:  11%|█         | 368/3295 [03:49<30:23,  1.61it/s]
Epoch 0:  11%|█         | 368/3295 [03:49<30:23,  1.61it/s]
Epoch 0:  11%|█         | 369/3295 [03:49<30:18,  1.61it/s]
Epoch 0:  11%|█         | 369/3295 [03:49<30:18,  1.61it/s]
Epoch 0:  11%|█         | 370/3295 [03:49<30:12,  1.61it/s]
Epoch 0:  11%|█         | 370/3295 [03:49<30:12,  1.61it/s]
Epoch 0:  11%|█▏        | 371/3295 [03:51<30:21,  1.61it/s]
Epoch 0:  11%|█▏        | 371/3295 [03:51<30:21,  1.61it/s]
Epoch 0:  11%|█▏        | 372/3295 [03:51<30:16,  1.61it/s]
Epoch 0:  11%|█▏        | 372/3295 [03:51<30:16,  1.61it/s]
Epoch 0:  11%|█▏        | 373/3295 [03:51<30:10,  1.61it/s]
Epoch 0:  11%|█▏        | 373/3295 [03:51<30:10,  1.61it/s]
Epoch 0:  11%|█▏        | 374/3295 [03:53<30:19,  1.61it/s]
Epoch 0:  11%|█▏        | 374/3295 [03:53<30:19,  1.61it/s]
Epoch 0:  11%|█▏        | 375/3295 [03:53<30:14,  1.61it/s]
Epoch 0:  11%|█▏        | 375/3295 [03:53<30:14,  1.61it/s]
Epoch 0:  11%|█▏        | 376/3295 [03:53<30:09,  1.61it/s]
Epoch 0:  11%|█▏        | 376/3295 [03:53<30:09,  1.61it/s]
Epoch 0:  11%|█▏        | 377/3295 [03:54<30:17,  1.61it/s]
Epoch 0:  11%|█▏        | 377/3295 [03:54<30:17,  1.61it/s]
Epoch 0:  11%|█▏        | 378/3295 [03:54<30:12,  1.61it/s]
Epoch 0:  11%|█▏        | 378/3295 [03:54<30:12,  1.61it/s]
Epoch 0:  12%|█▏        | 379/3295 [03:54<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 379/3295 [03:54<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 380/3295 [03:56<30:16,  1.60it/s]
Epoch 0:  12%|█▏        | 380/3295 [03:56<30:16,  1.60it/s]
Epoch 0:  12%|█▏        | 381/3295 [03:56<30:11,  1.61it/s]
Epoch 0:  12%|█▏        | 381/3295 [03:56<30:11,  1.61it/s]
Epoch 0:  12%|█▏        | 382/3295 [03:56<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 382/3295 [03:56<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 383/3295 [03:58<30:15,  1.60it/s]
Epoch 0:  12%|█▏        | 383/3295 [03:58<30:15,  1.60it/s]
Epoch 0:  12%|█▏        | 384/3295 [03:58<30:10,  1.61it/s]
Epoch 0:  12%|█▏        | 384/3295 [03:58<30:10,  1.61it/s]
Epoch 0:  12%|█▏        | 385/3295 [03:58<30:05,  1.61it/s]
Epoch 0:  12%|█▏        | 385/3295 [03:58<30:05,  1.61it/s]
Epoch 0:  12%|█▏        | 386/3295 [04:00<30:13,  1.60it/s]
Epoch 0:  12%|█▏        | 386/3295 [04:00<30:13,  1.60it/s]
Epoch 0:  12%|█▏        | 387/3295 [04:00<30:08,  1.61it/s]
Epoch 0:  12%|█▏        | 387/3295 [04:00<30:08,  1.61it/s]
Epoch 0:  12%|█▏        | 388/3295 [04:00<30:03,  1.61it/s]
Epoch 0:  12%|█▏        | 388/3295 [04:00<30:03,  1.61it/s]
Epoch 0:  12%|█▏        | 389/3295 [04:02<30:11,  1.60it/s]
Epoch 0:  12%|█▏        | 389/3295 [04:02<30:11,  1.60it/s]
Epoch 0:  12%|█▏        | 390/3295 [04:02<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 390/3295 [04:02<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 391/3295 [04:02<30:01,  1.61it/s]
Epoch 0:  12%|█▏        | 391/3295 [04:02<30:01,  1.61it/s]
Epoch 0:  12%|█▏        | 392/3295 [04:04<30:08,  1.60it/s]
Epoch 0:  12%|█▏        | 392/3295 [04:04<30:08,  1.60it/s]
Epoch 0:  12%|█▏        | 393/3295 [04:04<30:03,  1.61it/s]
Epoch 0:  12%|█▏        | 393/3295 [04:04<30:03,  1.61it/s]
Epoch 0:  12%|█▏        | 394/3295 [04:04<29:59,  1.61it/s]
Epoch 0:  12%|█▏        | 394/3295 [04:04<29:59,  1.61it/s]
Epoch 0:  12%|█▏        | 395/3295 [04:06<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 395/3295 [04:06<30:06,  1.61it/s]
Epoch 0:  12%|█▏        | 396/3295 [04:06<30:01,  1.61it/s]
Epoch 0:  12%|█▏        | 396/3295 [04:06<30:01,  1.61it/s]
Epoch 0:  12%|█▏        | 397/3295 [04:06<29:56,  1.61it/s]
Epoch 0:  12%|█▏        | 397/3295 [04:06<29:56,  1.61it/s]
Epoch 0:  12%|█▏        | 398/3295 [04:08<30:05,  1.60it/s]
Epoch 0:  12%|█▏        | 398/3295 [04:08<30:05,  1.60it/s]
Epoch 0:  12%|█▏        | 399/3295 [04:08<30:00,  1.61it/s]
Epoch 0:  12%|█▏        | 399/3295 [04:08<30:00,  1.61it/s]
Epoch 0:  12%|█▏        | 400/3295 [04:08<29:55,  1.61it/s]
Epoch 0:  12%|█▏        | 400/3295 [04:08<29:55,  1.61it/s]
Epoch 0:  12%|█▏        | 401/3295 [04:09<30:03,  1.61it/s]
Epoch 0:  12%|█▏        | 401/3295 [04:09<30:03,  1.61it/s]
Epoch 0:  12%|█▏        | 402/3295 [04:09<29:58,  1.61it/s]
Epoch 0:  12%|█▏        | 402/3295 [04:09<29:58,  1.61it/s]
Epoch 0:  12%|█▏        | 403/3295 [04:09<29:53,  1.61it/s]
Epoch 0:  12%|█▏        | 403/3295 [04:09<29:53,  1.61it/s]
Epoch 0:  12%|█▏        | 404/3295 [04:11<30:02,  1.60it/s]
Epoch 0:  12%|█▏        | 404/3295 [04:11<30:02,  1.60it/s]
Epoch 0:  12%|█▏        | 405/3295 [04:11<29:57,  1.61it/s]
Epoch 0:  12%|█▏        | 405/3295 [04:11<29:57,  1.61it/s]
Epoch 0:  12%|█▏        | 406/3295 [04:11<29:52,  1.61it/s]
Epoch 0:  12%|█▏        | 406/3295 [04:11<29:52,  1.61it/s]
Epoch 0:  12%|█▏        | 407/3295 [04:13<30:00,  1.60it/s]
Epoch 0:  12%|█▏        | 407/3295 [04:13<30:00,  1.60it/s]
Epoch 0:  12%|█▏        | 408/3295 [04:13<29:55,  1.61it/s]
Epoch 0:  12%|█▏        | 408/3295 [04:13<29:55,  1.61it/s]
Epoch 0:  12%|█▏        | 409/3295 [04:13<29:50,  1.61it/s]
Epoch 0:  12%|█▏        | 409/3295 [04:13<29:50,  1.61it/s]
Epoch 0:  12%|█▏        | 410/3295 [04:15<29:59,  1.60it/s]
Epoch 0:  12%|█▏        | 410/3295 [04:15<29:59,  1.60it/s]
Epoch 0:  12%|█▏        | 411/3295 [04:15<29:54,  1.61it/s]
Epoch 0:  12%|█▏        | 411/3295 [04:15<29:54,  1.61it/s]
Epoch 0:  13%|█▎        | 412/3295 [04:15<29:49,  1.61it/s]
Epoch 0:  13%|█▎        | 412/3295 [04:15<29:49,  1.61it/s]
Epoch 0:  13%|█▎        | 413/3295 [04:17<29:56,  1.60it/s]
Epoch 0:  13%|█▎        | 413/3295 [04:17<29:56,  1.60it/s]
Epoch 0:  13%|█▎        | 414/3295 [04:17<29:51,  1.61it/s]
Epoch 0:  13%|█▎        | 414/3295 [04:17<29:51,  1.61it/s]
Epoch 0:  13%|█▎        | 415/3295 [04:17<29:47,  1.61it/s]
Epoch 0:  13%|█▎        | 415/3295 [04:17<29:47,  1.61it/s]
Epoch 0:  13%|█▎        | 416/3295 [04:19<29:54,  1.60it/s]
Epoch 0:  13%|█▎        | 416/3295 [04:19<29:54,  1.60it/s]
Epoch 0:  13%|█▎        | 417/3295 [04:19<29:50,  1.61it/s]
Epoch 0:  13%|█▎        | 417/3295 [04:19<29:50,  1.61it/s]
Epoch 0:  13%|█▎        | 418/3295 [04:19<29:45,  1.61it/s]
Epoch 0:  13%|█▎        | 418/3295 [04:19<29:45,  1.61it/s]
Epoch 0:  13%|█▎        | 419/3295 [04:21<29:51,  1.61it/s]
Epoch 0:  13%|█▎        | 419/3295 [04:21<29:51,  1.61it/s]
Epoch 0:  13%|█▎        | 420/3295 [04:21<29:46,  1.61it/s]
Epoch 0:  13%|█▎        | 420/3295 [04:21<29:46,  1.61it/s]
Epoch 0:  13%|█▎        | 421/3295 [04:21<29:42,  1.61it/s]
Epoch 0:  13%|█▎        | 421/3295 [04:21<29:42,  1.61it/s]
Epoch 0:  13%|█▎        | 422/3295 [04:22<29:49,  1.61it/s]
Epoch 0:  13%|█▎        | 422/3295 [04:22<29:49,  1.61it/s]
Epoch 0:  13%|█▎        | 423/3295 [04:22<29:45,  1.61it/s]
Epoch 0:  13%|█▎        | 423/3295 [04:22<29:45,  1.61it/s]
Epoch 0:  13%|█▎        | 424/3295 [04:22<29:40,  1.61it/s]
Epoch 0:  13%|█▎        | 424/3295 [04:22<29:40,  1.61it/s]
Epoch 0:  13%|█▎        | 425/3295 [04:24<29:48,  1.60it/s]
Epoch 0:  13%|█▎        | 425/3295 [04:24<29:48,  1.60it/s]
Epoch 0:  13%|█▎        | 426/3295 [04:24<29:43,  1.61it/s]
Epoch 0:  13%|█▎        | 426/3295 [04:24<29:43,  1.61it/s]
Epoch 0:  13%|█▎        | 427/3295 [04:24<29:39,  1.61it/s]
Epoch 0:  13%|█▎        | 427/3295 [04:24<29:39,  1.61it/s]
Epoch 0:  13%|█▎        | 428/3295 [04:26<29:47,  1.60it/s]
Epoch 0:  13%|█▎        | 428/3295 [04:26<29:47,  1.60it/s]
Epoch 0:  13%|█▎        | 429/3295 [04:26<29:42,  1.61it/s]
Epoch 0:  13%|█▎        | 429/3295 [04:26<29:42,  1.61it/s]
Epoch 0:  13%|█▎        | 430/3295 [04:26<29:38,  1.61it/s]
Epoch 0:  13%|█▎        | 430/3295 [04:26<29:38,  1.61it/s]
Epoch 0:  13%|█▎        | 431/3295 [04:28<29:46,  1.60it/s]
Epoch 0:  13%|█▎        | 431/3295 [04:28<29:46,  1.60it/s]
Epoch 0:  13%|█▎        | 432/3295 [04:28<29:41,  1.61it/s]
Epoch 0:  13%|█▎        | 432/3295 [04:28<29:41,  1.61it/s]
Epoch 0:  13%|█▎        | 433/3295 [04:28<29:37,  1.61it/s]
Epoch 0:  13%|█▎        | 433/3295 [04:28<29:37,  1.61it/s]
Epoch 0:  13%|█▎        | 434/3295 [04:30<29:43,  1.60it/s]
Epoch 0:  13%|█▎        | 434/3295 [04:30<29:43,  1.60it/s]
Epoch 0:  13%|█▎        | 435/3295 [04:30<29:39,  1.61it/s]
Epoch 0:  13%|█▎        | 435/3295 [04:30<29:39,  1.61it/s]
Epoch 0:  13%|█▎        | 436/3295 [04:30<29:34,  1.61it/s]
Epoch 0:  13%|█▎        | 436/3295 [04:30<29:34,  1.61it/s]
Epoch 0:  13%|█▎        | 437/3295 [04:32<29:41,  1.60it/s]
Epoch 0:  13%|█▎        | 437/3295 [04:32<29:41,  1.60it/s]
Epoch 0:  13%|█▎        | 438/3295 [04:32<29:37,  1.61it/s]
Epoch 0:  13%|█▎        | 438/3295 [04:32<29:37,  1.61it/s]
Epoch 0:  13%|█▎        | 439/3295 [04:32<29:32,  1.61it/s]
Epoch 0:  13%|█▎        | 439/3295 [04:32<29:32,  1.61it/s]
Epoch 0:  13%|█▎        | 440/3295 [04:34<29:38,  1.61it/s]
Epoch 0:  13%|█▎        | 440/3295 [04:34<29:38,  1.61it/s]
Epoch 0:  13%|█▎        | 441/3295 [04:34<29:34,  1.61it/s]
Epoch 0:  13%|█▎        | 441/3295 [04:34<29:34,  1.61it/s]
Epoch 0:  13%|█▎        | 442/3295 [04:34<29:29,  1.61it/s]
Epoch 0:  13%|█▎        | 442/3295 [04:34<29:29,  1.61it/s]
Epoch 0:  13%|█▎        | 443/3295 [04:36<29:37,  1.60it/s]
Epoch 0:  13%|█▎        | 443/3295 [04:36<29:37,  1.60it/s]
Epoch 0:  13%|█▎        | 444/3295 [04:36<29:32,  1.61it/s]
Epoch 0:  13%|█▎        | 444/3295 [04:36<29:32,  1.61it/s]
Epoch 0:  14%|█▎        | 445/3295 [04:36<29:28,  1.61it/s]
Epoch 0:  14%|█▎        | 445/3295 [04:36<29:28,  1.61it/s]
Epoch 0:  14%|█▎        | 446/3295 [04:37<29:35,  1.60it/s]
Epoch 0:  14%|█▎        | 446/3295 [04:37<29:35,  1.60it/s]
Epoch 0:  14%|█▎        | 447/3295 [04:37<29:30,  1.61it/s]
Epoch 0:  14%|█▎        | 447/3295 [04:37<29:30,  1.61it/s]
Epoch 0:  14%|█▎        | 448/3295 [04:37<29:26,  1.61it/s]
Epoch 0:  14%|█▎        | 448/3295 [04:37<29:26,  1.61it/s]
Epoch 0:  14%|█▎        | 449/3295 [04:39<29:32,  1.61it/s]
Epoch 0:  14%|█▎        | 449/3295 [04:39<29:32,  1.61it/s]
Epoch 0:  14%|█▎        | 450/3295 [04:39<29:28,  1.61it/s]
Epoch 0:  14%|█▎        | 450/3295 [04:39<29:28,  1.61it/s]
Epoch 0:  14%|█▎        | 451/3295 [04:39<29:24,  1.61it/s]
Epoch 0:  14%|█▎        | 451/3295 [04:39<29:24,  1.61it/s]
Epoch 0:  14%|█▎        | 452/3295 [04:41<29:30,  1.61it/s]
Epoch 0:  14%|█▎        | 452/3295 [04:41<29:30,  1.61it/s]
Epoch 0:  14%|█▎        | 453/3295 [04:41<29:26,  1.61it/s]
Epoch 0:  14%|█▎        | 453/3295 [04:41<29:26,  1.61it/s]
Epoch 0:  14%|█▍        | 454/3295 [04:41<29:21,  1.61it/s]
Epoch 0:  14%|█▍        | 454/3295 [04:41<29:21,  1.61it/s]
Epoch 0:  14%|█▍        | 455/3295 [04:43<29:28,  1.61it/s]
Epoch 0:  14%|█▍        | 455/3295 [04:43<29:28,  1.61it/s]
Epoch 0:  14%|█▍        | 456/3295 [04:43<29:24,  1.61it/s]
Epoch 0:  14%|█▍        | 456/3295 [04:43<29:24,  1.61it/s]
Epoch 0:  14%|█▍        | 457/3295 [04:43<29:19,  1.61it/s]
Epoch 0:  14%|█▍        | 457/3295 [04:43<29:19,  1.61it/s]
Epoch 0:  14%|█▍        | 458/3295 [04:45<29:26,  1.61it/s]
Epoch 0:  14%|█▍        | 458/3295 [04:45<29:26,  1.61it/s]
Epoch 0:  14%|█▍        | 459/3295 [04:45<29:22,  1.61it/s]
Epoch 0:  14%|█▍        | 459/3295 [04:45<29:22,  1.61it/s]
Epoch 0:  14%|█▍        | 460/3295 [04:45<29:17,  1.61it/s]
Epoch 0:  14%|█▍        | 460/3295 [04:45<29:17,  1.61it/s]
Epoch 0:  14%|█▍        | 461/3295 [04:46<29:23,  1.61it/s]
Epoch 0:  14%|█▍        | 461/3295 [04:46<29:23,  1.61it/s]
Epoch 0:  14%|█▍        | 462/3295 [04:47<29:20,  1.61it/s]
Epoch 0:  14%|█▍        | 462/3295 [04:47<29:20,  1.61it/s]
Epoch 0:  14%|█▍        | 463/3295 [04:47<29:15,  1.61it/s]
Epoch 0:  14%|█▍        | 463/3295 [04:47<29:15,  1.61it/s]
Epoch 0:  14%|█▍        | 464/3295 [04:48<29:21,  1.61it/s]
Epoch 0:  14%|█▍        | 464/3295 [04:48<29:21,  1.61it/s]
Epoch 0:  14%|█▍        | 465/3295 [04:48<29:18,  1.61it/s]
Epoch 0:  14%|█▍        | 465/3295 [04:48<29:18,  1.61it/s]
Epoch 0:  14%|█▍        | 466/3295 [04:48<29:13,  1.61it/s]
Epoch 0:  14%|█▍        | 466/3295 [04:48<29:13,  1.61it/s]
Epoch 0:  14%|█▍        | 467/3295 [04:50<29:19,  1.61it/s]
Epoch 0:  14%|█▍        | 467/3295 [04:50<29:19,  1.61it/s]
Epoch 0:  14%|█▍        | 468/3295 [04:50<29:16,  1.61it/s]
Epoch 0:  14%|█▍        | 468/3295 [04:50<29:16,  1.61it/s]
Epoch 0:  14%|█▍        | 469/3295 [04:50<29:11,  1.61it/s]
Epoch 0:  14%|█▍        | 469/3295 [04:50<29:11,  1.61it/s]
Epoch 0:  14%|█▍        | 470/3295 [04:52<29:17,  1.61it/s]
Epoch 0:  14%|█▍        | 470/3295 [04:52<29:17,  1.61it/s]
Epoch 0:  14%|█▍        | 471/3295 [04:52<29:13,  1.61it/s]
Epoch 0:  14%|█▍        | 471/3295 [04:52<29:13,  1.61it/s]
Epoch 0:  14%|█▍        | 472/3295 [04:52<29:09,  1.61it/s]
Epoch 0:  14%|█▍        | 472/3295 [04:52<29:09,  1.61it/s]
Epoch 0:  14%|█▍        | 473/3295 [04:54<29:14,  1.61it/s]
Epoch 0:  14%|█▍        | 473/3295 [04:54<29:14,  1.61it/s]
Epoch 0:  14%|█▍        | 474/3295 [04:54<29:11,  1.61it/s]
Epoch 0:  14%|█▍        | 474/3295 [04:54<29:11,  1.61it/s]
Epoch 0:  14%|█▍        | 475/3295 [04:54<29:07,  1.61it/s]
Epoch 0:  14%|█▍        | 475/3295 [04:54<29:07,  1.61it/s]
Epoch 0:  14%|█▍        | 476/3295 [04:56<29:13,  1.61it/s]
Epoch 0:  14%|█▍        | 476/3295 [04:56<29:13,  1.61it/s]
Epoch 0:  14%|█▍        | 477/3295 [04:56<29:10,  1.61it/s]
Epoch 0:  14%|█▍        | 477/3295 [04:56<29:10,  1.61it/s]
Epoch 0:  15%|█▍        | 478/3295 [04:56<29:06,  1.61it/s]
Epoch 0:  15%|█▍        | 478/3295 [04:56<29:06,  1.61it/s]
Epoch 0:  15%|█▍        | 479/3295 [04:58<29:12,  1.61it/s]
Epoch 0:  15%|█▍        | 479/3295 [04:58<29:12,  1.61it/s]
Epoch 0:  15%|█▍        | 480/3295 [04:58<29:08,  1.61it/s]
Epoch 0:  15%|█▍        | 480/3295 [04:58<29:08,  1.61it/s]
Epoch 0:  15%|█▍        | 481/3295 [04:58<29:04,  1.61it/s]
Epoch 0:  15%|█▍        | 481/3295 [04:58<29:04,  1.61it/s]
Epoch 0:  15%|█▍        | 482/3295 [05:00<29:11,  1.61it/s]
Epoch 0:  15%|█▍        | 482/3295 [05:00<29:11,  1.61it/s]
Epoch 0:  15%|█▍        | 483/3295 [05:00<29:07,  1.61it/s]
Epoch 0:  15%|█▍        | 483/3295 [05:00<29:07,  1.61it/s]
Epoch 0:  15%|█▍        | 484/3295 [05:00<29:03,  1.61it/s]
Epoch 0:  15%|█▍        | 484/3295 [05:00<29:03,  1.61it/s]
Epoch 0:  15%|█▍        | 485/3295 [05:01<29:09,  1.61it/s]
Epoch 0:  15%|█▍        | 485/3295 [05:01<29:09,  1.61it/s]
Epoch 0:  15%|█▍        | 486/3295 [05:02<29:05,  1.61it/s]
Epoch 0:  15%|█▍        | 486/3295 [05:02<29:05,  1.61it/s]
Epoch 0:  15%|█▍        | 487/3295 [05:02<29:01,  1.61it/s]
Epoch 0:  15%|█▍        | 487/3295 [05:02<29:01,  1.61it/s]
Epoch 0:  15%|█▍        | 488/3295 [05:03<29:07,  1.61it/s]
Epoch 0:  15%|█▍        | 488/3295 [05:03<29:07,  1.61it/s]
Epoch 0:  15%|█▍        | 489/3295 [05:03<29:03,  1.61it/s]
Epoch 0:  15%|█▍        | 489/3295 [05:03<29:03,  1.61it/s]
Epoch 0:  15%|█▍        | 490/3295 [05:03<28:59,  1.61it/s]
Epoch 0:  15%|█▍        | 490/3295 [05:03<28:59,  1.61it/s]
Epoch 0:  15%|█▍        | 491/3295 [05:05<29:05,  1.61it/s]
Epoch 0:  15%|█▍        | 491/3295 [05:05<29:05,  1.61it/s]
Epoch 0:  15%|█▍        | 492/3295 [05:05<29:01,  1.61it/s]
Epoch 0:  15%|█▍        | 492/3295 [05:05<29:01,  1.61it/s]
Epoch 0:  15%|█▍        | 493/3295 [05:05<28:57,  1.61it/s]
Epoch 0:  15%|█▍        | 493/3295 [05:05<28:57,  1.61it/s]
Epoch 0:  15%|█▍        | 494/3295 [05:07<29:03,  1.61it/s]
Epoch 0:  15%|█▍        | 494/3295 [05:07<29:03,  1.61it/s]
Epoch 0:  15%|█▌        | 495/3295 [05:07<28:59,  1.61it/s]
Epoch 0:  15%|█▌        | 495/3295 [05:07<28:59,  1.61it/s]
Epoch 0:  15%|█▌        | 496/3295 [05:07<28:55,  1.61it/s]
Epoch 0:  15%|█▌        | 496/3295 [05:07<28:55,  1.61it/s]
Epoch 0:  15%|█▌        | 497/3295 [05:09<29:02,  1.61it/s]
Epoch 0:  15%|█▌        | 497/3295 [05:09<29:02,  1.61it/s]
Epoch 0:  15%|█▌        | 498/3295 [05:09<28:58,  1.61it/s]
Epoch 0:  15%|█▌        | 498/3295 [05:09<28:58,  1.61it/s]
Epoch 0:  15%|█▌        | 499/3295 [05:09<28:54,  1.61it/s]
Epoch 0:  15%|█▌        | 499/3295 [05:09<28:54,  1.61it/s]
Epoch 0:  15%|█▌        | 500/3295 [05:11<29:00,  1.61it/s]
Epoch 0:  15%|█▌        | 500/3295 [05:11<29:00,  1.61it/s]
Epoch 0:  15%|█▌        | 501/3295 [05:11<28:56,  1.61it/s]
Epoch 0:  15%|█▌        | 501/3295 [05:11<28:56,  1.61it/s]
Epoch 0:  15%|█▌        | 502/3295 [05:11<28:52,  1.61it/s]
Epoch 0:  15%|█▌        | 502/3295 [05:11<28:52,  1.61it/s]
Epoch 0:  15%|█▌        | 503/3295 [05:13<28:57,  1.61it/s]
Epoch 0:  15%|█▌        | 503/3295 [05:13<28:57,  1.61it/s]
Epoch 0:  15%|█▌        | 504/3295 [05:13<28:54,  1.61it/s]
Epoch 0:  15%|█▌        | 504/3295 [05:13<28:54,  1.61it/s]
Epoch 0:  15%|█▌        | 505/3295 [05:13<28:50,  1.61it/s]
Epoch 0:  15%|█▌        | 505/3295 [05:13<28:50,  1.61it/s]
Epoch 0:  15%|█▌        | 506/3295 [05:14<28:55,  1.61it/s]
Epoch 0:  15%|█▌        | 506/3295 [05:14<28:55,  1.61it/s]
Epoch 0:  15%|█▌        | 507/3295 [05:15<28:52,  1.61it/s]
Epoch 0:  15%|█▌        | 507/3295 [05:15<28:52,  1.61it/s]
Epoch 0:  15%|█▌        | 508/3295 [05:15<28:48,  1.61it/s]
Epoch 0:  15%|█▌        | 508/3295 [05:15<28:48,  1.61it/s]
Epoch 0:  15%|█▌        | 509/3295 [05:16<28:53,  1.61it/s]
Epoch 0:  15%|█▌        | 509/3295 [05:16<28:53,  1.61it/s]
Epoch 0:  15%|█▌        | 510/3295 [05:16<28:50,  1.61it/s]
Epoch 0:  15%|█▌        | 510/3295 [05:16<28:50,  1.61it/s]
Epoch 0:  16%|█▌        | 511/3295 [05:16<28:46,  1.61it/s]
Epoch 0:  16%|█▌        | 511/3295 [05:16<28:46,  1.61it/s]
Epoch 0:  16%|█▌        | 512/3295 [05:18<28:51,  1.61it/s]
Epoch 0:  16%|█▌        | 512/3295 [05:18<28:51,  1.61it/s]
Epoch 0:  16%|█▌        | 513/3295 [05:18<28:49,  1.61it/s]
Epoch 0:  16%|█▌        | 513/3295 [05:18<28:49,  1.61it/s]
Epoch 0:  16%|█▌        | 514/3295 [05:18<28:45,  1.61it/s]
Epoch 0:  16%|█▌        | 514/3295 [05:18<28:45,  1.61it/s]
Epoch 0:  16%|█▌        | 515/3295 [05:20<28:49,  1.61it/s]
Epoch 0:  16%|█▌        | 515/3295 [05:20<28:49,  1.61it/s]
Epoch 0:  16%|█▌        | 516/3295 [05:20<28:48,  1.61it/s]
Epoch 0:  16%|█▌        | 516/3295 [05:20<28:48,  1.61it/s]
Epoch 0:  16%|█▌        | 517/3295 [05:20<28:44,  1.61it/s]
Epoch 0:  16%|█▌        | 517/3295 [05:20<28:44,  1.61it/s]
Epoch 0:  16%|█▌        | 518/3295 [05:22<28:47,  1.61it/s]
Epoch 0:  16%|█▌        | 518/3295 [05:22<28:47,  1.61it/s]
Epoch 0:  16%|█▌        | 519/3295 [05:22<28:46,  1.61it/s]
Epoch 0:  16%|█▌        | 519/3295 [05:22<28:46,  1.61it/s]
Epoch 0:  16%|█▌        | 520/3295 [05:22<28:42,  1.61it/s]
Epoch 0:  16%|█▌        | 520/3295 [05:22<28:42,  1.61it/s]
Epoch 0:  16%|█▌        | 521/3295 [05:24<28:46,  1.61it/s]
Epoch 0:  16%|█▌        | 521/3295 [05:24<28:46,  1.61it/s]
Epoch 0:  16%|█▌        | 522/3295 [05:24<28:45,  1.61it/s]
Epoch 0:  16%|█▌        | 522/3295 [05:24<28:45,  1.61it/s]
Epoch 0:  16%|█▌        | 523/3295 [05:24<28:41,  1.61it/s]
Epoch 0:  16%|█▌        | 523/3295 [05:24<28:41,  1.61it/s]
Epoch 0:  16%|█▌        | 524/3295 [05:26<28:44,  1.61it/s]
Epoch 0:  16%|█▌        | 524/3295 [05:26<28:44,  1.61it/s]
Epoch 0:  16%|█▌        | 525/3295 [05:26<28:42,  1.61it/s]
Epoch 0:  16%|█▌        | 525/3295 [05:26<28:42,  1.61it/s]
Epoch 0:  16%|█▌        | 526/3295 [05:26<28:38,  1.61it/s]
Epoch 0:  16%|█▌        | 526/3295 [05:26<28:38,  1.61it/s]
Epoch 0:  16%|█▌        | 527/3295 [05:27<28:42,  1.61it/s]
Epoch 0:  16%|█▌        | 527/3295 [05:27<28:42,  1.61it/s]
Epoch 0:  16%|█▌        | 528/3295 [05:28<28:40,  1.61it/s]
Epoch 0:  16%|█▌        | 528/3295 [05:28<28:40,  1.61it/s]
Epoch 0:  16%|█▌        | 529/3295 [05:28<28:37,  1.61it/s]
Epoch 0:  16%|█▌        | 529/3295 [05:28<28:37,  1.61it/s]
Epoch 0:  16%|█▌        | 530/3295 [05:29<28:40,  1.61it/s]
Epoch 0:  16%|█▌        | 530/3295 [05:29<28:40,  1.61it/s]
Epoch 0:  16%|█▌        | 531/3295 [05:30<28:39,  1.61it/s]
Epoch 0:  16%|█▌        | 531/3295 [05:30<28:39,  1.61it/s]
Epoch 0:  16%|█▌        | 532/3295 [05:30<28:35,  1.61it/s]
Epoch 0:  16%|█▌        | 532/3295 [05:30<28:35,  1.61it/s]
Epoch 0:  16%|█▌        | 533/3295 [05:31<28:38,  1.61it/s]
Epoch 0:  16%|█▌        | 533/3295 [05:31<28:38,  1.61it/s]
Epoch 0:  16%|█▌        | 534/3295 [05:32<28:37,  1.61it/s]
Epoch 0:  16%|█▌        | 534/3295 [05:32<28:37,  1.61it/s]
Epoch 0:  16%|█▌        | 535/3295 [05:32<28:33,  1.61it/s]
Epoch 0:  16%|█▌        | 535/3295 [05:32<28:33,  1.61it/s]
Epoch 0:  16%|█▋        | 536/3295 [05:33<28:36,  1.61it/s]
Epoch 0:  16%|█▋        | 536/3295 [05:33<28:36,  1.61it/s]
Epoch 0:  16%|█▋        | 537/3295 [05:34<28:36,  1.61it/s]
Epoch 0:  16%|█▋        | 537/3295 [05:34<28:36,  1.61it/s]
Epoch 0:  16%|█▋        | 538/3295 [05:34<28:32,  1.61it/s]
Epoch 0:  16%|█▋        | 538/3295 [05:34<28:32,  1.61it/s]
Epoch 0:  16%|█▋        | 539/3295 [05:35<28:34,  1.61it/s]
Epoch 0:  16%|█▋        | 539/3295 [05:35<28:34,  1.61it/s]
Epoch 0:  16%|█▋        | 540/3295 [05:35<28:33,  1.61it/s]
Epoch 0:  16%|█▋        | 540/3295 [05:35<28:33,  1.61it/s]
Epoch 0:  16%|█▋        | 541/3295 [05:35<28:30,  1.61it/s]
Epoch 0:  16%|█▋        | 541/3295 [05:35<28:30,  1.61it/s]
Epoch 0:  16%|█▋        | 542/3295 [05:37<28:32,  1.61it/s]
Epoch 0:  16%|█▋        | 542/3295 [05:37<28:32,  1.61it/s]
Epoch 0:  16%|█▋        | 543/3295 [05:37<28:32,  1.61it/s]
Epoch 0:  16%|█▋        | 543/3295 [05:37<28:32,  1.61it/s]
Epoch 0:  17%|█▋        | 544/3295 [05:37<28:28,  1.61it/s]
Epoch 0:  17%|█▋        | 544/3295 [05:37<28:28,  1.61it/s]
Epoch 0:  17%|█▋        | 545/3295 [05:39<28:30,  1.61it/s]
Epoch 0:  17%|█▋        | 545/3295 [05:39<28:30,  1.61it/s]
Epoch 0:  17%|█▋        | 546/3295 [05:39<28:30,  1.61it/s]
Epoch 0:  17%|█▋        | 546/3295 [05:39<28:30,  1.61it/s]
Epoch 0:  17%|█▋        | 547/3295 [05:39<28:26,  1.61it/s]
Epoch 0:  17%|█▋        | 547/3295 [05:39<28:26,  1.61it/s]
Epoch 0:  17%|█▋        | 548/3295 [05:40<28:28,  1.61it/s]
Epoch 0:  17%|█▋        | 548/3295 [05:40<28:28,  1.61it/s]
Epoch 0:  17%|█▋        | 549/3295 [05:41<28:29,  1.61it/s]
Epoch 0:  17%|█▋        | 549/3295 [05:41<28:29,  1.61it/s]
Epoch 0:  17%|█▋        | 550/3295 [05:41<28:25,  1.61it/s]
Epoch 0:  17%|█▋        | 550/3295 [05:41<28:25,  1.61it/s]
Epoch 0:  17%|█▋        | 551/3295 [05:42<28:27,  1.61it/s]
Epoch 0:  17%|█▋        | 551/3295 [05:42<28:27,  1.61it/s]
Epoch 0:  17%|█▋        | 552/3295 [05:43<28:27,  1.61it/s]
Epoch 0:  17%|█▋        | 552/3295 [05:43<28:27,  1.61it/s]
Epoch 0:  17%|█▋        | 553/3295 [05:43<28:23,  1.61it/s]
Epoch 0:  17%|█▋        | 553/3295 [05:43<28:23,  1.61it/s]
Epoch 0:  17%|█▋        | 554/3295 [05:44<28:26,  1.61it/s]
Epoch 0:  17%|█▋        | 554/3295 [05:44<28:26,  1.61it/s]
Epoch 0:  17%|█▋        | 555/3295 [05:45<28:24,  1.61it/s]
Epoch 0:  17%|█▋        | 555/3295 [05:45<28:24,  1.61it/s]
Epoch 0:  17%|█▋        | 556/3295 [05:45<28:20,  1.61it/s]
Epoch 0:  17%|█▋        | 556/3295 [05:45<28:21,  1.61it/s]
Epoch 0:  17%|█▋        | 557/3295 [05:46<28:24,  1.61it/s]
Epoch 0:  17%|█▋        | 557/3295 [05:46<28:24,  1.61it/s]
Epoch 0:  17%|█▋        | 558/3295 [05:47<28:22,  1.61it/s]
Epoch 0:  17%|█▋        | 558/3295 [05:47<28:22,  1.61it/s]
Epoch 0:  17%|█▋        | 559/3295 [05:47<28:18,  1.61it/s]
Epoch 0:  17%|█▋        | 559/3295 [05:47<28:18,  1.61it/s]
Epoch 0:  17%|█▋        | 560/3295 [05:48<28:22,  1.61it/s]
Epoch 0:  17%|█▋        | 560/3295 [05:48<28:22,  1.61it/s]
Epoch 0:  17%|█▋        | 561/3295 [05:48<28:20,  1.61it/s]
Epoch 0:  17%|█▋        | 561/3295 [05:48<28:20,  1.61it/s]
Epoch 0:  17%|█▋        | 562/3295 [05:48<28:16,  1.61it/s]
Epoch 0:  17%|█▋        | 562/3295 [05:48<28:16,  1.61it/s]
Epoch 0:  17%|█▋        | 563/3295 [05:50<28:20,  1.61it/s]
Epoch 0:  17%|█▋        | 563/3295 [05:50<28:20,  1.61it/s]
Epoch 0:  17%|█▋        | 564/3295 [05:50<28:18,  1.61it/s]
Epoch 0:  17%|█▋        | 564/3295 [05:50<28:18,  1.61it/s]
Epoch 0:  17%|█▋        | 565/3295 [05:50<28:15,  1.61it/s]
Epoch 0:  17%|█▋        | 565/3295 [05:50<28:15,  1.61it/s]
Epoch 0:  17%|█▋        | 566/3295 [05:52<28:18,  1.61it/s]
Epoch 0:  17%|█▋        | 566/3295 [05:52<28:18,  1.61it/s]
Epoch 0:  17%|█▋        | 567/3295 [05:52<28:16,  1.61it/s]
Epoch 0:  17%|█▋        | 567/3295 [05:52<28:16,  1.61it/s]
Epoch 0:  17%|█▋        | 568/3295 [05:52<28:12,  1.61it/s]
Epoch 0:  17%|█▋        | 568/3295 [05:52<28:12,  1.61it/s]
Epoch 0:  17%|█▋        | 569/3295 [05:54<28:16,  1.61it/s]
Epoch 0:  17%|█▋        | 569/3295 [05:54<28:16,  1.61it/s]
Epoch 0:  17%|█▋        | 570/3295 [05:54<28:14,  1.61it/s]
Epoch 0:  17%|█▋        | 570/3295 [05:54<28:14,  1.61it/s]
Epoch 0:  17%|█▋        | 571/3295 [05:54<28:10,  1.61it/s]
Epoch 0:  17%|█▋        | 571/3295 [05:54<28:10,  1.61it/s]
Epoch 0:  17%|█▋        | 572/3295 [05:55<28:13,  1.61it/s]
Epoch 0:  17%|█▋        | 572/3295 [05:55<28:13,  1.61it/s]
Epoch 0:  17%|█▋        | 573/3295 [05:56<28:12,  1.61it/s]
Epoch 0:  17%|█▋        | 573/3295 [05:56<28:12,  1.61it/s]
Epoch 0:  17%|█▋        | 574/3295 [05:56<28:09,  1.61it/s]
Epoch 0:  17%|█▋        | 574/3295 [05:56<28:09,  1.61it/s]
Epoch 0:  17%|█▋        | 575/3295 [05:57<28:11,  1.61it/s]
Epoch 0:  17%|█▋        | 575/3295 [05:57<28:11,  1.61it/s]
Epoch 0:  17%|█▋        | 576/3295 [05:58<28:10,  1.61it/s]
Epoch 0:  17%|█▋        | 576/3295 [05:58<28:10,  1.61it/s]
Epoch 0:  18%|█▊        | 577/3295 [05:58<28:07,  1.61it/s]
Epoch 0:  18%|█▊        | 577/3295 [05:58<28:07,  1.61it/s]
Epoch 0:  18%|█▊        | 578/3295 [05:59<28:10,  1.61it/s]
Epoch 0:  18%|█▊        | 578/3295 [05:59<28:10,  1.61it/s]
Epoch 0:  18%|█▊        | 579/3295 [05:59<28:08,  1.61it/s]
Epoch 0:  18%|█▊        | 579/3295 [05:59<28:08,  1.61it/s]
Epoch 0:  18%|█▊        | 580/3295 [05:59<28:04,  1.61it/s]
Epoch 0:  18%|█▊        | 580/3295 [05:59<28:04,  1.61it/s]
Epoch 0:  18%|█▊        | 581/3295 [06:01<28:08,  1.61it/s]
Epoch 0:  18%|█▊        | 581/3295 [06:01<28:08,  1.61it/s]
Epoch 0:  18%|█▊        | 582/3295 [06:01<28:06,  1.61it/s]
Epoch 0:  18%|█▊        | 582/3295 [06:01<28:06,  1.61it/s]
Epoch 0:  18%|█▊        | 583/3295 [06:01<28:02,  1.61it/s]
Epoch 0:  18%|█▊        | 583/3295 [06:01<28:02,  1.61it/s]
Epoch 0:  18%|█▊        | 584/3295 [06:03<28:06,  1.61it/s]
Epoch 0:  18%|█▊        | 584/3295 [06:03<28:06,  1.61it/s]
Epoch 0:  18%|█▊        | 585/3295 [06:03<28:04,  1.61it/s]
Epoch 0:  18%|█▊        | 585/3295 [06:03<28:04,  1.61it/s]
Epoch 0:  18%|█▊        | 586/3295 [06:03<28:00,  1.61it/s]
Epoch 0:  18%|█▊        | 586/3295 [06:03<28:00,  1.61it/s]
Epoch 0:  18%|█▊        | 587/3295 [06:05<28:05,  1.61it/s]
Epoch 0:  18%|█▊        | 587/3295 [06:05<28:05,  1.61it/s]
Epoch 0:  18%|█▊        | 588/3295 [06:05<28:02,  1.61it/s]
Epoch 0:  18%|█▊        | 588/3295 [06:05<28:02,  1.61it/s]
Epoch 0:  18%|█▊        | 589/3295 [06:05<27:58,  1.61it/s]
Epoch 0:  18%|█▊        | 589/3295 [06:05<27:58,  1.61it/s]
Epoch 0:  18%|█▊        | 590/3295 [06:07<28:03,  1.61it/s]
Epoch 0:  18%|█▊        | 590/3295 [06:07<28:03,  1.61it/s]
Epoch 0:  18%|█▊        | 591/3295 [06:07<28:00,  1.61it/s]
Epoch 0:  18%|█▊        | 591/3295 [06:07<28:00,  1.61it/s]
Epoch 0:  18%|█▊        | 592/3295 [06:07<27:56,  1.61it/s]
Epoch 0:  18%|█▊        | 592/3295 [06:07<27:56,  1.61it/s]
Epoch 0:  18%|█▊        | 593/3295 [06:09<28:01,  1.61it/s]
Epoch 0:  18%|█▊        | 593/3295 [06:09<28:01,  1.61it/s]
Epoch 0:  18%|█▊        | 594/3295 [06:09<27:58,  1.61it/s]
Epoch 0:  18%|█▊        | 594/3295 [06:09<27:58,  1.61it/s]
Epoch 0:  18%|█▊        | 595/3295 [06:09<27:54,  1.61it/s]
Epoch 0:  18%|█▊        | 595/3295 [06:09<27:54,  1.61it/s]
Epoch 0:  18%|█▊        | 596/3295 [06:11<28:00,  1.61it/s]
Epoch 0:  18%|█▊        | 596/3295 [06:11<28:00,  1.61it/s]
Epoch 0:  18%|█▊        | 597/3295 [06:11<27:56,  1.61it/s]
Epoch 0:  18%|█▊        | 597/3295 [06:11<27:56,  1.61it/s]
Epoch 0:  18%|█▊        | 598/3295 [06:11<27:53,  1.61it/s]
Epoch 0:  18%|█▊        | 598/3295 [06:11<27:53,  1.61it/s]
Epoch 0:  18%|█▊        | 599/3295 [06:12<27:58,  1.61it/s]
Epoch 0:  18%|█▊        | 599/3295 [06:12<27:58,  1.61it/s]
Epoch 0:  18%|█▊        | 600/3295 [06:12<27:54,  1.61it/s]
Epoch 0:  18%|█▊        | 600/3295 [06:12<27:54,  1.61it/s]
Epoch 0:  18%|█▊        | 601/3295 [06:12<27:51,  1.61it/s]
Epoch 0:  18%|█▊        | 601/3295 [06:12<27:51,  1.61it/s]
Epoch 0:  18%|█▊        | 602/3295 [06:14<27:56,  1.61it/s]
Epoch 0:  18%|█▊        | 602/3295 [06:14<27:56,  1.61it/s]
Epoch 0:  18%|█▊        | 603/3295 [06:14<27:53,  1.61it/s]
Epoch 0:  18%|█▊        | 603/3295 [06:14<27:53,  1.61it/s]
Epoch 0:  18%|█▊        | 604/3295 [06:14<27:50,  1.61it/s]
Epoch 0:  18%|█▊        | 604/3295 [06:14<27:50,  1.61it/s]
Epoch 0:  18%|█▊        | 605/3295 [06:16<27:55,  1.61it/s]
Epoch 0:  18%|█▊        | 605/3295 [06:16<27:55,  1.61it/s]
Epoch 0:  18%|█▊        | 606/3295 [06:16<27:51,  1.61it/s]
Epoch 0:  18%|█▊        | 606/3295 [06:16<27:51,  1.61it/s]
Epoch 0:  18%|█▊        | 607/3295 [06:16<27:48,  1.61it/s]
Epoch 0:  18%|█▊        | 607/3295 [06:16<27:48,  1.61it/s]
Epoch 0:  18%|█▊        | 608/3295 [06:18<27:53,  1.61it/s]
Epoch 0:  18%|█▊        | 608/3295 [06:18<27:53,  1.61it/s]
Epoch 0:  18%|█▊        | 609/3295 [06:18<27:50,  1.61it/s]
Epoch 0:  18%|█▊        | 609/3295 [06:18<27:50,  1.61it/s]
Epoch 0:  19%|█▊        | 610/3295 [06:18<27:47,  1.61it/s]
Epoch 0:  19%|█▊        | 610/3295 [06:18<27:47,  1.61it/s]
Epoch 0:  19%|█▊        | 611/3295 [06:20<27:51,  1.61it/s]
Epoch 0:  19%|█▊        | 611/3295 [06:20<27:51,  1.61it/s]
Epoch 0:  19%|█▊        | 612/3295 [06:20<27:48,  1.61it/s]
Epoch 0:  19%|█▊        | 612/3295 [06:20<27:48,  1.61it/s]
Epoch 0:  19%|█▊        | 613/3295 [06:20<27:45,  1.61it/s]
Epoch 0:  19%|█▊        | 613/3295 [06:20<27:45,  1.61it/s]
Epoch 0:  19%|█▊        | 614/3295 [06:22<27:49,  1.61it/s]
Epoch 0:  19%|█▊        | 614/3295 [06:22<27:49,  1.61it/s]
Epoch 0:  19%|█▊        | 615/3295 [06:22<27:46,  1.61it/s]
Epoch 0:  19%|█▊        | 615/3295 [06:22<27:46,  1.61it/s]
Epoch 0:  19%|█▊        | 616/3295 [06:22<27:42,  1.61it/s]
Epoch 0:  19%|█▊        | 616/3295 [06:22<27:42,  1.61it/s]
Epoch 0:  19%|█▊        | 617/3295 [06:24<27:47,  1.61it/s]
Epoch 0:  19%|█▊        | 617/3295 [06:24<27:47,  1.61it/s]
Epoch 0:  19%|█▉        | 618/3295 [06:24<27:44,  1.61it/s]
Epoch 0:  19%|█▉        | 618/3295 [06:24<27:44,  1.61it/s]
Epoch 0:  19%|█▉        | 619/3295 [06:24<27:41,  1.61it/s]
Epoch 0:  19%|█▉        | 619/3295 [06:24<27:41,  1.61it/s]
Epoch 0:  19%|█▉        | 620/3295 [06:26<27:45,  1.61it/s]
Epoch 0:  19%|█▉        | 620/3295 [06:26<27:45,  1.61it/s]
Epoch 0:  19%|█▉        | 621/3295 [06:26<27:42,  1.61it/s]
Epoch 0:  19%|█▉        | 621/3295 [06:26<27:42,  1.61it/s]
Epoch 0:  19%|█▉        | 622/3295 [06:26<27:39,  1.61it/s]
Epoch 0:  19%|█▉        | 622/3295 [06:26<27:39,  1.61it/s]
Epoch 0:  19%|█▉        | 623/3295 [06:28<27:44,  1.61it/s]
Epoch 0:  19%|█▉        | 623/3295 [06:28<27:44,  1.61it/s]
Epoch 0:  19%|█▉        | 624/3295 [06:28<27:41,  1.61it/s]
Epoch 0:  19%|█▉        | 624/3295 [06:28<27:41,  1.61it/s]
Epoch 0:  19%|█▉        | 625/3295 [06:28<27:38,  1.61it/s]
Epoch 0:  19%|█▉        | 625/3295 [06:28<27:38,  1.61it/s]
Epoch 0:  19%|█▉        | 626/3295 [06:29<27:42,  1.61it/s]
Epoch 0:  19%|█▉        | 626/3295 [06:29<27:42,  1.61it/s]
Epoch 0:  19%|█▉        | 627/3295 [06:30<27:39,  1.61it/s]
Epoch 0:  19%|█▉        | 627/3295 [06:30<27:39,  1.61it/s]
Epoch 0:  19%|█▉        | 628/3295 [06:30<27:36,  1.61it/s]
Epoch 0:  19%|█▉        | 628/3295 [06:30<27:36,  1.61it/s]
Epoch 0:  19%|█▉        | 629/3295 [06:31<27:40,  1.61it/s]
Epoch 0:  19%|█▉        | 629/3295 [06:31<27:40,  1.61it/s]
Epoch 0:  19%|█▉        | 630/3295 [06:31<27:37,  1.61it/s]
Epoch 0:  19%|█▉        | 630/3295 [06:31<27:37,  1.61it/s]
Epoch 0:  19%|█▉        | 631/3295 [06:31<27:34,  1.61it/s]
Epoch 0:  19%|█▉        | 631/3295 [06:31<27:34,  1.61it/s]
Epoch 0:  19%|█▉        | 632/3295 [06:33<27:38,  1.61it/s]
Epoch 0:  19%|█▉        | 632/3295 [06:33<27:38,  1.61it/s]
Epoch 0:  19%|█▉        | 633/3295 [06:33<27:35,  1.61it/s]
Epoch 0:  19%|█▉        | 633/3295 [06:33<27:35,  1.61it/s]
Epoch 0:  19%|█▉        | 634/3295 [06:33<27:32,  1.61it/s]
Epoch 0:  19%|█▉        | 634/3295 [06:33<27:32,  1.61it/s]
Epoch 0:  19%|█▉        | 635/3295 [06:35<27:36,  1.61it/s]
Epoch 0:  19%|█▉        | 635/3295 [06:35<27:36,  1.61it/s]
Epoch 0:  19%|█▉        | 636/3295 [06:35<27:32,  1.61it/s]
Epoch 0:  19%|█▉        | 636/3295 [06:35<27:32,  1.61it/s]
Epoch 0:  19%|█▉        | 637/3295 [06:35<27:29,  1.61it/s]
Epoch 0:  19%|█▉        | 637/3295 [06:35<27:29,  1.61it/s]
Epoch 0:  19%|█▉        | 638/3295 [06:37<27:33,  1.61it/s]
Epoch 0:  19%|█▉        | 638/3295 [06:37<27:33,  1.61it/s]
Epoch 0:  19%|█▉        | 639/3295 [06:37<27:30,  1.61it/s]
Epoch 0:  19%|█▉        | 639/3295 [06:37<27:30,  1.61it/s]
Epoch 0:  19%|█▉        | 640/3295 [06:37<27:27,  1.61it/s]
Epoch 0:  19%|█▉        | 640/3295 [06:37<27:27,  1.61it/s]
Epoch 0:  19%|█▉        | 641/3295 [06:38<27:31,  1.61it/s]
Epoch 0:  19%|█▉        | 641/3295 [06:38<27:31,  1.61it/s]
Epoch 0:  19%|█▉        | 642/3295 [06:38<27:28,  1.61it/s]
Epoch 0:  19%|█▉        | 642/3295 [06:38<27:28,  1.61it/s]
Epoch 0:  20%|█▉        | 643/3295 [06:38<27:25,  1.61it/s]
Epoch 0:  20%|█▉        | 643/3295 [06:38<27:25,  1.61it/s]
Epoch 0:  20%|█▉        | 644/3295 [06:40<27:30,  1.61it/s]
Epoch 0:  20%|█▉        | 644/3295 [06:40<27:30,  1.61it/s]
Epoch 0:  20%|█▉        | 645/3295 [06:40<27:27,  1.61it/s]
Epoch 0:  20%|█▉        | 645/3295 [06:40<27:27,  1.61it/s]
Epoch 0:  20%|█▉        | 646/3295 [06:40<27:23,  1.61it/s]
Epoch 0:  20%|█▉        | 646/3295 [06:40<27:23,  1.61it/s]
Epoch 0:  20%|█▉        | 647/3295 [06:42<27:27,  1.61it/s]
Epoch 0:  20%|█▉        | 647/3295 [06:42<27:27,  1.61it/s]
Epoch 0:  20%|█▉        | 648/3295 [06:42<27:24,  1.61it/s]
Epoch 0:  20%|█▉        | 648/3295 [06:42<27:24,  1.61it/s]
Epoch 0:  20%|█▉        | 649/3295 [06:42<27:21,  1.61it/s]
Epoch 0:  20%|█▉        | 649/3295 [06:42<27:21,  1.61it/s]
Epoch 0:  20%|█▉        | 650/3295 [06:44<27:25,  1.61it/s]
Epoch 0:  20%|█▉        | 650/3295 [06:44<27:25,  1.61it/s]
Epoch 0:  20%|█▉        | 651/3295 [06:44<27:22,  1.61it/s]
Epoch 0:  20%|█▉        | 651/3295 [06:44<27:22,  1.61it/s]
Epoch 0:  20%|█▉        | 652/3295 [06:44<27:19,  1.61it/s]
Epoch 0:  20%|█▉        | 652/3295 [06:44<27:19,  1.61it/s]
Epoch 0:  20%|█▉        | 653/3295 [06:46<27:23,  1.61it/s]
Epoch 0:  20%|█▉        | 653/3295 [06:46<27:23,  1.61it/s]
Epoch 0:  20%|█▉        | 654/3295 [06:46<27:20,  1.61it/s]
Epoch 0:  20%|█▉        | 654/3295 [06:46<27:20,  1.61it/s]
Epoch 0:  20%|█▉        | 655/3295 [06:46<27:17,  1.61it/s]
Epoch 0:  20%|█▉        | 655/3295 [06:46<27:17,  1.61it/s]
Epoch 0:  20%|█▉        | 656/3295 [06:48<27:21,  1.61it/s]
Epoch 0:  20%|█▉        | 656/3295 [06:48<27:21,  1.61it/s]
Epoch 0:  20%|█▉        | 657/3295 [06:48<27:18,  1.61it/s]
Epoch 0:  20%|█▉        | 657/3295 [06:48<27:18,  1.61it/s]
Epoch 0:  20%|█▉        | 658/3295 [06:48<27:15,  1.61it/s]
Epoch 0:  20%|█▉        | 658/3295 [06:48<27:15,  1.61it/s]
Epoch 0:  20%|██        | 659/3295 [06:49<27:19,  1.61it/s]
Epoch 0:  20%|██        | 659/3295 [06:49<27:19,  1.61it/s]
Epoch 0:  20%|██        | 660/3295 [06:49<27:16,  1.61it/s]
Epoch 0:  20%|██        | 660/3295 [06:49<27:16,  1.61it/s]
Epoch 0:  20%|██        | 661/3295 [06:49<27:13,  1.61it/s]
Epoch 0:  20%|██        | 661/3295 [06:49<27:13,  1.61it/s]
Epoch 0:  20%|██        | 662/3295 [06:51<27:17,  1.61it/s]
Epoch 0:  20%|██        | 662/3295 [06:51<27:17,  1.61it/s]
Epoch 0:  20%|██        | 663/3295 [06:51<27:14,  1.61it/s]
Epoch 0:  20%|██        | 663/3295 [06:51<27:14,  1.61it/s]
Epoch 0:  20%|██        | 664/3295 [06:51<27:11,  1.61it/s]
Epoch 0:  20%|██        | 664/3295 [06:51<27:11,  1.61it/s]
Epoch 0:  20%|██        | 665/3295 [06:53<27:15,  1.61it/s]
Epoch 0:  20%|██        | 665/3295 [06:53<27:15,  1.61it/s]
Epoch 0:  20%|██        | 666/3295 [06:53<27:13,  1.61it/s]
Epoch 0:  20%|██        | 666/3295 [06:53<27:13,  1.61it/s]
Epoch 0:  20%|██        | 667/3295 [06:53<27:10,  1.61it/s]
Epoch 0:  20%|██        | 667/3295 [06:53<27:10,  1.61it/s]
Epoch 0:  20%|██        | 668/3295 [06:55<27:13,  1.61it/s]
Epoch 0:  20%|██        | 668/3295 [06:55<27:13,  1.61it/s]
Epoch 0:  20%|██        | 669/3295 [06:55<27:10,  1.61it/s]
Epoch 0:  20%|██        | 669/3295 [06:55<27:10,  1.61it/s]
Epoch 0:  20%|██        | 670/3295 [06:55<27:07,  1.61it/s]
Epoch 0:  20%|██        | 670/3295 [06:55<27:07,  1.61it/s]
Epoch 0:  20%|██        | 671/3295 [06:57<27:12,  1.61it/s]
Epoch 0:  20%|██        | 671/3295 [06:57<27:12,  1.61it/s]
Epoch 0:  20%|██        | 672/3295 [06:57<27:09,  1.61it/s]
Epoch 0:  20%|██        | 672/3295 [06:57<27:09,  1.61it/s]
Epoch 0:  20%|██        | 673/3295 [06:57<27:06,  1.61it/s]
Epoch 0:  20%|██        | 673/3295 [06:57<27:06,  1.61it/s]
Epoch 0:  20%|██        | 674/3295 [06:59<27:10,  1.61it/s]
Epoch 0:  20%|██        | 674/3295 [06:59<27:10,  1.61it/s]
Epoch 0:  20%|██        | 675/3295 [06:59<27:07,  1.61it/s]
Epoch 0:  20%|██        | 675/3295 [06:59<27:07,  1.61it/s]
Epoch 0:  21%|██        | 676/3295 [06:59<27:04,  1.61it/s]
Epoch 0:  21%|██        | 676/3295 [06:59<27:04,  1.61it/s]
Epoch 0:  21%|██        | 677/3295 [07:01<27:08,  1.61it/s]
Epoch 0:  21%|██        | 677/3295 [07:01<27:08,  1.61it/s]
Epoch 0:  21%|██        | 678/3295 [07:01<27:05,  1.61it/s]
Epoch 0:  21%|██        | 678/3295 [07:01<27:05,  1.61it/s]
Epoch 0:  21%|██        | 679/3295 [07:01<27:02,  1.61it/s]
Epoch 0:  21%|██        | 679/3295 [07:01<27:02,  1.61it/s]
Epoch 0:  21%|██        | 680/3295 [07:02<27:06,  1.61it/s]
Epoch 0:  21%|██        | 680/3295 [07:02<27:06,  1.61it/s]
Epoch 0:  21%|██        | 681/3295 [07:02<27:03,  1.61it/s]
Epoch 0:  21%|██        | 681/3295 [07:02<27:03,  1.61it/s]
Epoch 0:  21%|██        | 682/3295 [07:03<27:00,  1.61it/s]
Epoch 0:  21%|██        | 682/3295 [07:03<27:00,  1.61it/s]
Epoch 0:  21%|██        | 683/3295 [07:04<27:04,  1.61it/s]
Epoch 0:  21%|██        | 683/3295 [07:04<27:04,  1.61it/s]
Epoch 0:  21%|██        | 684/3295 [07:04<27:01,  1.61it/s]
Epoch 0:  21%|██        | 684/3295 [07:04<27:01,  1.61it/s]
Epoch 0:  21%|██        | 685/3295 [07:04<26:58,  1.61it/s]
Epoch 0:  21%|██        | 685/3295 [07:04<26:58,  1.61it/s]
Epoch 0:  21%|██        | 686/3295 [07:06<27:02,  1.61it/s]
Epoch 0:  21%|██        | 686/3295 [07:06<27:02,  1.61it/s]
Epoch 0:  21%|██        | 687/3295 [07:06<26:59,  1.61it/s]
Epoch 0:  21%|██        | 687/3295 [07:06<26:59,  1.61it/s]
Epoch 0:  21%|██        | 688/3295 [07:06<26:56,  1.61it/s]
Epoch 0:  21%|██        | 688/3295 [07:06<26:56,  1.61it/s]
Epoch 0:  21%|██        | 689/3295 [07:08<27:01,  1.61it/s]
Epoch 0:  21%|██        | 689/3295 [07:08<27:01,  1.61it/s]
Epoch 0:  21%|██        | 690/3295 [07:08<26:58,  1.61it/s]
Epoch 0:  21%|██        | 690/3295 [07:08<26:58,  1.61it/s]
Epoch 0:  21%|██        | 691/3295 [07:08<26:55,  1.61it/s]
Epoch 0:  21%|██        | 691/3295 [07:08<26:55,  1.61it/s]
Epoch 0:  21%|██        | 692/3295 [07:10<26:59,  1.61it/s]
Epoch 0:  21%|██        | 692/3295 [07:10<26:59,  1.61it/s]
Epoch 0:  21%|██        | 693/3295 [07:10<26:56,  1.61it/s]
Epoch 0:  21%|██        | 693/3295 [07:10<26:56,  1.61it/s]
Epoch 0:  21%|██        | 694/3295 [07:10<26:53,  1.61it/s]
Epoch 0:  21%|██        | 694/3295 [07:10<26:53,  1.61it/s]
Epoch 0:  21%|██        | 695/3295 [07:12<26:57,  1.61it/s]
Epoch 0:  21%|██        | 695/3295 [07:12<26:57,  1.61it/s]
Epoch 0:  21%|██        | 696/3295 [07:12<26:54,  1.61it/s]
Epoch 0:  21%|██        | 696/3295 [07:12<26:54,  1.61it/s]
Epoch 0:  21%|██        | 697/3295 [07:12<26:51,  1.61it/s]
Epoch 0:  21%|██        | 697/3295 [07:12<26:51,  1.61it/s]
Epoch 0:  21%|██        | 698/3295 [07:14<26:55,  1.61it/s]
Epoch 0:  21%|██        | 698/3295 [07:14<26:55,  1.61it/s]
Epoch 0:  21%|██        | 699/3295 [07:14<26:52,  1.61it/s]
Epoch 0:  21%|██        | 699/3295 [07:14<26:52,  1.61it/s]
Epoch 0:  21%|██        | 700/3295 [07:14<26:49,  1.61it/s]
Epoch 0:  21%|██        | 700/3295 [07:14<26:49,  1.61it/s]
Epoch 0:  21%|██▏       | 701/3295 [07:16<26:53,  1.61it/s]
Epoch 0:  21%|██▏       | 701/3295 [07:16<26:53,  1.61it/s]
Epoch 0:  21%|██▏       | 702/3295 [07:16<26:50,  1.61it/s]
Epoch 0:  21%|██▏       | 702/3295 [07:16<26:50,  1.61it/s]
Epoch 0:  21%|██▏       | 703/3295 [07:16<26:48,  1.61it/s]
Epoch 0:  21%|██▏       | 703/3295 [07:16<26:48,  1.61it/s]
Epoch 0:  21%|██▏       | 704/3295 [07:18<26:52,  1.61it/s]
Epoch 0:  21%|██▏       | 704/3295 [07:18<26:52,  1.61it/s]
Epoch 0:  21%|██▏       | 705/3295 [07:18<26:49,  1.61it/s]
Epoch 0:  21%|██▏       | 705/3295 [07:18<26:49,  1.61it/s]
Epoch 0:  21%|██▏       | 706/3295 [07:18<26:46,  1.61it/s]
Epoch 0:  21%|██▏       | 706/3295 [07:18<26:46,  1.61it/s]
Epoch 0:  21%|██▏       | 707/3295 [07:19<26:50,  1.61it/s]
Epoch 0:  21%|██▏       | 707/3295 [07:19<26:50,  1.61it/s]
Epoch 0:  21%|██▏       | 708/3295 [07:19<26:47,  1.61it/s]
Epoch 0:  21%|██▏       | 708/3295 [07:19<26:47,  1.61it/s]
Epoch 0:  22%|██▏       | 709/3295 [07:19<26:44,  1.61it/s]
Epoch 0:  22%|██▏       | 709/3295 [07:19<26:44,  1.61it/s]
Epoch 0:  22%|██▏       | 710/3295 [07:21<26:47,  1.61it/s]
Epoch 0:  22%|██▏       | 710/3295 [07:21<26:47,  1.61it/s]
Epoch 0:  22%|██▏       | 711/3295 [07:21<26:45,  1.61it/s]
Epoch 0:  22%|██▏       | 711/3295 [07:21<26:45,  1.61it/s]
Epoch 0:  22%|██▏       | 712/3295 [07:21<26:42,  1.61it/s]
Epoch 0:  22%|██▏       | 712/3295 [07:21<26:42,  1.61it/s]
Epoch 0:  22%|██▏       | 713/3295 [07:23<26:46,  1.61it/s]
Epoch 0:  22%|██▏       | 713/3295 [07:23<26:46,  1.61it/s]
Epoch 0:  22%|██▏       | 714/3295 [07:23<26:43,  1.61it/s]
Epoch 0:  22%|██▏       | 714/3295 [07:23<26:43,  1.61it/s]
Epoch 0:  22%|██▏       | 715/3295 [07:23<26:40,  1.61it/s]
Epoch 0:  22%|██▏       | 715/3295 [07:23<26:40,  1.61it/s]
Epoch 0:  22%|██▏       | 716/3295 [07:25<26:44,  1.61it/s]
Epoch 0:  22%|██▏       | 716/3295 [07:25<26:44,  1.61it/s]
Epoch 0:  22%|██▏       | 717/3295 [07:25<26:41,  1.61it/s]
Epoch 0:  22%|██▏       | 717/3295 [07:25<26:41,  1.61it/s]
Epoch 0:  22%|██▏       | 718/3295 [07:25<26:39,  1.61it/s]
Epoch 0:  22%|██▏       | 718/3295 [07:25<26:39,  1.61it/s]
Epoch 0:  22%|██▏       | 719/3295 [07:27<26:42,  1.61it/s]
Epoch 0:  22%|██▏       | 719/3295 [07:27<26:42,  1.61it/s]
Epoch 0:  22%|██▏       | 720/3295 [07:27<26:39,  1.61it/s]
Epoch 0:  22%|██▏       | 720/3295 [07:27<26:39,  1.61it/s]
Epoch 0:  22%|██▏       | 721/3295 [07:27<26:37,  1.61it/s]
Epoch 0:  22%|██▏       | 721/3295 [07:27<26:37,  1.61it/s]
Epoch 0:  22%|██▏       | 722/3295 [07:29<26:40,  1.61it/s]
Epoch 0:  22%|██▏       | 722/3295 [07:29<26:40,  1.61it/s]
Epoch 0:  22%|██▏       | 723/3295 [07:29<26:37,  1.61it/s]
Epoch 0:  22%|██▏       | 723/3295 [07:29<26:37,  1.61it/s]
Epoch 0:  22%|██▏       | 724/3295 [07:29<26:35,  1.61it/s]
Epoch 0:  22%|██▏       | 724/3295 [07:29<26:35,  1.61it/s]
Epoch 0:  22%|██▏       | 725/3295 [07:30<26:38,  1.61it/s]
Epoch 0:  22%|██▏       | 725/3295 [07:30<26:38,  1.61it/s]
Epoch 0:  22%|██▏       | 726/3295 [07:30<26:35,  1.61it/s]
Epoch 0:  22%|██▏       | 726/3295 [07:30<26:35,  1.61it/s]
Epoch 0:  22%|██▏       | 727/3295 [07:30<26:33,  1.61it/s]
Epoch 0:  22%|██▏       | 727/3295 [07:30<26:33,  1.61it/s]
Epoch 0:  22%|██▏       | 728/3295 [07:32<26:36,  1.61it/s]
Epoch 0:  22%|██▏       | 728/3295 [07:32<26:36,  1.61it/s]
Epoch 0:  22%|██▏       | 729/3295 [07:32<26:33,  1.61it/s]
Epoch 0:  22%|██▏       | 729/3295 [07:32<26:33,  1.61it/s]
Epoch 0:  22%|██▏       | 730/3295 [07:32<26:30,  1.61it/s]
Epoch 0:  22%|██▏       | 730/3295 [07:32<26:30,  1.61it/s]
Epoch 0:  22%|██▏       | 731/3295 [07:34<26:34,  1.61it/s]
Epoch 0:  22%|██▏       | 731/3295 [07:34<26:34,  1.61it/s]
Epoch 0:  22%|██▏       | 732/3295 [07:34<26:31,  1.61it/s]
Epoch 0:  22%|██▏       | 732/3295 [07:34<26:31,  1.61it/s]
Epoch 0:  22%|██▏       | 733/3295 [07:34<26:28,  1.61it/s]
Epoch 0:  22%|██▏       | 733/3295 [07:34<26:28,  1.61it/s]
Epoch 0:  22%|██▏       | 734/3295 [07:36<26:32,  1.61it/s]
Epoch 0:  22%|██▏       | 734/3295 [07:36<26:32,  1.61it/s]
Epoch 0:  22%|██▏       | 735/3295 [07:36<26:29,  1.61it/s]
Epoch 0:  22%|██▏       | 735/3295 [07:36<26:29,  1.61it/s]
Epoch 0:  22%|██▏       | 736/3295 [07:36<26:26,  1.61it/s]
Epoch 0:  22%|██▏       | 736/3295 [07:36<26:26,  1.61it/s]
Epoch 0:  22%|██▏       | 737/3295 [07:38<26:30,  1.61it/s]
Epoch 0:  22%|██▏       | 737/3295 [07:38<26:30,  1.61it/s]
Epoch 0:  22%|██▏       | 738/3295 [07:38<26:27,  1.61it/s]
Epoch 0:  22%|██▏       | 738/3295 [07:38<26:27,  1.61it/s]
Epoch 0:  22%|██▏       | 739/3295 [07:38<26:25,  1.61it/s]
Epoch 0:  22%|██▏       | 739/3295 [07:38<26:25,  1.61it/s]
Epoch 0:  22%|██▏       | 740/3295 [07:40<26:29,  1.61it/s]
Epoch 0:  22%|██▏       | 740/3295 [07:40<26:29,  1.61it/s]
Epoch 0:  22%|██▏       | 741/3295 [07:40<26:26,  1.61it/s]
Epoch 0:  22%|██▏       | 741/3295 [07:40<26:26,  1.61it/s]
Epoch 0:  23%|██▎       | 742/3295 [07:40<26:23,  1.61it/s]
Epoch 0:  23%|██▎       | 742/3295 [07:40<26:23,  1.61it/s]
Epoch 0:  23%|██▎       | 743/3295 [07:42<26:27,  1.61it/s]
Epoch 0:  23%|██▎       | 743/3295 [07:42<26:27,  1.61it/s]
Epoch 0:  23%|██▎       | 744/3295 [07:42<26:24,  1.61it/s]
Epoch 0:  23%|██▎       | 744/3295 [07:42<26:24,  1.61it/s]
Epoch 0:  23%|██▎       | 745/3295 [07:42<26:22,  1.61it/s]
Epoch 0:  23%|██▎       | 745/3295 [07:42<26:22,  1.61it/s]
Epoch 0:  23%|██▎       | 746/3295 [07:44<26:25,  1.61it/s]
Epoch 0:  23%|██▎       | 746/3295 [07:44<26:25,  1.61it/s]
Epoch 0:  23%|██▎       | 747/3295 [07:44<26:23,  1.61it/s]
Epoch 0:  23%|██▎       | 747/3295 [07:44<26:23,  1.61it/s]
Epoch 0:  23%|██▎       | 748/3295 [07:44<26:20,  1.61it/s]
Epoch 0:  23%|██▎       | 748/3295 [07:44<26:20,  1.61it/s]
Epoch 0:  23%|██▎       | 749/3295 [07:45<26:23,  1.61it/s]
Epoch 0:  23%|██▎       | 749/3295 [07:45<26:23,  1.61it/s]
Epoch 0:  23%|██▎       | 750/3295 [07:45<26:21,  1.61it/s]
Epoch 0:  23%|██▎       | 750/3295 [07:45<26:21,  1.61it/s]
Epoch 0:  23%|██▎       | 751/3295 [07:46<26:18,  1.61it/s]
Epoch 0:  23%|██▎       | 751/3295 [07:46<26:18,  1.61it/s]
Epoch 0:  23%|██▎       | 752/3295 [07:47<26:22,  1.61it/s]
Epoch 0:  23%|██▎       | 752/3295 [07:47<26:22,  1.61it/s]
Epoch 0:  23%|██▎       | 753/3295 [07:47<26:19,  1.61it/s]
Epoch 0:  23%|██▎       | 753/3295 [07:47<26:19,  1.61it/s]
Epoch 0:  23%|██▎       | 754/3295 [07:47<26:16,  1.61it/s]
Epoch 0:  23%|██▎       | 754/3295 [07:47<26:16,  1.61it/s]
Epoch 0:  23%|██▎       | 755/3295 [07:49<26:20,  1.61it/s]
Epoch 0:  23%|██▎       | 755/3295 [07:49<26:20,  1.61it/s]
Epoch 0:  23%|██▎       | 756/3295 [07:49<26:17,  1.61it/s]
Epoch 0:  23%|██▎       | 756/3295 [07:49<26:17,  1.61it/s]
Epoch 0:  23%|██▎       | 757/3295 [07:49<26:14,  1.61it/s]
Epoch 0:  23%|██▎       | 757/3295 [07:49<26:14,  1.61it/s]
Epoch 0:  23%|██▎       | 758/3295 [07:51<26:18,  1.61it/s]
Epoch 0:  23%|██▎       | 758/3295 [07:51<26:18,  1.61it/s]
Epoch 0:  23%|██▎       | 759/3295 [07:51<26:15,  1.61it/s]
Epoch 0:  23%|██▎       | 759/3295 [07:51<26:15,  1.61it/s]
Epoch 0:  23%|██▎       | 760/3295 [07:51<26:13,  1.61it/s]
Epoch 0:  23%|██▎       | 760/3295 [07:51<26:13,  1.61it/s]
Epoch 0:  23%|██▎       | 761/3295 [07:53<26:17,  1.61it/s]
Epoch 0:  23%|██▎       | 761/3295 [07:53<26:17,  1.61it/s]
Epoch 0:  23%|██▎       | 762/3295 [07:53<26:14,  1.61it/s]
Epoch 0:  23%|██▎       | 762/3295 [07:53<26:14,  1.61it/s]
Epoch 0:  23%|██▎       | 763/3295 [07:53<26:12,  1.61it/s]
Epoch 0:  23%|██▎       | 763/3295 [07:53<26:12,  1.61it/s]
Epoch 0:  23%|██▎       | 764/3295 [07:55<26:15,  1.61it/s]
Epoch 0:  23%|██▎       | 764/3295 [07:55<26:15,  1.61it/s]
Epoch 0:  23%|██▎       | 765/3295 [07:55<26:12,  1.61it/s]
Epoch 0:  23%|██▎       | 765/3295 [07:55<26:12,  1.61it/s]
Epoch 0:  23%|██▎       | 766/3295 [07:55<26:10,  1.61it/s]
Epoch 0:  23%|██▎       | 766/3295 [07:55<26:10,  1.61it/s]
Epoch 0:  23%|██▎       | 767/3295 [07:57<26:13,  1.61it/s]
Epoch 0:  23%|██▎       | 767/3295 [07:57<26:13,  1.61it/s]
Epoch 0:  23%|██▎       | 768/3295 [07:57<26:11,  1.61it/s]
Epoch 0:  23%|██▎       | 768/3295 [07:57<26:11,  1.61it/s]
Epoch 0:  23%|██▎       | 769/3295 [07:57<26:08,  1.61it/s]
Epoch 0:  23%|██▎       | 769/3295 [07:57<26:08,  1.61it/s]
Epoch 0:  23%|██▎       | 770/3295 [07:59<26:11,  1.61it/s]
Epoch 0:  23%|██▎       | 770/3295 [07:59<26:11,  1.61it/s]
Epoch 0:  23%|██▎       | 771/3295 [07:59<26:08,  1.61it/s]
Epoch 0:  23%|██▎       | 771/3295 [07:59<26:08,  1.61it/s]
Epoch 0:  23%|██▎       | 772/3295 [07:59<26:05,  1.61it/s]
Epoch 0:  23%|██▎       | 772/3295 [07:59<26:05,  1.61it/s]
Epoch 0:  23%|██▎       | 773/3295 [08:01<26:09,  1.61it/s]
Epoch 0:  23%|██▎       | 773/3295 [08:01<26:09,  1.61it/s]
Epoch 0:  23%|██▎       | 774/3295 [08:01<26:06,  1.61it/s]
Epoch 0:  23%|██▎       | 774/3295 [08:01<26:06,  1.61it/s]
Epoch 0:  24%|██▎       | 775/3295 [08:01<26:04,  1.61it/s]
Epoch 0:  24%|██▎       | 775/3295 [08:01<26:04,  1.61it/s]
Epoch 0:  24%|██▎       | 776/3295 [08:02<26:07,  1.61it/s]
Epoch 0:  24%|██▎       | 776/3295 [08:02<26:07,  1.61it/s]
Epoch 0:  24%|██▎       | 777/3295 [08:02<26:04,  1.61it/s]
Epoch 0:  24%|██▎       | 777/3295 [08:02<26:04,  1.61it/s]
Epoch 0:  24%|██▎       | 778/3295 [08:02<26:02,  1.61it/s]
Epoch 0:  24%|██▎       | 778/3295 [08:02<26:02,  1.61it/s]
Epoch 0:  24%|██▎       | 779/3295 [08:04<26:05,  1.61it/s]
Epoch 0:  24%|██▎       | 779/3295 [08:04<26:05,  1.61it/s]
Epoch 0:  24%|██▎       | 780/3295 [08:04<26:02,  1.61it/s]
Epoch 0:  24%|██▎       | 780/3295 [08:04<26:02,  1.61it/s]
Epoch 0:  24%|██▎       | 781/3295 [08:04<26:00,  1.61it/s]
Epoch 0:  24%|██▎       | 781/3295 [08:04<26:00,  1.61it/s]
Epoch 0:  24%|██▎       | 782/3295 [08:06<26:03,  1.61it/s]
Epoch 0:  24%|██▎       | 782/3295 [08:06<26:03,  1.61it/s]
Epoch 0:  24%|██▍       | 783/3295 [08:06<26:01,  1.61it/s]
Epoch 0:  24%|██▍       | 783/3295 [08:06<26:01,  1.61it/s]
Epoch 0:  24%|██▍       | 784/3295 [08:06<25:58,  1.61it/s]
Epoch 0:  24%|██▍       | 784/3295 [08:06<25:58,  1.61it/s]
Epoch 0:  24%|██▍       | 785/3295 [08:08<26:01,  1.61it/s]
Epoch 0:  24%|██▍       | 785/3295 [08:08<26:01,  1.61it/s]
Epoch 0:  24%|██▍       | 786/3295 [08:08<25:59,  1.61it/s]
Epoch 0:  24%|██▍       | 786/3295 [08:08<25:59,  1.61it/s]
Epoch 0:  24%|██▍       | 787/3295 [08:08<25:56,  1.61it/s]
Epoch 0:  24%|██▍       | 787/3295 [08:08<25:56,  1.61it/s]
Epoch 0:  24%|██▍       | 788/3295 [08:10<25:59,  1.61it/s]
Epoch 0:  24%|██▍       | 788/3295 [08:10<25:59,  1.61it/s]
Epoch 0:  24%|██▍       | 789/3295 [08:10<25:57,  1.61it/s]
Epoch 0:  24%|██▍       | 789/3295 [08:10<25:57,  1.61it/s]
Epoch 0:  24%|██▍       | 790/3295 [08:10<25:54,  1.61it/s]
Epoch 0:  24%|██▍       | 790/3295 [08:10<25:54,  1.61it/s]
Epoch 0:  24%|██▍       | 791/3295 [08:12<25:57,  1.61it/s]
Epoch 0:  24%|██▍       | 791/3295 [08:12<25:57,  1.61it/s]
Epoch 0:  24%|██▍       | 792/3295 [08:12<25:55,  1.61it/s]
Epoch 0:  24%|██▍       | 792/3295 [08:12<25:55,  1.61it/s]
Epoch 0:  24%|██▍       | 793/3295 [08:12<25:52,  1.61it/s]
Epoch 0:  24%|██▍       | 793/3295 [08:12<25:52,  1.61it/s]
Epoch 0:  24%|██▍       | 794/3295 [08:13<25:55,  1.61it/s]
Epoch 0:  24%|██▍       | 794/3295 [08:13<25:55,  1.61it/s]
Epoch 0:  24%|██▍       | 795/3295 [08:13<25:53,  1.61it/s]
Epoch 0:  24%|██▍       | 795/3295 [08:13<25:53,  1.61it/s]
Epoch 0:  24%|██▍       | 796/3295 [08:14<25:50,  1.61it/s]
Epoch 0:  24%|██▍       | 796/3295 [08:14<25:50,  1.61it/s]
Epoch 0:  24%|██▍       | 797/3295 [08:15<25:54,  1.61it/s]
Epoch 0:  24%|██▍       | 797/3295 [08:15<25:54,  1.61it/s]
Epoch 0:  24%|██▍       | 798/3295 [08:15<25:51,  1.61it/s]
Epoch 0:  24%|██▍       | 798/3295 [08:15<25:51,  1.61it/s]
Epoch 0:  24%|██▍       | 799/3295 [08:15<25:49,  1.61it/s]
Epoch 0:  24%|██▍       | 799/3295 [08:15<25:49,  1.61it/s]
Epoch 0:  24%|██▍       | 800/3295 [08:17<25:51,  1.61it/s]
Epoch 0:  24%|██▍       | 800/3295 [08:17<25:51,  1.61it/s]
Epoch 0:  24%|██▍       | 801/3295 [08:17<25:49,  1.61it/s]
Epoch 0:  24%|██▍       | 801/3295 [08:17<25:49,  1.61it/s]
Epoch 0:  24%|██▍       | 802/3295 [08:17<25:46,  1.61it/s]
Epoch 0:  24%|██▍       | 802/3295 [08:17<25:46,  1.61it/s]
Epoch 0:  24%|██▍       | 803/3295 [08:19<25:49,  1.61it/s]
Epoch 0:  24%|██▍       | 803/3295 [08:19<25:49,  1.61it/s]
Epoch 0:  24%|██▍       | 804/3295 [08:19<25:47,  1.61it/s]
Epoch 0:  24%|██▍       | 804/3295 [08:19<25:47,  1.61it/s]
Epoch 0:  24%|██▍       | 805/3295 [08:19<25:44,  1.61it/s]
Epoch 0:  24%|██▍       | 805/3295 [08:19<25:44,  1.61it/s]
Epoch 0:  24%|██▍       | 806/3295 [08:21<25:47,  1.61it/s]
Epoch 0:  24%|██▍       | 806/3295 [08:21<25:47,  1.61it/s]
Epoch 0:  24%|██▍       | 807/3295 [08:21<25:45,  1.61it/s]
Epoch 0:  24%|██▍       | 807/3295 [08:21<25:45,  1.61it/s]
Epoch 0:  25%|██▍       | 808/3295 [08:21<25:42,  1.61it/s]
Epoch 0:  25%|██▍       | 808/3295 [08:21<25:42,  1.61it/s]
Epoch 0:  25%|██▍       | 809/3295 [08:23<25:46,  1.61it/s]
Epoch 0:  25%|██▍       | 809/3295 [08:23<25:46,  1.61it/s]
Epoch 0:  25%|██▍       | 810/3295 [08:23<25:43,  1.61it/s]
Epoch 0:  25%|██▍       | 810/3295 [08:23<25:43,  1.61it/s]
Epoch 0:  25%|██▍       | 811/3295 [08:23<25:41,  1.61it/s]
Epoch 0:  25%|██▍       | 811/3295 [08:23<25:41,  1.61it/s]
Epoch 0:  25%|██▍       | 812/3295 [08:24<25:44,  1.61it/s]
Epoch 0:  25%|██▍       | 812/3295 [08:24<25:44,  1.61it/s]
Epoch 0:  25%|██▍       | 813/3295 [08:25<25:41,  1.61it/s]
Epoch 0:  25%|██▍       | 813/3295 [08:25<25:41,  1.61it/s]
Epoch 0:  25%|██▍       | 814/3295 [08:25<25:39,  1.61it/s]
Epoch 0:  25%|██▍       | 814/3295 [08:25<25:39,  1.61it/s]
Epoch 0:  25%|██▍       | 815/3295 [08:26<25:42,  1.61it/s]
Epoch 0:  25%|██▍       | 815/3295 [08:26<25:42,  1.61it/s]
Epoch 0:  25%|██▍       | 816/3295 [08:26<25:39,  1.61it/s]
Epoch 0:  25%|██▍       | 816/3295 [08:26<25:39,  1.61it/s]
Epoch 0:  25%|██▍       | 817/3295 [08:26<25:37,  1.61it/s]
Epoch 0:  25%|██▍       | 817/3295 [08:26<25:37,  1.61it/s]
Epoch 0:  25%|██▍       | 818/3295 [08:28<25:40,  1.61it/s]
Epoch 0:  25%|██▍       | 818/3295 [08:28<25:40,  1.61it/s]
Epoch 0:  25%|██▍       | 819/3295 [08:28<25:37,  1.61it/s]
Epoch 0:  25%|██▍       | 819/3295 [08:28<25:37,  1.61it/s]
Epoch 0:  25%|██▍       | 820/3295 [08:28<25:35,  1.61it/s]
Epoch 0:  25%|██▍       | 820/3295 [08:28<25:35,  1.61it/s]
Epoch 0:  25%|██▍       | 821/3295 [08:30<25:38,  1.61it/s]
Epoch 0:  25%|██▍       | 821/3295 [08:30<25:38,  1.61it/s]
Epoch 0:  25%|██▍       | 822/3295 [08:30<25:35,  1.61it/s]
Epoch 0:  25%|██▍       | 822/3295 [08:30<25:35,  1.61it/s]
Epoch 0:  25%|██▍       | 823/3295 [08:30<25:33,  1.61it/s]
Epoch 0:  25%|██▍       | 823/3295 [08:30<25:33,  1.61it/s]
Epoch 0:  25%|██▌       | 824/3295 [08:32<25:36,  1.61it/s]
Epoch 0:  25%|██▌       | 824/3295 [08:32<25:36,  1.61it/s]
Epoch 0:  25%|██▌       | 825/3295 [08:32<25:33,  1.61it/s]
Epoch 0:  25%|██▌       | 825/3295 [08:32<25:33,  1.61it/s]
Epoch 0:  25%|██▌       | 826/3295 [08:32<25:31,  1.61it/s]
Epoch 0:  25%|██▌       | 826/3295 [08:32<25:31,  1.61it/s]
Epoch 0:  25%|██▌       | 827/3295 [08:34<25:33,  1.61it/s]
Epoch 0:  25%|██▌       | 827/3295 [08:34<25:33,  1.61it/s]
Epoch 0:  25%|██▌       | 828/3295 [08:34<25:31,  1.61it/s]
Epoch 0:  25%|██▌       | 828/3295 [08:34<25:31,  1.61it/s]
Epoch 0:  25%|██▌       | 829/3295 [08:34<25:29,  1.61it/s]
Epoch 0:  25%|██▌       | 829/3295 [08:34<25:29,  1.61it/s]
Epoch 0:  25%|██▌       | 830/3295 [08:35<25:31,  1.61it/s]
Epoch 0:  25%|██▌       | 830/3295 [08:35<25:31,  1.61it/s]
Epoch 0:  25%|██▌       | 831/3295 [08:35<25:29,  1.61it/s]
Epoch 0:  25%|██▌       | 831/3295 [08:35<25:29,  1.61it/s]
Epoch 0:  25%|██▌       | 832/3295 [08:35<25:26,  1.61it/s]
Epoch 0:  25%|██▌       | 832/3295 [08:35<25:26,  1.61it/s]
Epoch 0:  25%|██▌       | 833/3295 [08:37<25:29,  1.61it/s]
Epoch 0:  25%|██▌       | 833/3295 [08:37<25:29,  1.61it/s]
Epoch 0:  25%|██▌       | 834/3295 [08:37<25:27,  1.61it/s]
Epoch 0:  25%|██▌       | 834/3295 [08:37<25:27,  1.61it/s]
Epoch 0:  25%|██▌       | 835/3295 [08:37<25:24,  1.61it/s]
Epoch 0:  25%|██▌       | 835/3295 [08:37<25:24,  1.61it/s]
Epoch 0:  25%|██▌       | 836/3295 [08:39<25:27,  1.61it/s]
Epoch 0:  25%|██▌       | 836/3295 [08:39<25:27,  1.61it/s]
Epoch 0:  25%|██▌       | 837/3295 [08:39<25:24,  1.61it/s]
Epoch 0:  25%|██▌       | 837/3295 [08:39<25:24,  1.61it/s]
Epoch 0:  25%|██▌       | 838/3295 [08:39<25:22,  1.61it/s]
Epoch 0:  25%|██▌       | 838/3295 [08:39<25:22,  1.61it/s]
Epoch 0:  25%|██▌       | 839/3295 [08:41<25:25,  1.61it/s]
Epoch 0:  25%|██▌       | 839/3295 [08:41<25:25,  1.61it/s]
Epoch 0:  25%|██▌       | 840/3295 [08:41<25:23,  1.61it/s]
Epoch 0:  25%|██▌       | 840/3295 [08:41<25:23,  1.61it/s]
Epoch 0:  26%|██▌       | 841/3295 [08:41<25:20,  1.61it/s]
Epoch 0:  26%|██▌       | 841/3295 [08:41<25:20,  1.61it/s]
Epoch 0:  26%|██▌       | 842/3295 [08:43<25:23,  1.61it/s]
Epoch 0:  26%|██▌       | 842/3295 [08:43<25:23,  1.61it/s]
Epoch 0:  26%|██▌       | 843/3295 [08:43<25:21,  1.61it/s]
Epoch 0:  26%|██▌       | 843/3295 [08:43<25:21,  1.61it/s]
Epoch 0:  26%|██▌       | 844/3295 [08:43<25:19,  1.61it/s]
Epoch 0:  26%|██▌       | 844/3295 [08:43<25:19,  1.61it/s]
Epoch 0:  26%|██▌       | 845/3295 [08:44<25:21,  1.61it/s]
Epoch 0:  26%|██▌       | 845/3295 [08:44<25:21,  1.61it/s]
Epoch 0:  26%|██▌       | 846/3295 [08:44<25:19,  1.61it/s]
Epoch 0:  26%|██▌       | 846/3295 [08:44<25:19,  1.61it/s]
Epoch 0:  26%|██▌       | 847/3295 [08:44<25:17,  1.61it/s]
Epoch 0:  26%|██▌       | 847/3295 [08:44<25:17,  1.61it/s]
Epoch 0:  26%|██▌       | 848/3295 [08:46<25:20,  1.61it/s]
Epoch 0:  26%|██▌       | 848/3295 [08:46<25:20,  1.61it/s]
Epoch 0:  26%|██▌       | 849/3295 [08:46<25:17,  1.61it/s]
Epoch 0:  26%|██▌       | 849/3295 [08:46<25:17,  1.61it/s]
Epoch 0:  26%|██▌       | 850/3295 [08:46<25:15,  1.61it/s]
Epoch 0:  26%|██▌       | 850/3295 [08:46<25:15,  1.61it/s]
Epoch 0:  26%|██▌       | 851/3295 [08:48<25:18,  1.61it/s]
Epoch 0:  26%|██▌       | 851/3295 [08:48<25:18,  1.61it/s]
Epoch 0:  26%|██▌       | 852/3295 [08:48<25:15,  1.61it/s]
Epoch 0:  26%|██▌       | 852/3295 [08:48<25:15,  1.61it/s]
Epoch 0:  26%|██▌       | 853/3295 [08:48<25:13,  1.61it/s]
Epoch 0:  26%|██▌       | 853/3295 [08:48<25:13,  1.61it/s]
Epoch 0:  26%|██▌       | 854/3295 [08:50<25:15,  1.61it/s]
Epoch 0:  26%|██▌       | 854/3295 [08:50<25:15,  1.61it/s]
Epoch 0:  26%|██▌       | 855/3295 [08:50<25:13,  1.61it/s]
Epoch 0:  26%|██▌       | 855/3295 [08:50<25:13,  1.61it/s]
Epoch 0:  26%|██▌       | 856/3295 [08:50<25:11,  1.61it/s]
Epoch 0:  26%|██▌       | 856/3295 [08:50<25:11,  1.61it/s]
Epoch 0:  26%|██▌       | 857/3295 [08:52<25:13,  1.61it/s]
Epoch 0:  26%|██▌       | 857/3295 [08:52<25:13,  1.61it/s]
Epoch 0:  26%|██▌       | 858/3295 [08:52<25:11,  1.61it/s]
Epoch 0:  26%|██▌       | 858/3295 [08:52<25:11,  1.61it/s]
Epoch 0:  26%|██▌       | 859/3295 [08:52<25:09,  1.61it/s]
Epoch 0:  26%|██▌       | 859/3295 [08:52<25:09,  1.61it/s]
Epoch 0:  26%|██▌       | 860/3295 [08:54<25:12,  1.61it/s]
Epoch 0:  26%|██▌       | 860/3295 [08:54<25:12,  1.61it/s]
Epoch 0:  26%|██▌       | 861/3295 [08:54<25:10,  1.61it/s]
Epoch 0:  26%|██▌       | 861/3295 [08:54<25:10,  1.61it/s]
Epoch 0:  26%|██▌       | 862/3295 [08:54<25:07,  1.61it/s]
Epoch 0:  26%|██▌       | 862/3295 [08:54<25:07,  1.61it/s]
Epoch 0:  26%|██▌       | 863/3295 [08:55<25:10,  1.61it/s]
Epoch 0:  26%|██▌       | 863/3295 [08:55<25:10,  1.61it/s]
Epoch 0:  26%|██▌       | 864/3295 [08:55<25:08,  1.61it/s]
Epoch 0:  26%|██▌       | 864/3295 [08:55<25:08,  1.61it/s]
Epoch 0:  26%|██▋       | 865/3295 [08:56<25:05,  1.61it/s]
Epoch 0:  26%|██▋       | 865/3295 [08:56<25:05,  1.61it/s]
Epoch 0:  26%|██▋       | 866/3295 [08:57<25:08,  1.61it/s]
Epoch 0:  26%|██▋       | 866/3295 [08:57<25:08,  1.61it/s]
Epoch 0:  26%|██▋       | 867/3295 [08:57<25:06,  1.61it/s]
Epoch 0:  26%|██▋       | 867/3295 [08:57<25:06,  1.61it/s]
Epoch 0:  26%|██▋       | 868/3295 [08:57<25:03,  1.61it/s]
Epoch 0:  26%|██▋       | 868/3295 [08:57<25:03,  1.61it/s]
Epoch 0:  26%|██▋       | 869/3295 [08:59<25:06,  1.61it/s]
Epoch 0:  26%|██▋       | 869/3295 [08:59<25:06,  1.61it/s]
Epoch 0:  26%|██▋       | 870/3295 [08:59<25:03,  1.61it/s]
Epoch 0:  26%|██▋       | 870/3295 [08:59<25:03,  1.61it/s]
Epoch 0:  26%|██▋       | 871/3295 [08:59<25:01,  1.61it/s]
Epoch 0:  26%|██▋       | 871/3295 [08:59<25:01,  1.61it/s]
Epoch 0:  26%|██▋       | 872/3295 [09:01<25:04,  1.61it/s]
Epoch 0:  26%|██▋       | 872/3295 [09:01<25:04,  1.61it/s]
Epoch 0:  26%|██▋       | 873/3295 [09:01<25:02,  1.61it/s]
Epoch 0:  26%|██▋       | 873/3295 [09:01<25:02,  1.61it/s]
Epoch 0:  27%|██▋       | 874/3295 [09:01<25:00,  1.61it/s]
Epoch 0:  27%|██▋       | 874/3295 [09:01<25:00,  1.61it/s]
Epoch 0:  27%|██▋       | 875/3295 [09:03<25:02,  1.61it/s]
Epoch 0:  27%|██▋       | 875/3295 [09:03<25:02,  1.61it/s]
Epoch 0:  27%|██▋       | 876/3295 [09:03<25:00,  1.61it/s]
Epoch 0:  27%|██▋       | 876/3295 [09:03<25:00,  1.61it/s]
Epoch 0:  27%|██▋       | 877/3295 [09:03<24:58,  1.61it/s]
Epoch 0:  27%|██▋       | 877/3295 [09:03<24:58,  1.61it/s]
Epoch 0:  27%|██▋       | 878/3295 [09:05<25:01,  1.61it/s]
Epoch 0:  27%|██▋       | 878/3295 [09:05<25:01,  1.61it/s]
Epoch 0:  27%|██▋       | 879/3295 [09:05<24:59,  1.61it/s]
Epoch 0:  27%|██▋       | 879/3295 [09:05<24:59,  1.61it/s]
Epoch 0:  27%|██▋       | 880/3295 [09:05<24:56,  1.61it/s]
Epoch 0:  27%|██▋       | 880/3295 [09:05<24:56,  1.61it/s]
Epoch 0:  27%|██▋       | 881/3295 [09:07<24:59,  1.61it/s]
Epoch 0:  27%|██▋       | 881/3295 [09:07<24:59,  1.61it/s]
Epoch 0:  27%|██▋       | 882/3295 [09:07<24:57,  1.61it/s]
Epoch 0:  27%|██▋       | 882/3295 [09:07<24:57,  1.61it/s]
Epoch 0:  27%|██▋       | 883/3295 [09:07<24:55,  1.61it/s]
Epoch 0:  27%|██▋       | 883/3295 [09:07<24:55,  1.61it/s]
Epoch 0:  27%|██▋       | 884/3295 [09:09<24:57,  1.61it/s]
Epoch 0:  27%|██▋       | 884/3295 [09:09<24:57,  1.61it/s]
Epoch 0:  27%|██▋       | 885/3295 [09:09<24:55,  1.61it/s]
Epoch 0:  27%|██▋       | 885/3295 [09:09<24:55,  1.61it/s]
Epoch 0:  27%|██▋       | 886/3295 [09:09<24:53,  1.61it/s]
Epoch 0:  27%|██▋       | 886/3295 [09:09<24:53,  1.61it/s]
Epoch 0:  27%|██▋       | 887/3295 [09:10<24:55,  1.61it/s]
Epoch 0:  27%|██▋       | 887/3295 [09:10<24:55,  1.61it/s]
Epoch 0:  27%|██▋       | 888/3295 [09:11<24:53,  1.61it/s]
Epoch 0:  27%|██▋       | 888/3295 [09:11<24:53,  1.61it/s]
Epoch 0:  27%|██▋       | 889/3295 [09:11<24:51,  1.61it/s]
Epoch 0:  27%|██▋       | 889/3295 [09:11<24:51,  1.61it/s]
Epoch 0:  27%|██▋       | 890/3295 [09:12<24:54,  1.61it/s]
Epoch 0:  27%|██▋       | 890/3295 [09:12<24:54,  1.61it/s]
Epoch 0:  27%|██▋       | 891/3295 [09:12<24:51,  1.61it/s]
Epoch 0:  27%|██▋       | 891/3295 [09:12<24:51,  1.61it/s]
Epoch 0:  27%|██▋       | 892/3295 [09:12<24:49,  1.61it/s]
Epoch 0:  27%|██▋       | 892/3295 [09:12<24:49,  1.61it/s]
Epoch 0:  27%|██▋       | 893/3295 [09:14<24:52,  1.61it/s]
Epoch 0:  27%|██▋       | 893/3295 [09:14<24:52,  1.61it/s]
Epoch 0:  27%|██▋       | 894/3295 [09:14<24:49,  1.61it/s]
Epoch 0:  27%|██▋       | 894/3295 [09:14<24:49,  1.61it/s]
Epoch 0:  27%|██▋       | 895/3295 [09:14<24:47,  1.61it/s]
Epoch 0:  27%|██▋       | 895/3295 [09:14<24:47,  1.61it/s]
Epoch 0:  27%|██▋       | 896/3295 [09:16<24:50,  1.61it/s]
Epoch 0:  27%|██▋       | 896/3295 [09:16<24:50,  1.61it/s]
Epoch 0:  27%|██▋       | 897/3295 [09:16<24:47,  1.61it/s]
Epoch 0:  27%|██▋       | 897/3295 [09:16<24:47,  1.61it/s]
Epoch 0:  27%|██▋       | 898/3295 [09:16<24:45,  1.61it/s]
Epoch 0:  27%|██▋       | 898/3295 [09:16<24:45,  1.61it/s]
Epoch 0:  27%|██▋       | 899/3295 [09:18<24:48,  1.61it/s]
Epoch 0:  27%|██▋       | 899/3295 [09:18<24:48,  1.61it/s]
Epoch 0:  27%|██▋       | 900/3295 [09:18<24:46,  1.61it/s]
Epoch 0:  27%|██▋       | 900/3295 [09:18<24:46,  1.61it/s]
Epoch 0:  27%|██▋       | 901/3295 [09:18<24:43,  1.61it/s]
Epoch 0:  27%|██▋       | 901/3295 [09:18<24:43,  1.61it/s]
Epoch 0:  27%|██▋       | 902/3295 [09:20<24:46,  1.61it/s]
Epoch 0:  27%|██▋       | 902/3295 [09:20<24:46,  1.61it/s]
Epoch 0:  27%|██▋       | 903/3295 [09:20<24:43,  1.61it/s]
Epoch 0:  27%|██▋       | 903/3295 [09:20<24:43,  1.61it/s]
Epoch 0:  27%|██▋       | 904/3295 [09:20<24:41,  1.61it/s]
Epoch 0:  27%|██▋       | 904/3295 [09:20<24:41,  1.61it/s]
Epoch 0:  27%|██▋       | 905/3295 [09:21<24:44,  1.61it/s]
Epoch 0:  27%|██▋       | 905/3295 [09:21<24:44,  1.61it/s]
Epoch 0:  27%|██▋       | 906/3295 [09:22<24:41,  1.61it/s]
Epoch 0:  27%|██▋       | 906/3295 [09:22<24:41,  1.61it/s]
Epoch 0:  28%|██▊       | 907/3295 [09:22<24:39,  1.61it/s]
Epoch 0:  28%|██▊       | 907/3295 [09:22<24:39,  1.61it/s]
Epoch 0:  28%|██▊       | 908/3295 [09:23<24:42,  1.61it/s]
Epoch 0:  28%|██▊       | 908/3295 [09:23<24:42,  1.61it/s]
Epoch 0:  28%|██▊       | 909/3295 [09:23<24:39,  1.61it/s]
Epoch 0:  28%|██▊       | 909/3295 [09:23<24:39,  1.61it/s]
Epoch 0:  28%|██▊       | 910/3295 [09:23<24:37,  1.61it/s]
Epoch 0:  28%|██▊       | 910/3295 [09:23<24:37,  1.61it/s]
Epoch 0:  28%|██▊       | 911/3295 [09:25<24:40,  1.61it/s]
Epoch 0:  28%|██▊       | 911/3295 [09:25<24:40,  1.61it/s]
Epoch 0:  28%|██▊       | 912/3295 [09:25<24:38,  1.61it/s]
Epoch 0:  28%|██▊       | 912/3295 [09:25<24:38,  1.61it/s]
Epoch 0:  28%|██▊       | 913/3295 [09:25<24:36,  1.61it/s]
Epoch 0:  28%|██▊       | 913/3295 [09:25<24:36,  1.61it/s]
Epoch 0:  28%|██▊       | 914/3295 [09:27<24:38,  1.61it/s]
Epoch 0:  28%|██▊       | 914/3295 [09:27<24:38,  1.61it/s]
Epoch 0:  28%|██▊       | 915/3295 [09:27<24:36,  1.61it/s]
Epoch 0:  28%|██▊       | 915/3295 [09:27<24:36,  1.61it/s]
Epoch 0:  28%|██▊       | 916/3295 [09:27<24:34,  1.61it/s]
Epoch 0:  28%|██▊       | 916/3295 [09:27<24:34,  1.61it/s]
Epoch 0:  28%|██▊       | 917/3295 [09:29<24:37,  1.61it/s]
Epoch 0:  28%|██▊       | 917/3295 [09:29<24:37,  1.61it/s]
Epoch 0:  28%|██▊       | 918/3295 [09:29<24:35,  1.61it/s]
Epoch 0:  28%|██▊       | 918/3295 [09:29<24:35,  1.61it/s]
Epoch 0:  28%|██▊       | 919/3295 [09:29<24:33,  1.61it/s]
Epoch 0:  28%|██▊       | 919/3295 [09:29<24:33,  1.61it/s]
Epoch 0:  28%|██▊       | 920/3295 [09:31<24:35,  1.61it/s]
Epoch 0:  28%|██▊       | 920/3295 [09:31<24:35,  1.61it/s]
Epoch 0:  28%|██▊       | 921/3295 [09:31<24:33,  1.61it/s]
Epoch 0:  28%|██▊       | 921/3295 [09:31<24:33,  1.61it/s]
Epoch 0:  28%|██▊       | 922/3295 [09:31<24:31,  1.61it/s]
Epoch 0:  28%|██▊       | 922/3295 [09:31<24:31,  1.61it/s]
Epoch 0:  28%|██▊       | 923/3295 [09:33<24:33,  1.61it/s]
Epoch 0:  28%|██▊       | 923/3295 [09:33<24:33,  1.61it/s]
Epoch 0:  28%|██▊       | 924/3295 [09:33<24:31,  1.61it/s]
Epoch 0:  28%|██▊       | 924/3295 [09:33<24:31,  1.61it/s]
Epoch 0:  28%|██▊       | 925/3295 [09:33<24:29,  1.61it/s]
Epoch 0:  28%|██▊       | 925/3295 [09:33<24:29,  1.61it/s]
Epoch 0:  28%|██▊       | 926/3295 [09:35<24:31,  1.61it/s]
Epoch 0:  28%|██▊       | 926/3295 [09:35<24:31,  1.61it/s]
Epoch 0:  28%|██▊       | 927/3295 [09:35<24:29,  1.61it/s]
Epoch 0:  28%|██▊       | 927/3295 [09:35<24:29,  1.61it/s]
Epoch 0:  28%|██▊       | 928/3295 [09:35<24:27,  1.61it/s]
Epoch 0:  28%|██▊       | 928/3295 [09:35<24:27,  1.61it/s]
Epoch 0:  28%|██▊       | 929/3295 [09:36<24:29,  1.61it/s]
Epoch 0:  28%|██▊       | 929/3295 [09:36<24:29,  1.61it/s]
Epoch 0:  28%|██▊       | 930/3295 [09:37<24:27,  1.61it/s]
Epoch 0:  28%|██▊       | 930/3295 [09:37<24:27,  1.61it/s]
Epoch 0:  28%|██▊       | 931/3295 [09:37<24:25,  1.61it/s]
Epoch 0:  28%|██▊       | 931/3295 [09:37<24:25,  1.61it/s]
Epoch 0:  28%|██▊       | 932/3295 [09:38<24:27,  1.61it/s]
Epoch 0:  28%|██▊       | 932/3295 [09:38<24:27,  1.61it/s]
Epoch 0:  28%|██▊       | 933/3295 [09:38<24:25,  1.61it/s]
Epoch 0:  28%|██▊       | 933/3295 [09:38<24:25,  1.61it/s]
Epoch 0:  28%|██▊       | 934/3295 [09:38<24:23,  1.61it/s]
Epoch 0:  28%|██▊       | 934/3295 [09:38<24:23,  1.61it/s]
Epoch 0:  28%|██▊       | 935/3295 [09:40<24:25,  1.61it/s]
Epoch 0:  28%|██▊       | 935/3295 [09:40<24:25,  1.61it/s]
Epoch 0:  28%|██▊       | 936/3295 [09:40<24:23,  1.61it/s]
Epoch 0:  28%|██▊       | 936/3295 [09:40<24:23,  1.61it/s]
Epoch 0:  28%|██▊       | 937/3295 [09:40<24:21,  1.61it/s]
Epoch 0:  28%|██▊       | 937/3295 [09:40<24:21,  1.61it/s]
Epoch 0:  28%|██▊       | 938/3295 [09:42<24:23,  1.61it/s]
Epoch 0:  28%|██▊       | 938/3295 [09:42<24:23,  1.61it/s]
Epoch 0:  28%|██▊       | 939/3295 [09:42<24:21,  1.61it/s]
Epoch 0:  28%|██▊       | 939/3295 [09:42<24:21,  1.61it/s]
Epoch 0:  29%|██▊       | 940/3295 [09:42<24:19,  1.61it/s]
Epoch 0:  29%|██▊       | 940/3295 [09:42<24:19,  1.61it/s]
Epoch 0:  29%|██▊       | 941/3295 [09:44<24:21,  1.61it/s]
Epoch 0:  29%|██▊       | 941/3295 [09:44<24:21,  1.61it/s]
Epoch 0:  29%|██▊       | 942/3295 [09:44<24:19,  1.61it/s]
Epoch 0:  29%|██▊       | 942/3295 [09:44<24:19,  1.61it/s]
Epoch 0:  29%|██▊       | 943/3295 [09:44<24:17,  1.61it/s]
Epoch 0:  29%|██▊       | 943/3295 [09:44<24:17,  1.61it/s]
Epoch 0:  29%|██▊       | 944/3295 [09:45<24:19,  1.61it/s]
Epoch 0:  29%|██▊       | 944/3295 [09:45<24:19,  1.61it/s]
Epoch 0:  29%|██▊       | 945/3295 [09:46<24:17,  1.61it/s]
Epoch 0:  29%|██▊       | 945/3295 [09:46<24:17,  1.61it/s]
Epoch 0:  29%|██▊       | 946/3295 [09:46<24:15,  1.61it/s]
Epoch 0:  29%|██▊       | 946/3295 [09:46<24:15,  1.61it/s]
Epoch 0:  29%|██▊       | 947/3295 [09:47<24:17,  1.61it/s]
Epoch 0:  29%|██▊       | 947/3295 [09:47<24:17,  1.61it/s]
Epoch 0:  29%|██▉       | 948/3295 [09:47<24:15,  1.61it/s]
Epoch 0:  29%|██▉       | 948/3295 [09:47<24:15,  1.61it/s]
Epoch 0:  29%|██▉       | 949/3295 [09:47<24:13,  1.61it/s]
Epoch 0:  29%|██▉       | 949/3295 [09:47<24:13,  1.61it/s]
Epoch 0:  29%|██▉       | 950/3295 [09:49<24:15,  1.61it/s]
Epoch 0:  29%|██▉       | 950/3295 [09:49<24:15,  1.61it/s]
Epoch 0:  29%|██▉       | 951/3295 [09:49<24:13,  1.61it/s]
Epoch 0:  29%|██▉       | 951/3295 [09:49<24:13,  1.61it/s]
Epoch 0:  29%|██▉       | 952/3295 [09:49<24:11,  1.61it/s]
Epoch 0:  29%|██▉       | 952/3295 [09:49<24:11,  1.61it/s]
Epoch 0:  29%|██▉       | 953/3295 [09:51<24:13,  1.61it/s]
Epoch 0:  29%|██▉       | 953/3295 [09:51<24:13,  1.61it/s]
Epoch 0:  29%|██▉       | 954/3295 [09:51<24:11,  1.61it/s]
Epoch 0:  29%|██▉       | 954/3295 [09:51<24:11,  1.61it/s]
Epoch 0:  29%|██▉       | 955/3295 [09:51<24:09,  1.61it/s]
Epoch 0:  29%|██▉       | 955/3295 [09:51<24:09,  1.61it/s]
Epoch 0:  29%|██▉       | 956/3295 [09:53<24:11,  1.61it/s]
Epoch 0:  29%|██▉       | 956/3295 [09:53<24:11,  1.61it/s]
Epoch 0:  29%|██▉       | 957/3295 [09:53<24:09,  1.61it/s]
Epoch 0:  29%|██▉       | 957/3295 [09:53<24:09,  1.61it/s]
Epoch 0:  29%|██▉       | 958/3295 [09:53<24:07,  1.61it/s]
Epoch 0:  29%|██▉       | 958/3295 [09:53<24:07,  1.61it/s]
Epoch 0:  29%|██▉       | 959/3295 [09:54<24:09,  1.61it/s]
Epoch 0:  29%|██▉       | 959/3295 [09:54<24:09,  1.61it/s]
Epoch 0:  29%|██▉       | 960/3295 [09:55<24:07,  1.61it/s]
Epoch 0:  29%|██▉       | 960/3295 [09:55<24:07,  1.61it/s]
Epoch 0:  29%|██▉       | 961/3295 [09:55<24:05,  1.61it/s]
Epoch 0:  29%|██▉       | 961/3295 [09:55<24:05,  1.61it/s]
Epoch 0:  29%|██▉       | 962/3295 [09:56<24:07,  1.61it/s]
Epoch 0:  29%|██▉       | 962/3295 [09:56<24:07,  1.61it/s]
Epoch 0:  29%|██▉       | 963/3295 [09:56<24:05,  1.61it/s]
Epoch 0:  29%|██▉       | 963/3295 [09:56<24:05,  1.61it/s]
Epoch 0:  29%|██▉       | 964/3295 [09:56<24:03,  1.61it/s]
Epoch 0:  29%|██▉       | 964/3295 [09:56<24:03,  1.61it/s]
Epoch 0:  29%|██▉       | 965/3295 [09:58<24:05,  1.61it/s]
Epoch 0:  29%|██▉       | 965/3295 [09:58<24:05,  1.61it/s]
Epoch 0:  29%|██▉       | 966/3295 [09:58<24:03,  1.61it/s]
Epoch 0:  29%|██▉       | 966/3295 [09:58<24:03,  1.61it/s]
Epoch 0:  29%|██▉       | 967/3295 [09:58<24:01,  1.61it/s]
Epoch 0:  29%|██▉       | 967/3295 [09:58<24:01,  1.61it/s]
Epoch 0:  29%|██▉       | 968/3295 [10:00<24:04,  1.61it/s]
Epoch 0:  29%|██▉       | 968/3295 [10:00<24:04,  1.61it/s]
Epoch 0:  29%|██▉       | 969/3295 [10:00<24:01,  1.61it/s]
Epoch 0:  29%|██▉       | 969/3295 [10:00<24:01,  1.61it/s]
Epoch 0:  29%|██▉       | 970/3295 [10:00<23:59,  1.61it/s]
Epoch 0:  29%|██▉       | 970/3295 [10:00<23:59,  1.61it/s]
Epoch 0:  29%|██▉       | 971/3295 [10:02<24:02,  1.61it/s]
Epoch 0:  29%|██▉       | 971/3295 [10:02<24:02,  1.61it/s]
Epoch 0:  29%|██▉       | 972/3295 [10:02<24:00,  1.61it/s]
Epoch 0:  29%|██▉       | 972/3295 [10:02<24:00,  1.61it/s]
Epoch 0:  30%|██▉       | 973/3295 [10:02<23:58,  1.61it/s]
Epoch 0:  30%|██▉       | 973/3295 [10:02<23:58,  1.61it/s]
Epoch 0:  30%|██▉       | 974/3295 [10:04<24:00,  1.61it/s]
Epoch 0:  30%|██▉       | 974/3295 [10:04<24:00,  1.61it/s]
Epoch 0:  30%|██▉       | 975/3295 [10:04<23:58,  1.61it/s]
Epoch 0:  30%|██▉       | 975/3295 [10:04<23:58,  1.61it/s]
Epoch 0:  30%|██▉       | 976/3295 [10:04<23:56,  1.61it/s]
Epoch 0:  30%|██▉       | 976/3295 [10:04<23:56,  1.61it/s]
Epoch 0:  30%|██▉       | 977/3295 [10:06<23:59,  1.61it/s]
Epoch 0:  30%|██▉       | 977/3295 [10:06<23:59,  1.61it/s]
Epoch 0:  30%|██▉       | 978/3295 [10:06<23:56,  1.61it/s]
Epoch 0:  30%|██▉       | 978/3295 [10:06<23:56,  1.61it/s]
Epoch 0:  30%|██▉       | 979/3295 [10:06<23:54,  1.61it/s]
Epoch 0:  30%|██▉       | 979/3295 [10:06<23:54,  1.61it/s]
Epoch 0:  30%|██▉       | 980/3295 [10:08<23:57,  1.61it/s]
Epoch 0:  30%|██▉       | 980/3295 [10:08<23:57,  1.61it/s]
Epoch 0:  30%|██▉       | 981/3295 [10:08<23:55,  1.61it/s]
Epoch 0:  30%|██▉       | 981/3295 [10:08<23:55,  1.61it/s]
Epoch 0:  30%|██▉       | 982/3295 [10:08<23:53,  1.61it/s]
Epoch 0:  30%|██▉       | 982/3295 [10:08<23:53,  1.61it/s]
Epoch 0:  30%|██▉       | 983/3295 [10:10<23:55,  1.61it/s]
Epoch 0:  30%|██▉       | 983/3295 [10:10<23:55,  1.61it/s]
Epoch 0:  30%|██▉       | 984/3295 [10:10<23:53,  1.61it/s]
Epoch 0:  30%|██▉       | 984/3295 [10:10<23:53,  1.61it/s]
Epoch 0:  30%|██▉       | 985/3295 [10:10<23:51,  1.61it/s]
Epoch 0:  30%|██▉       | 985/3295 [10:10<23:51,  1.61it/s]
Epoch 0:  30%|██▉       | 986/3295 [10:12<23:53,  1.61it/s]
Epoch 0:  30%|██▉       | 986/3295 [10:12<23:53,  1.61it/s]
Epoch 0:  30%|██▉       | 987/3295 [10:12<23:51,  1.61it/s]
Epoch 0:  30%|██▉       | 987/3295 [10:12<23:51,  1.61it/s]
Epoch 0:  30%|██▉       | 988/3295 [10:12<23:49,  1.61it/s]
Epoch 0:  30%|██▉       | 988/3295 [10:12<23:49,  1.61it/s]
Epoch 0:  30%|███       | 989/3295 [10:14<23:52,  1.61it/s]
Epoch 0:  30%|███       | 989/3295 [10:14<23:52,  1.61it/s]
Epoch 0:  30%|███       | 990/3295 [10:14<23:50,  1.61it/s]
Epoch 0:  30%|███       | 990/3295 [10:14<23:50,  1.61it/s]
Epoch 0:  30%|███       | 991/3295 [10:14<23:48,  1.61it/s]
Epoch 0:  30%|███       | 991/3295 [10:14<23:48,  1.61it/s]
Epoch 0:  30%|███       | 992/3295 [10:16<23:50,  1.61it/s]
Epoch 0:  30%|███       | 992/3295 [10:16<23:50,  1.61it/s]
Epoch 0:  30%|███       | 993/3295 [10:16<23:48,  1.61it/s]
Epoch 0:  30%|███       | 993/3295 [10:16<23:48,  1.61it/s]
Epoch 0:  30%|███       | 994/3295 [10:16<23:46,  1.61it/s]
Epoch 0:  30%|███       | 994/3295 [10:16<23:46,  1.61it/s]
Epoch 0:  30%|███       | 995/3295 [10:17<23:48,  1.61it/s]
Epoch 0:  30%|███       | 995/3295 [10:17<23:48,  1.61it/s]
Epoch 0:  30%|███       | 996/3295 [10:17<23:46,  1.61it/s]
Epoch 0:  30%|███       | 996/3295 [10:17<23:46,  1.61it/s]
Epoch 0:  30%|███       | 997/3295 [10:18<23:44,  1.61it/s]
Epoch 0:  30%|███       | 997/3295 [10:18<23:44,  1.61it/s]
Epoch 0:  30%|███       | 998/3295 [10:19<23:46,  1.61it/s]
Epoch 0:  30%|███       | 998/3295 [10:19<23:46,  1.61it/s]
Epoch 0:  30%|███       | 999/3295 [10:19<23:44,  1.61it/s]
Epoch 0:  30%|███       | 999/3295 [10:19<23:44,  1.61it/s]
Epoch 0:  30%|███       | 1000/3295 [10:20<23:42,  1.61it/s]
Epoch 0:  30%|███       | 1000/3295 [10:20<23:42,  1.61it/s]
Epoch 0:  30%|███       | 1001/3295 [10:21<23:45,  1.61it/s]
Epoch 0:  30%|███       | 1001/3295 [10:21<23:45,  1.61it/s]
Epoch 0:  30%|███       | 1002/3295 [10:21<23:43,  1.61it/s]
Epoch 0:  30%|███       | 1002/3295 [10:21<23:43,  1.61it/s]
Epoch 0:  30%|███       | 1003/3295 [10:21<23:41,  1.61it/s]
Epoch 0:  30%|███       | 1003/3295 [10:21<23:41,  1.61it/s]
Epoch 0:  30%|███       | 1004/3295 [10:24<23:44,  1.61it/s]
Epoch 0:  30%|███       | 1004/3295 [10:24<23:44,  1.61it/s]
Epoch 0:  31%|███       | 1005/3295 [10:24<23:42,  1.61it/s]
Epoch 0:  31%|███       | 1005/3295 [10:24<23:42,  1.61it/s]
Epoch 0:  31%|███       | 1006/3295 [10:24<23:40,  1.61it/s]
Epoch 0:  31%|███       | 1006/3295 [10:24<23:40,  1.61it/s]
Epoch 0:  31%|███       | 1007/3295 [10:26<23:43,  1.61it/s]
Epoch 0:  31%|███       | 1007/3295 [10:26<23:43,  1.61it/s]
Epoch 0:  31%|███       | 1008/3295 [10:26<23:41,  1.61it/s]
Epoch 0:  31%|███       | 1008/3295 [10:26<23:41,  1.61it/s]
Epoch 0:  31%|███       | 1009/3295 [10:26<23:39,  1.61it/s]
Epoch 0:  31%|███       | 1009/3295 [10:26<23:39,  1.61it/s]
Epoch 0:  31%|███       | 1010/3295 [10:28<23:42,  1.61it/s]
Epoch 0:  31%|███       | 1010/3295 [10:28<23:42,  1.61it/s]
Epoch 0:  31%|███       | 1011/3295 [10:28<23:40,  1.61it/s]
Epoch 0:  31%|███       | 1011/3295 [10:28<23:40,  1.61it/s]
Epoch 0:  31%|███       | 1012/3295 [10:28<23:38,  1.61it/s]
Epoch 0:  31%|███       | 1012/3295 [10:28<23:38,  1.61it/s]
Epoch 0:  31%|███       | 1013/3295 [10:30<23:41,  1.61it/s]
Epoch 0:  31%|███       | 1013/3295 [10:30<23:41,  1.61it/s]
Epoch 0:  31%|███       | 1014/3295 [10:31<23:39,  1.61it/s]
Epoch 0:  31%|███       | 1014/3295 [10:31<23:39,  1.61it/s]
Epoch 0:  31%|███       | 1015/3295 [10:31<23:37,  1.61it/s]
Epoch 0:  31%|███       | 1015/3295 [10:31<23:37,  1.61it/s]
Epoch 0:  31%|███       | 1016/3295 [10:33<23:40,  1.60it/s]
Epoch 0:  31%|███       | 1016/3295 [10:33<23:40,  1.60it/s]
Epoch 0:  31%|███       | 1017/3295 [10:33<23:38,  1.61it/s]
Epoch 0:  31%|███       | 1017/3295 [10:33<23:38,  1.61it/s]
Epoch 0:  31%|███       | 1018/3295 [10:33<23:36,  1.61it/s]
Epoch 0:  31%|███       | 1018/3295 [10:33<23:36,  1.61it/s]
Epoch 0:  31%|███       | 1019/3295 [10:35<23:39,  1.60it/s]
Epoch 0:  31%|███       | 1019/3295 [10:35<23:39,  1.60it/s]
Epoch 0:  31%|███       | 1020/3295 [10:35<23:37,  1.60it/s]
Epoch 0:  31%|███       | 1020/3295 [10:35<23:37,  1.60it/s]
Epoch 0:  31%|███       | 1021/3295 [10:35<23:35,  1.61it/s]
Epoch 0:  31%|███       | 1021/3295 [10:35<23:35,  1.61it/s]
Epoch 0:  31%|███       | 1022/3295 [10:38<23:39,  1.60it/s]
Epoch 0:  31%|███       | 1022/3295 [10:38<23:39,  1.60it/s]
Epoch 0:  31%|███       | 1023/3295 [10:38<23:37,  1.60it/s]
Epoch 0:  31%|███       | 1023/3295 [10:38<23:37,  1.60it/s]
Epoch 0:  31%|███       | 1024/3295 [10:38<23:35,  1.60it/s]
Epoch 0:  31%|███       | 1024/3295 [10:38<23:35,  1.60it/s]
Epoch 0:  31%|███       | 1025/3295 [10:40<23:38,  1.60it/s]
Epoch 0:  31%|███       | 1025/3295 [10:40<23:38,  1.60it/s]
Epoch 0:  31%|███       | 1026/3295 [10:40<23:36,  1.60it/s]
Epoch 0:  31%|███       | 1026/3295 [10:40<23:36,  1.60it/s]
Epoch 0:  31%|███       | 1027/3295 [10:40<23:34,  1.60it/s]
Epoch 0:  31%|███       | 1027/3295 [10:40<23:34,  1.60it/s]
Epoch 0:  31%|███       | 1028/3295 [10:42<23:37,  1.60it/s]
Epoch 0:  31%|███       | 1028/3295 [10:42<23:37,  1.60it/s]
Epoch 0:  31%|███       | 1029/3295 [10:43<23:35,  1.60it/s]
Epoch 0:  31%|███       | 1029/3295 [10:43<23:35,  1.60it/s]
Epoch 0:  31%|███▏      | 1030/3295 [10:43<23:34,  1.60it/s]
Epoch 0:  31%|███▏      | 1030/3295 [10:43<23:34,  1.60it/s]
Epoch 0:  31%|███▏      | 1031/3295 [10:45<23:36,  1.60it/s]
Epoch 0:  31%|███▏      | 1031/3295 [10:45<23:36,  1.60it/s]
Epoch 0:  31%|███▏      | 1032/3295 [10:45<23:34,  1.60it/s]
Epoch 0:  31%|███▏      | 1032/3295 [10:45<23:34,  1.60it/s]
Epoch 0:  31%|███▏      | 1033/3295 [10:45<23:32,  1.60it/s]
Epoch 0:  31%|███▏      | 1033/3295 [10:45<23:32,  1.60it/s]
Epoch 0:  31%|███▏      | 1034/3295 [10:47<23:35,  1.60it/s]
Epoch 0:  31%|███▏      | 1034/3295 [10:47<23:35,  1.60it/s]
Epoch 0:  31%|███▏      | 1035/3295 [10:47<23:33,  1.60it/s]
Epoch 0:  31%|███▏      | 1035/3295 [10:47<23:33,  1.60it/s]
Epoch 0:  31%|███▏      | 1036/3295 [10:47<23:31,  1.60it/s]
Epoch 0:  31%|███▏      | 1036/3295 [10:47<23:31,  1.60it/s]
Epoch 0:  31%|███▏      | 1037/3295 [10:49<23:34,  1.60it/s]
Epoch 0:  31%|███▏      | 1037/3295 [10:49<23:34,  1.60it/s]
Epoch 0:  32%|███▏      | 1038/3295 [10:49<23:32,  1.60it/s]
Epoch 0:  32%|███▏      | 1038/3295 [10:49<23:32,  1.60it/s]
Epoch 0:  32%|███▏      | 1039/3295 [10:49<23:30,  1.60it/s]
Epoch 0:  32%|███▏      | 1039/3295 [10:49<23:30,  1.60it/s]
Epoch 0:  32%|███▏      | 1040/3295 [10:51<23:33,  1.60it/s]
Epoch 0:  32%|███▏      | 1040/3295 [10:51<23:33,  1.60it/s]
Epoch 0:  32%|███▏      | 1041/3295 [10:51<23:31,  1.60it/s]
Epoch 0:  32%|███▏      | 1041/3295 [10:51<23:31,  1.60it/s]
Epoch 0:  32%|███▏      | 1042/3295 [10:51<23:29,  1.60it/s]
Epoch 0:  32%|███▏      | 1042/3295 [10:51<23:29,  1.60it/s]
Epoch 0:  32%|███▏      | 1043/3295 [10:53<23:32,  1.59it/s]
Epoch 0:  32%|███▏      | 1043/3295 [10:53<23:32,  1.59it/s]
Epoch 0:  32%|███▏      | 1044/3295 [10:54<23:30,  1.60it/s]
Epoch 0:  32%|███▏      | 1044/3295 [10:54<23:30,  1.60it/s]
Epoch 0:  32%|███▏      | 1045/3295 [10:54<23:28,  1.60it/s]
Epoch 0:  32%|███▏      | 1045/3295 [10:54<23:28,  1.60it/s]
Epoch 0:  32%|███▏      | 1046/3295 [10:56<23:31,  1.59it/s]
Epoch 0:  32%|███▏      | 1046/3295 [10:56<23:31,  1.59it/s]
Epoch 0:  32%|███▏      | 1047/3295 [10:56<23:29,  1.60it/s]
Epoch 0:  32%|███▏      | 1047/3295 [10:56<23:29,  1.60it/s]
Epoch 0:  32%|███▏      | 1048/3295 [10:56<23:27,  1.60it/s]
Epoch 0:  32%|███▏      | 1048/3295 [10:56<23:27,  1.60it/s]
Epoch 0:  32%|███▏      | 1049/3295 [10:58<23:30,  1.59it/s]
Epoch 0:  32%|███▏      | 1049/3295 [10:58<23:30,  1.59it/s]
Epoch 0:  32%|███▏      | 1050/3295 [10:58<23:28,  1.59it/s]
Epoch 0:  32%|███▏      | 1050/3295 [10:58<23:28,  1.59it/s]
Epoch 0:  32%|███▏      | 1051/3295 [10:58<23:26,  1.60it/s]
Epoch 0:  32%|███▏      | 1051/3295 [10:58<23:26,  1.60it/s]
Epoch 0:  32%|███▏      | 1052/3295 [11:00<23:28,  1.59it/s]
Epoch 0:  32%|███▏      | 1052/3295 [11:00<23:28,  1.59it/s]
Epoch 0:  32%|███▏      | 1053/3295 [11:00<23:26,  1.59it/s]
Epoch 0:  32%|███▏      | 1053/3295 [11:00<23:26,  1.59it/s]
Epoch 0:  32%|███▏      | 1054/3295 [11:00<23:24,  1.60it/s]
Epoch 0:  32%|███▏      | 1054/3295 [11:00<23:24,  1.60it/s]
Epoch 0:  32%|███▏      | 1055/3295 [11:02<23:27,  1.59it/s]
Epoch 0:  32%|███▏      | 1055/3295 [11:02<23:27,  1.59it/s]
Epoch 0:  32%|███▏      | 1056/3295 [11:02<23:25,  1.59it/s]
Epoch 0:  32%|███▏      | 1056/3295 [11:02<23:25,  1.59it/s]
Epoch 0:  32%|███▏      | 1057/3295 [11:02<23:23,  1.59it/s]
Epoch 0:  32%|███▏      | 1057/3295 [11:02<23:23,  1.59it/s]
Epoch 0:  32%|███▏      | 1058/3295 [11:04<23:25,  1.59it/s]
Epoch 0:  32%|███▏      | 1058/3295 [11:04<23:25,  1.59it/s]
Epoch 0:  32%|███▏      | 1059/3295 [11:04<23:23,  1.59it/s]
Epoch 0:  32%|███▏      | 1059/3295 [11:04<23:23,  1.59it/s]
Epoch 0:  32%|███▏      | 1060/3295 [11:04<23:21,  1.59it/s]
Epoch 0:  32%|███▏      | 1060/3295 [11:04<23:21,  1.59it/s]
Epoch 0:  32%|███▏      | 1061/3295 [11:07<23:24,  1.59it/s]
Epoch 0:  32%|███▏      | 1061/3295 [11:07<23:24,  1.59it/s]
Epoch 0:  32%|███▏      | 1062/3295 [11:07<23:22,  1.59it/s]
Epoch 0:  32%|███▏      | 1062/3295 [11:07<23:22,  1.59it/s]
Epoch 0:  32%|███▏      | 1063/3295 [11:07<23:20,  1.59it/s]
Epoch 0:  32%|███▏      | 1063/3295 [11:07<23:20,  1.59it/s]
Epoch 0:  32%|███▏      | 1064/3295 [11:09<23:23,  1.59it/s]
Epoch 0:  32%|███▏      | 1064/3295 [11:09<23:23,  1.59it/s]
Epoch 0:  32%|███▏      | 1065/3295 [11:09<23:21,  1.59it/s]
Epoch 0:  32%|███▏      | 1065/3295 [11:09<23:21,  1.59it/s]
Epoch 0:  32%|███▏      | 1066/3295 [11:09<23:19,  1.59it/s]
Epoch 0:  32%|███▏      | 1066/3295 [11:09<23:19,  1.59it/s]
Epoch 0:  32%|███▏      | 1067/3295 [11:11<23:22,  1.59it/s]
Epoch 0:  32%|███▏      | 1067/3295 [11:11<23:22,  1.59it/s]
Epoch 0:  32%|███▏      | 1068/3295 [11:11<23:20,  1.59it/s]
Epoch 0:  32%|███▏      | 1068/3295 [11:11<23:20,  1.59it/s]
Epoch 0:  32%|███▏      | 1069/3295 [11:11<23:18,  1.59it/s]
Epoch 0:  32%|███▏      | 1069/3295 [11:11<23:18,  1.59it/s]
Epoch 0:  32%|███▏      | 1070/3295 [11:13<23:21,  1.59it/s]
Epoch 0:  32%|███▏      | 1070/3295 [11:13<23:21,  1.59it/s]
Epoch 0:  33%|███▎      | 1071/3295 [11:14<23:19,  1.59it/s]
Epoch 0:  33%|███▎      | 1071/3295 [11:14<23:19,  1.59it/s]
Epoch 0:  33%|███▎      | 1072/3295 [11:14<23:17,  1.59it/s]
Epoch 0:  33%|███▎      | 1072/3295 [11:14<23:17,  1.59it/s]
Epoch 0:  33%|███▎      | 1073/3295 [11:16<23:20,  1.59it/s]
Epoch 0:  33%|███▎      | 1073/3295 [11:16<23:20,  1.59it/s]
Epoch 0:  33%|███▎      | 1074/3295 [11:16<23:18,  1.59it/s]
Epoch 0:  33%|███▎      | 1074/3295 [11:16<23:18,  1.59it/s]
Epoch 0:  33%|███▎      | 1075/3295 [11:16<23:16,  1.59it/s]
Epoch 0:  33%|███▎      | 1075/3295 [11:16<23:16,  1.59it/s]
Epoch 0:  33%|███▎      | 1076/3295 [11:18<23:18,  1.59it/s]
Epoch 0:  33%|███▎      | 1076/3295 [11:18<23:18,  1.59it/s]
Epoch 0:  33%|███▎      | 1077/3295 [11:18<23:16,  1.59it/s]
Epoch 0:  33%|███▎      | 1077/3295 [11:18<23:16,  1.59it/s]
Epoch 0:  33%|███▎      | 1078/3295 [11:18<23:14,  1.59it/s]
Epoch 0:  33%|███▎      | 1078/3295 [11:18<23:14,  1.59it/s]
Epoch 0:  33%|███▎      | 1079/3295 [11:20<23:17,  1.59it/s]
Epoch 0:  33%|███▎      | 1079/3295 [11:20<23:17,  1.59it/s]
Epoch 0:  33%|███▎      | 1080/3295 [11:20<23:15,  1.59it/s]
Epoch 0:  33%|███▎      | 1080/3295 [11:20<23:15,  1.59it/s]
Epoch 0:  33%|███▎      | 1081/3295 [11:20<23:13,  1.59it/s]
Epoch 0:  33%|███▎      | 1081/3295 [11:20<23:13,  1.59it/s]
Epoch 0:  33%|███▎      | 1082/3295 [11:22<23:16,  1.59it/s]
Epoch 0:  33%|███▎      | 1082/3295 [11:22<23:16,  1.59it/s]
Epoch 0:  33%|███▎      | 1083/3295 [11:22<23:14,  1.59it/s]
Epoch 0:  33%|███▎      | 1083/3295 [11:22<23:14,  1.59it/s]
Epoch 0:  33%|███▎      | 1084/3295 [11:22<23:12,  1.59it/s]
Epoch 0:  33%|███▎      | 1084/3295 [11:22<23:12,  1.59it/s]
Epoch 0:  33%|███▎      | 1085/3295 [11:25<23:15,  1.58it/s]
Epoch 0:  33%|███▎      | 1085/3295 [11:25<23:15,  1.58it/s]
Epoch 0:  33%|███▎      | 1086/3295 [11:25<23:13,  1.59it/s]
Epoch 0:  33%|███▎      | 1086/3295 [11:25<23:13,  1.59it/s]
Epoch 0:  33%|███▎      | 1087/3295 [11:25<23:11,  1.59it/s]
Epoch 0:  33%|███▎      | 1087/3295 [11:25<23:11,  1.59it/s]
Epoch 0:  33%|███▎      | 1088/3295 [11:27<23:14,  1.58it/s]
Epoch 0:  33%|███▎      | 1088/3295 [11:27<23:14,  1.58it/s]
Epoch 0:  33%|███▎      | 1089/3295 [11:27<23:12,  1.58it/s]
Epoch 0:  33%|███▎      | 1089/3295 [11:27<23:12,  1.58it/s]
Epoch 0:  33%|███▎      | 1090/3295 [11:27<23:10,  1.59it/s]
Epoch 0:  33%|███▎      | 1090/3295 [11:27<23:10,  1.59it/s]
Epoch 0:  33%|███▎      | 1091/3295 [11:29<23:12,  1.58it/s]
Epoch 0:  33%|███▎      | 1091/3295 [11:29<23:12,  1.58it/s]
Epoch 0:  33%|███▎      | 1092/3295 [11:29<23:11,  1.58it/s]
Epoch 0:  33%|███▎      | 1092/3295 [11:29<23:11,  1.58it/s]
Epoch 0:  33%|███▎      | 1093/3295 [11:29<23:09,  1.58it/s]
Epoch 0:  33%|███▎      | 1093/3295 [11:29<23:09,  1.58it/s]
Epoch 0:  33%|███▎      | 1094/3295 [11:31<23:11,  1.58it/s]
Epoch 0:  33%|███▎      | 1094/3295 [11:31<23:11,  1.58it/s]
Epoch 0:  33%|███▎      | 1095/3295 [11:31<23:09,  1.58it/s]
Epoch 0:  33%|███▎      | 1095/3295 [11:31<23:09,  1.58it/s]
Epoch 0:  33%|███▎      | 1096/3295 [11:31<23:07,  1.58it/s]
Epoch 0:  33%|███▎      | 1096/3295 [11:31<23:07,  1.58it/s]
Epoch 0:  33%|███▎      | 1097/3295 [11:33<23:10,  1.58it/s]
Epoch 0:  33%|███▎      | 1097/3295 [11:33<23:10,  1.58it/s]
Epoch 0:  33%|███▎      | 1098/3295 [11:33<23:08,  1.58it/s]
Epoch 0:  33%|███▎      | 1098/3295 [11:33<23:08,  1.58it/s]
Epoch 0:  33%|███▎      | 1099/3295 [11:33<23:06,  1.58it/s]
Epoch 0:  33%|███▎      | 1099/3295 [11:33<23:06,  1.58it/s]
Epoch 0:  33%|███▎      | 1100/3295 [11:36<23:09,  1.58it/s]
Epoch 0:  33%|███▎      | 1100/3295 [11:36<23:09,  1.58it/s]
Epoch 0:  33%|███▎      | 1101/3295 [11:36<23:07,  1.58it/s]
Epoch 0:  33%|███▎      | 1101/3295 [11:36<23:07,  1.58it/s]
Epoch 0:  33%|███▎      | 1102/3295 [11:36<23:05,  1.58it/s]
Epoch 0:  33%|███▎      | 1102/3295 [11:36<23:05,  1.58it/s]
Epoch 0:  33%|███▎      | 1103/3295 [11:38<23:07,  1.58it/s]
Epoch 0:  33%|███▎      | 1103/3295 [11:38<23:07,  1.58it/s]
Epoch 0:  34%|███▎      | 1104/3295 [11:38<23:05,  1.58it/s]
Epoch 0:  34%|███▎      | 1104/3295 [11:38<23:05,  1.58it/s]
Epoch 0:  34%|███▎      | 1105/3295 [11:38<23:03,  1.58it/s]
Epoch 0:  34%|███▎      | 1105/3295 [11:38<23:03,  1.58it/s]
Epoch 0:  34%|███▎      | 1106/3295 [11:40<23:05,  1.58it/s]
Epoch 0:  34%|███▎      | 1106/3295 [11:40<23:05,  1.58it/s]
Epoch 0:  34%|███▎      | 1107/3295 [11:40<23:03,  1.58it/s]
Epoch 0:  34%|███▎      | 1107/3295 [11:40<23:03,  1.58it/s]
Epoch 0:  34%|███▎      | 1108/3295 [11:40<23:02,  1.58it/s]
Epoch 0:  34%|███▎      | 1108/3295 [11:40<23:02,  1.58it/s]
Epoch 0:  34%|███▎      | 1109/3295 [11:41<23:03,  1.58it/s]
Epoch 0:  34%|███▎      | 1109/3295 [11:41<23:03,  1.58it/s]
Epoch 0:  34%|███▎      | 1110/3295 [11:41<23:01,  1.58it/s]
Epoch 0:  34%|███▎      | 1110/3295 [11:41<23:01,  1.58it/s]
Epoch 0:  34%|███▎      | 1111/3295 [11:41<22:59,  1.58it/s]
Epoch 0:  34%|███▎      | 1111/3295 [11:41<22:59,  1.58it/s]
Epoch 0:  34%|███▎      | 1112/3295 [11:43<23:01,  1.58it/s]
Epoch 0:  34%|███▎      | 1112/3295 [11:43<23:01,  1.58it/s]
Epoch 0:  34%|███▍      | 1113/3295 [11:43<22:59,  1.58it/s]
Epoch 0:  34%|███▍      | 1113/3295 [11:43<22:59,  1.58it/s]
Epoch 0:  34%|███▍      | 1114/3295 [11:43<22:57,  1.58it/s]
Epoch 0:  34%|███▍      | 1114/3295 [11:43<22:57,  1.58it/s]
Epoch 0:  34%|███▍      | 1115/3295 [11:45<22:59,  1.58it/s]
Epoch 0:  34%|███▍      | 1115/3295 [11:45<22:59,  1.58it/s]
Epoch 0:  34%|███▍      | 1116/3295 [11:45<22:57,  1.58it/s]
Epoch 0:  34%|███▍      | 1116/3295 [11:45<22:57,  1.58it/s]
Epoch 0:  34%|███▍      | 1117/3295 [11:45<22:56,  1.58it/s]
Epoch 0:  34%|███▍      | 1117/3295 [11:45<22:56,  1.58it/s]
Epoch 0:  34%|███▍      | 1118/3295 [11:47<22:57,  1.58it/s]
Epoch 0:  34%|███▍      | 1118/3295 [11:47<22:57,  1.58it/s]
Epoch 0:  34%|███▍      | 1119/3295 [11:47<22:56,  1.58it/s]
Epoch 0:  34%|███▍      | 1119/3295 [11:47<22:56,  1.58it/s]
Epoch 0:  34%|███▍      | 1120/3295 [11:47<22:54,  1.58it/s]
Epoch 0:  34%|███▍      | 1120/3295 [11:47<22:54,  1.58it/s]
Epoch 0:  34%|███▍      | 1121/3295 [11:49<22:56,  1.58it/s]
Epoch 0:  34%|███▍      | 1121/3295 [11:49<22:56,  1.58it/s]
Epoch 0:  34%|███▍      | 1122/3295 [11:49<22:54,  1.58it/s]
Epoch 0:  34%|███▍      | 1122/3295 [11:49<22:54,  1.58it/s]
Epoch 0:  34%|███▍      | 1123/3295 [11:49<22:52,  1.58it/s]
Epoch 0:  34%|███▍      | 1123/3295 [11:49<22:52,  1.58it/s]
Epoch 0:  34%|███▍      | 1124/3295 [11:51<22:54,  1.58it/s]
Epoch 0:  34%|███▍      | 1124/3295 [11:51<22:54,  1.58it/s]
Epoch 0:  34%|███▍      | 1125/3295 [11:51<22:52,  1.58it/s]
Epoch 0:  34%|███▍      | 1125/3295 [11:51<22:52,  1.58it/s]
Epoch 0:  34%|███▍      | 1126/3295 [11:51<22:50,  1.58it/s]
Epoch 0:  34%|███▍      | 1126/3295 [11:51<22:50,  1.58it/s]
Epoch 0:  34%|███▍      | 1127/3295 [11:53<22:52,  1.58it/s]
Epoch 0:  34%|███▍      | 1127/3295 [11:53<22:52,  1.58it/s]
Epoch 0:  34%|███▍      | 1128/3295 [11:53<22:51,  1.58it/s]
Epoch 0:  34%|███▍      | 1128/3295 [11:53<22:51,  1.58it/s]
Epoch 0:  34%|███▍      | 1129/3295 [11:53<22:49,  1.58it/s]
Epoch 0:  34%|███▍      | 1129/3295 [11:53<22:49,  1.58it/s]
Epoch 0:  34%|███▍      | 1130/3295 [11:55<22:50,  1.58it/s]
Epoch 0:  34%|███▍      | 1130/3295 [11:55<22:50,  1.58it/s]
Epoch 0:  34%|███▍      | 1131/3295 [11:55<22:49,  1.58it/s]
Epoch 0:  34%|███▍      | 1131/3295 [11:55<22:49,  1.58it/s]
Epoch 0:  34%|███▍      | 1132/3295 [11:55<22:47,  1.58it/s]
Epoch 0:  34%|███▍      | 1132/3295 [11:55<22:47,  1.58it/s]
Epoch 0:  34%|███▍      | 1133/3295 [11:57<22:49,  1.58it/s]
Epoch 0:  34%|███▍      | 1133/3295 [11:57<22:49,  1.58it/s]
Epoch 0:  34%|███▍      | 1134/3295 [11:57<22:47,  1.58it/s]
Epoch 0:  34%|███▍      | 1134/3295 [11:57<22:47,  1.58it/s]
Epoch 0:  34%|███▍      | 1135/3295 [11:57<22:45,  1.58it/s]
Epoch 0:  34%|███▍      | 1135/3295 [11:57<22:45,  1.58it/s]
Epoch 0:  34%|███▍      | 1136/3295 [11:59<22:47,  1.58it/s]
Epoch 0:  34%|███▍      | 1136/3295 [11:59<22:47,  1.58it/s]
Epoch 0:  35%|███▍      | 1137/3295 [11:59<22:45,  1.58it/s]
Epoch 0:  35%|███▍      | 1137/3295 [11:59<22:45,  1.58it/s]
Epoch 0:  35%|███▍      | 1138/3295 [11:59<22:43,  1.58it/s]
Epoch 0:  35%|███▍      | 1138/3295 [11:59<22:43,  1.58it/s]
Epoch 0:  35%|███▍      | 1139/3295 [12:01<22:45,  1.58it/s]
Epoch 0:  35%|███▍      | 1139/3295 [12:01<22:45,  1.58it/s]
Epoch 0:  35%|███▍      | 1140/3295 [12:01<22:43,  1.58it/s]
Epoch 0:  35%|███▍      | 1140/3295 [12:01<22:43,  1.58it/s]
Epoch 0:  35%|███▍      | 1141/3295 [12:01<22:41,  1.58it/s]
Epoch 0:  35%|███▍      | 1141/3295 [12:01<22:41,  1.58it/s]
Epoch 0:  35%|███▍      | 1142/3295 [12:03<22:43,  1.58it/s]
Epoch 0:  35%|███▍      | 1142/3295 [12:03<22:43,  1.58it/s]
Epoch 0:  35%|███▍      | 1143/3295 [12:03<22:41,  1.58it/s]
Epoch 0:  35%|███▍      | 1143/3295 [12:03<22:41,  1.58it/s]
Epoch 0:  35%|███▍      | 1144/3295 [12:03<22:39,  1.58it/s]
Epoch 0:  35%|███▍      | 1144/3295 [12:03<22:39,  1.58it/s]
Epoch 0:  35%|███▍      | 1145/3295 [12:05<22:41,  1.58it/s]
Epoch 0:  35%|███▍      | 1145/3295 [12:05<22:41,  1.58it/s]
Epoch 0:  35%|███▍      | 1146/3295 [12:05<22:39,  1.58it/s]
Epoch 0:  35%|███▍      | 1146/3295 [12:05<22:39,  1.58it/s]
Epoch 0:  35%|███▍      | 1147/3295 [12:05<22:37,  1.58it/s]
Epoch 0:  35%|███▍      | 1147/3295 [12:05<22:37,  1.58it/s]
Epoch 0:  35%|███▍      | 1148/3295 [12:07<22:39,  1.58it/s]
Epoch 0:  35%|███▍      | 1148/3295 [12:07<22:39,  1.58it/s]
Epoch 0:  35%|███▍      | 1149/3295 [12:07<22:37,  1.58it/s]
Epoch 0:  35%|███▍      | 1149/3295 [12:07<22:37,  1.58it/s]
Epoch 0:  35%|███▍      | 1150/3295 [12:07<22:36,  1.58it/s]
Epoch 0:  35%|███▍      | 1150/3295 [12:07<22:36,  1.58it/s]
Epoch 0:  35%|███▍      | 1151/3295 [12:08<22:37,  1.58it/s]
Epoch 0:  35%|███▍      | 1151/3295 [12:08<22:37,  1.58it/s]
Epoch 0:  35%|███▍      | 1152/3295 [12:09<22:36,  1.58it/s]
Epoch 0:  35%|███▍      | 1152/3295 [12:09<22:36,  1.58it/s]
Epoch 0:  35%|███▍      | 1153/3295 [12:09<22:34,  1.58it/s]
Epoch 0:  35%|███▍      | 1153/3295 [12:09<22:34,  1.58it/s]
Epoch 0:  35%|███▌      | 1154/3295 [12:10<22:35,  1.58it/s]
Epoch 0:  35%|███▌      | 1154/3295 [12:10<22:35,  1.58it/s]
Epoch 0:  35%|███▌      | 1155/3295 [12:10<22:33,  1.58it/s]
Epoch 0:  35%|███▌      | 1155/3295 [12:10<22:33,  1.58it/s]
Epoch 0:  35%|███▌      | 1156/3295 [12:10<22:32,  1.58it/s]
Epoch 0:  35%|███▌      | 1156/3295 [12:10<22:32,  1.58it/s]
Epoch 0:  35%|███▌      | 1157/3295 [12:12<22:33,  1.58it/s]
Epoch 0:  35%|███▌      | 1157/3295 [12:12<22:33,  1.58it/s]
Epoch 0:  35%|███▌      | 1158/3295 [12:12<22:32,  1.58it/s]
Epoch 0:  35%|███▌      | 1158/3295 [12:12<22:32,  1.58it/s]
Epoch 0:  35%|███▌      | 1159/3295 [12:12<22:30,  1.58it/s]
Epoch 0:  35%|███▌      | 1159/3295 [12:12<22:30,  1.58it/s]
Epoch 0:  35%|███▌      | 1160/3295 [12:14<22:31,  1.58it/s]
Epoch 0:  35%|███▌      | 1160/3295 [12:14<22:31,  1.58it/s]
Epoch 0:  35%|███▌      | 1161/3295 [12:14<22:30,  1.58it/s]
Epoch 0:  35%|███▌      | 1161/3295 [12:14<22:30,  1.58it/s]
Epoch 0:  35%|███▌      | 1162/3295 [12:14<22:28,  1.58it/s]
Epoch 0:  35%|███▌      | 1162/3295 [12:14<22:28,  1.58it/s]
Epoch 0:  35%|███▌      | 1163/3295 [12:16<22:29,  1.58it/s]
Epoch 0:  35%|███▌      | 1163/3295 [12:16<22:29,  1.58it/s]
Epoch 0:  35%|███▌      | 1164/3295 [12:16<22:27,  1.58it/s]
Epoch 0:  35%|███▌      | 1164/3295 [12:16<22:27,  1.58it/s]
Epoch 0:  35%|███▌      | 1165/3295 [12:16<22:26,  1.58it/s]
Epoch 0:  35%|███▌      | 1165/3295 [12:16<22:26,  1.58it/s]
Epoch 0:  35%|███▌      | 1166/3295 [12:17<22:27,  1.58it/s]
Epoch 0:  35%|███▌      | 1166/3295 [12:17<22:27,  1.58it/s]
Epoch 0:  35%|███▌      | 1167/3295 [12:18<22:25,  1.58it/s]
Epoch 0:  35%|███▌      | 1167/3295 [12:18<22:25,  1.58it/s]
Epoch 0:  35%|███▌      | 1168/3295 [12:18<22:24,  1.58it/s]
Epoch 0:  35%|███▌      | 1168/3295 [12:18<22:24,  1.58it/s]
Epoch 0:  35%|███▌      | 1169/3295 [12:19<22:25,  1.58it/s]
Epoch 0:  35%|███▌      | 1169/3295 [12:19<22:25,  1.58it/s]
Epoch 0:  36%|███▌      | 1170/3295 [12:19<22:23,  1.58it/s]
Epoch 0:  36%|███▌      | 1170/3295 [12:19<22:23,  1.58it/s]
Epoch 0:  36%|███▌      | 1171/3295 [12:19<22:21,  1.58it/s]
Epoch 0:  36%|███▌      | 1171/3295 [12:19<22:21,  1.58it/s]
Epoch 0:  36%|███▌      | 1172/3295 [12:21<22:23,  1.58it/s]
Epoch 0:  36%|███▌      | 1172/3295 [12:21<22:23,  1.58it/s]
Epoch 0:  36%|███▌      | 1173/3295 [12:21<22:21,  1.58it/s]
Epoch 0:  36%|███▌      | 1173/3295 [12:21<22:21,  1.58it/s]
Epoch 0:  36%|███▌      | 1174/3295 [12:21<22:20,  1.58it/s]
Epoch 0:  36%|███▌      | 1174/3295 [12:21<22:20,  1.58it/s]
Epoch 0:  36%|███▌      | 1175/3295 [12:23<22:21,  1.58it/s]
Epoch 0:  36%|███▌      | 1175/3295 [12:23<22:21,  1.58it/s]
Epoch 0:  36%|███▌      | 1176/3295 [12:23<22:20,  1.58it/s]
Epoch 0:  36%|███▌      | 1176/3295 [12:23<22:20,  1.58it/s]
Epoch 0:  36%|███▌      | 1177/3295 [12:23<22:18,  1.58it/s]
Epoch 0:  36%|███▌      | 1177/3295 [12:23<22:18,  1.58it/s]
Epoch 0:  36%|███▌      | 1178/3295 [12:25<22:20,  1.58it/s]
Epoch 0:  36%|███▌      | 1178/3295 [12:25<22:20,  1.58it/s]
Epoch 0:  36%|███▌      | 1179/3295 [12:25<22:18,  1.58it/s]
Epoch 0:  36%|███▌      | 1179/3295 [12:25<22:18,  1.58it/s]
Epoch 0:  36%|███▌      | 1180/3295 [12:25<22:16,  1.58it/s]
Epoch 0:  36%|███▌      | 1180/3295 [12:25<22:16,  1.58it/s]
Epoch 0:  36%|███▌      | 1181/3295 [12:27<22:18,  1.58it/s]
Epoch 0:  36%|███▌      | 1181/3295 [12:27<22:18,  1.58it/s]
Epoch 0:  36%|███▌      | 1182/3295 [12:27<22:16,  1.58it/s]
Epoch 0:  36%|███▌      | 1182/3295 [12:27<22:16,  1.58it/s]
Epoch 0:  36%|███▌      | 1183/3295 [12:27<22:14,  1.58it/s]
Epoch 0:  36%|███▌      | 1183/3295 [12:27<22:14,  1.58it/s]
Epoch 0:  36%|███▌      | 1184/3295 [12:29<22:16,  1.58it/s]
Epoch 0:  36%|███▌      | 1184/3295 [12:29<22:16,  1.58it/s]
Epoch 0:  36%|███▌      | 1185/3295 [12:29<22:14,  1.58it/s]
Epoch 0:  36%|███▌      | 1185/3295 [12:29<22:14,  1.58it/s]
Epoch 0:  36%|███▌      | 1186/3295 [12:29<22:12,  1.58it/s]
Epoch 0:  36%|███▌      | 1186/3295 [12:29<22:12,  1.58it/s]
Epoch 0:  36%|███▌      | 1187/3295 [12:31<22:14,  1.58it/s]
Epoch 0:  36%|███▌      | 1187/3295 [12:31<22:14,  1.58it/s]
Epoch 0:  36%|███▌      | 1188/3295 [12:31<22:12,  1.58it/s]
Epoch 0:  36%|███▌      | 1188/3295 [12:31<22:12,  1.58it/s]
Epoch 0:  36%|███▌      | 1189/3295 [12:31<22:10,  1.58it/s]
Epoch 0:  36%|███▌      | 1189/3295 [12:31<22:10,  1.58it/s]
Epoch 0:  36%|███▌      | 1190/3295 [12:33<22:12,  1.58it/s]
Epoch 0:  36%|███▌      | 1190/3295 [12:33<22:12,  1.58it/s]
Epoch 0:  36%|███▌      | 1191/3295 [12:33<22:10,  1.58it/s]
Epoch 0:  36%|███▌      | 1191/3295 [12:33<22:10,  1.58it/s]
Epoch 0:  36%|███▌      | 1192/3295 [12:33<22:09,  1.58it/s]
Epoch 0:  36%|███▌      | 1192/3295 [12:33<22:09,  1.58it/s]
Epoch 0:  36%|███▌      | 1193/3295 [12:35<22:11,  1.58it/s]
Epoch 0:  36%|███▌      | 1193/3295 [12:35<22:11,  1.58it/s]
Epoch 0:  36%|███▌      | 1194/3295 [12:35<22:09,  1.58it/s]
Epoch 0:  36%|███▌      | 1194/3295 [12:35<22:09,  1.58it/s]
Epoch 0:  36%|███▋      | 1195/3295 [12:35<22:08,  1.58it/s]
Epoch 0:  36%|███▋      | 1195/3295 [12:35<22:08,  1.58it/s]
Epoch 0:  36%|███▋      | 1196/3295 [12:38<22:10,  1.58it/s]
Epoch 0:  36%|███▋      | 1196/3295 [12:38<22:10,  1.58it/s]
Epoch 0:  36%|███▋      | 1197/3295 [12:38<22:08,  1.58it/s]
Epoch 0:  36%|███▋      | 1197/3295 [12:38<22:08,  1.58it/s]
Epoch 0:  36%|███▋      | 1198/3295 [12:38<22:07,  1.58it/s]
Epoch 0:  36%|███▋      | 1198/3295 [12:38<22:07,  1.58it/s]
Epoch 0:  36%|███▋      | 1199/3295 [12:39<22:08,  1.58it/s]
Epoch 0:  36%|███▋      | 1199/3295 [12:39<22:08,  1.58it/s]
Epoch 0:  36%|███▋      | 1200/3295 [12:39<22:06,  1.58it/s]
Epoch 0:  36%|███▋      | 1200/3295 [12:39<22:06,  1.58it/s]
Epoch 0:  36%|███▋      | 1201/3295 [12:40<22:05,  1.58it/s]
Epoch 0:  36%|███▋      | 1201/3295 [12:40<22:05,  1.58it/s]
Epoch 0:  36%|███▋      | 1202/3295 [12:41<22:06,  1.58it/s]
Epoch 0:  36%|███▋      | 1202/3295 [12:41<22:06,  1.58it/s]
Epoch 0:  37%|███▋      | 1203/3295 [12:41<22:04,  1.58it/s]
Epoch 0:  37%|███▋      | 1203/3295 [12:41<22:04,  1.58it/s]
Epoch 0:  37%|███▋      | 1204/3295 [12:41<22:03,  1.58it/s]
Epoch 0:  37%|███▋      | 1204/3295 [12:41<22:03,  1.58it/s]
Epoch 0:  37%|███▋      | 1205/3295 [12:43<22:04,  1.58it/s]
Epoch 0:  37%|███▋      | 1205/3295 [12:43<22:04,  1.58it/s]
Epoch 0:  37%|███▋      | 1206/3295 [12:43<22:03,  1.58it/s]
Epoch 0:  37%|███▋      | 1206/3295 [12:43<22:03,  1.58it/s]
Epoch 0:  37%|███▋      | 1207/3295 [12:43<22:01,  1.58it/s]
Epoch 0:  37%|███▋      | 1207/3295 [12:43<22:01,  1.58it/s]
Epoch 0:  37%|███▋      | 1208/3295 [12:45<22:02,  1.58it/s]
Epoch 0:  37%|███▋      | 1208/3295 [12:45<22:02,  1.58it/s]
Epoch 0:  37%|███▋      | 1209/3295 [12:45<22:01,  1.58it/s]
Epoch 0:  37%|███▋      | 1209/3295 [12:45<22:01,  1.58it/s]
Epoch 0:  37%|███▋      | 1210/3295 [12:45<21:59,  1.58it/s]
Epoch 0:  37%|███▋      | 1210/3295 [12:45<21:59,  1.58it/s]
Epoch 0:  37%|███▋      | 1211/3295 [12:47<22:00,  1.58it/s]
Epoch 0:  37%|███▋      | 1211/3295 [12:47<22:00,  1.58it/s]
Epoch 0:  37%|███▋      | 1212/3295 [12:47<21:59,  1.58it/s]
Epoch 0:  37%|███▋      | 1212/3295 [12:47<21:59,  1.58it/s]
Epoch 0:  37%|███▋      | 1213/3295 [12:47<21:57,  1.58it/s]
Epoch 0:  37%|███▋      | 1213/3295 [12:47<21:57,  1.58it/s]
Epoch 0:  37%|███▋      | 1214/3295 [12:49<21:58,  1.58it/s]
Epoch 0:  37%|███▋      | 1214/3295 [12:49<21:58,  1.58it/s]
Epoch 0:  37%|███▋      | 1215/3295 [12:49<21:57,  1.58it/s]
Epoch 0:  37%|███▋      | 1215/3295 [12:49<21:57,  1.58it/s]
Epoch 0:  37%|███▋      | 1216/3295 [12:49<21:55,  1.58it/s]
Epoch 0:  37%|███▋      | 1216/3295 [12:49<21:55,  1.58it/s]
Epoch 0:  37%|███▋      | 1217/3295 [12:51<21:57,  1.58it/s]
Epoch 0:  37%|███▋      | 1217/3295 [12:51<21:57,  1.58it/s]
Epoch 0:  37%|███▋      | 1218/3295 [12:51<21:55,  1.58it/s]
Epoch 0:  37%|███▋      | 1218/3295 [12:51<21:55,  1.58it/s]
Epoch 0:  37%|███▋      | 1219/3295 [12:51<21:53,  1.58it/s]
Epoch 0:  37%|███▋      | 1219/3295 [12:51<21:53,  1.58it/s]
Epoch 0:  37%|███▋      | 1220/3295 [12:53<21:54,  1.58it/s]
Epoch 0:  37%|███▋      | 1220/3295 [12:53<21:54,  1.58it/s]
Epoch 0:  37%|███▋      | 1221/3295 [12:53<21:53,  1.58it/s]
Epoch 0:  37%|███▋      | 1221/3295 [12:53<21:53,  1.58it/s]
Epoch 0:  37%|███▋      | 1222/3295 [12:53<21:51,  1.58it/s]
Epoch 0:  37%|███▋      | 1222/3295 [12:53<21:51,  1.58it/s]
Epoch 0:  37%|███▋      | 1223/3295 [12:55<21:53,  1.58it/s]
Epoch 0:  37%|███▋      | 1223/3295 [12:55<21:53,  1.58it/s]
Epoch 0:  37%|███▋      | 1224/3295 [12:55<21:51,  1.58it/s]
Epoch 0:  37%|███▋      | 1224/3295 [12:55<21:51,  1.58it/s]
Epoch 0:  37%|███▋      | 1225/3295 [12:55<21:49,  1.58it/s]
Epoch 0:  37%|███▋      | 1225/3295 [12:55<21:49,  1.58it/s]
Epoch 0:  37%|███▋      | 1226/3295 [12:56<21:51,  1.58it/s]
Epoch 0:  37%|███▋      | 1226/3295 [12:56<21:51,  1.58it/s]
Epoch 0:  37%|███▋      | 1227/3295 [12:56<21:49,  1.58it/s]
Epoch 0:  37%|███▋      | 1227/3295 [12:56<21:49,  1.58it/s]
Epoch 0:  37%|███▋      | 1228/3295 [12:56<21:47,  1.58it/s]
Epoch 0:  37%|███▋      | 1228/3295 [12:56<21:47,  1.58it/s]
Epoch 0:  37%|███▋      | 1229/3295 [12:58<21:49,  1.58it/s]
Epoch 0:  37%|███▋      | 1229/3295 [12:58<21:49,  1.58it/s]
Epoch 0:  37%|███▋      | 1230/3295 [12:58<21:47,  1.58it/s]
Epoch 0:  37%|███▋      | 1230/3295 [12:58<21:47,  1.58it/s]
Epoch 0:  37%|███▋      | 1231/3295 [12:58<21:45,  1.58it/s]
Epoch 0:  37%|███▋      | 1231/3295 [12:58<21:45,  1.58it/s]
Epoch 0:  37%|███▋      | 1232/3295 [13:00<21:47,  1.58it/s]
Epoch 0:  37%|███▋      | 1232/3295 [13:00<21:47,  1.58it/s]
Epoch 0:  37%|███▋      | 1233/3295 [13:00<21:45,  1.58it/s]
Epoch 0:  37%|███▋      | 1233/3295 [13:00<21:45,  1.58it/s]
Epoch 0:  37%|███▋      | 1234/3295 [13:00<21:43,  1.58it/s]
Epoch 0:  37%|███▋      | 1234/3295 [13:00<21:43,  1.58it/s]
Epoch 0:  37%|███▋      | 1235/3295 [13:02<21:44,  1.58it/s]
Epoch 0:  37%|███▋      | 1235/3295 [13:02<21:44,  1.58it/s]
Epoch 0:  38%|███▊      | 1236/3295 [13:02<21:43,  1.58it/s]
Epoch 0:  38%|███▊      | 1236/3295 [13:02<21:43,  1.58it/s]
Epoch 0:  38%|███▊      | 1237/3295 [13:02<21:41,  1.58it/s]
Epoch 0:  38%|███▊      | 1237/3295 [13:02<21:41,  1.58it/s]
Epoch 0:  38%|███▊      | 1238/3295 [13:04<21:42,  1.58it/s]
Epoch 0:  38%|███▊      | 1238/3295 [13:04<21:42,  1.58it/s]
Epoch 0:  38%|███▊      | 1239/3295 [13:04<21:41,  1.58it/s]
Epoch 0:  38%|███▊      | 1239/3295 [13:04<21:41,  1.58it/s]
Epoch 0:  38%|███▊      | 1240/3295 [13:04<21:39,  1.58it/s]
Epoch 0:  38%|███▊      | 1240/3295 [13:04<21:39,  1.58it/s]
Epoch 0:  38%|███▊      | 1241/3295 [13:06<21:40,  1.58it/s]
Epoch 0:  38%|███▊      | 1241/3295 [13:06<21:40,  1.58it/s]
Epoch 0:  38%|███▊      | 1242/3295 [13:06<21:39,  1.58it/s]
Epoch 0:  38%|███▊      | 1242/3295 [13:06<21:39,  1.58it/s]
Epoch 0:  38%|███▊      | 1243/3295 [13:06<21:37,  1.58it/s]
Epoch 0:  38%|███▊      | 1243/3295 [13:06<21:37,  1.58it/s]
Epoch 0:  38%|███▊      | 1244/3295 [13:07<21:38,  1.58it/s]
Epoch 0:  38%|███▊      | 1244/3295 [13:07<21:38,  1.58it/s]
Epoch 0:  38%|███▊      | 1245/3295 [13:07<21:37,  1.58it/s]
Epoch 0:  38%|███▊      | 1245/3295 [13:07<21:37,  1.58it/s]
Epoch 0:  38%|███▊      | 1246/3295 [13:07<21:35,  1.58it/s]
Epoch 0:  38%|███▊      | 1246/3295 [13:07<21:35,  1.58it/s]
Epoch 0:  38%|███▊      | 1247/3295 [13:09<21:36,  1.58it/s]
Epoch 0:  38%|███▊      | 1247/3295 [13:09<21:36,  1.58it/s]
Epoch 0:  38%|███▊      | 1248/3295 [13:09<21:35,  1.58it/s]
Epoch 0:  38%|███▊      | 1248/3295 [13:09<21:35,  1.58it/s]
Epoch 0:  38%|███▊      | 1249/3295 [13:09<21:33,  1.58it/s]
Epoch 0:  38%|███▊      | 1249/3295 [13:09<21:33,  1.58it/s]
Epoch 0:  38%|███▊      | 1250/3295 [13:11<21:34,  1.58it/s]
Epoch 0:  38%|███▊      | 1250/3295 [13:11<21:34,  1.58it/s]
Epoch 0:  38%|███▊      | 1251/3295 [13:11<21:33,  1.58it/s]
Epoch 0:  38%|███▊      | 1251/3295 [13:11<21:33,  1.58it/s]
Epoch 0:  38%|███▊      | 1252/3295 [13:11<21:31,  1.58it/s]
Epoch 0:  38%|███▊      | 1252/3295 [13:11<21:31,  1.58it/s]
Epoch 0:  38%|███▊      | 1253/3295 [13:13<21:32,  1.58it/s]
Epoch 0:  38%|███▊      | 1253/3295 [13:13<21:32,  1.58it/s]
Epoch 0:  38%|███▊      | 1254/3295 [13:13<21:31,  1.58it/s]
Epoch 0:  38%|███▊      | 1254/3295 [13:13<21:31,  1.58it/s]
Epoch 0:  38%|███▊      | 1255/3295 [13:13<21:29,  1.58it/s]
Epoch 0:  38%|███▊      | 1255/3295 [13:13<21:29,  1.58it/s]
Epoch 0:  38%|███▊      | 1256/3295 [13:15<21:30,  1.58it/s]
Epoch 0:  38%|███▊      | 1256/3295 [13:15<21:30,  1.58it/s]
Epoch 0:  38%|███▊      | 1257/3295 [13:15<21:29,  1.58it/s]
Epoch 0:  38%|███▊      | 1257/3295 [13:15<21:29,  1.58it/s]
Epoch 0:  38%|███▊      | 1258/3295 [13:15<21:27,  1.58it/s]
Epoch 0:  38%|███▊      | 1258/3295 [13:15<21:27,  1.58it/s]
Epoch 0:  38%|███▊      | 1259/3295 [13:17<21:29,  1.58it/s]
Epoch 0:  38%|███▊      | 1259/3295 [13:17<21:29,  1.58it/s]
Epoch 0:  38%|███▊      | 1260/3295 [13:17<21:27,  1.58it/s]
Epoch 0:  38%|███▊      | 1260/3295 [13:17<21:27,  1.58it/s]
Epoch 0:  38%|███▊      | 1261/3295 [13:17<21:25,  1.58it/s]
Epoch 0:  38%|███▊      | 1261/3295 [13:17<21:25,  1.58it/s]
Epoch 0:  38%|███▊      | 1262/3295 [13:18<21:26,  1.58it/s]
Epoch 0:  38%|███▊      | 1262/3295 [13:18<21:26,  1.58it/s]
Epoch 0:  38%|███▊      | 1263/3295 [13:18<21:25,  1.58it/s]
Epoch 0:  38%|███▊      | 1263/3295 [13:18<21:25,  1.58it/s]
Epoch 0:  38%|███▊      | 1264/3295 [13:18<21:23,  1.58it/s]
Epoch 0:  38%|███▊      | 1264/3295 [13:18<21:23,  1.58it/s]
Epoch 0:  38%|███▊      | 1265/3295 [13:20<21:25,  1.58it/s]
Epoch 0:  38%|███▊      | 1265/3295 [13:20<21:25,  1.58it/s]
Epoch 0:  38%|███▊      | 1266/3295 [13:20<21:23,  1.58it/s]
Epoch 0:  38%|███▊      | 1266/3295 [13:20<21:23,  1.58it/s]
Epoch 0:  38%|███▊      | 1267/3295 [13:20<21:21,  1.58it/s]
Epoch 0:  38%|███▊      | 1267/3295 [13:20<21:21,  1.58it/s]
Epoch 0:  38%|███▊      | 1268/3295 [13:22<21:23,  1.58it/s]
Epoch 0:  38%|███▊      | 1268/3295 [13:22<21:23,  1.58it/s]
Epoch 0:  39%|███▊      | 1269/3295 [13:22<21:21,  1.58it/s]
Epoch 0:  39%|███▊      | 1269/3295 [13:22<21:21,  1.58it/s]
Epoch 0:  39%|███▊      | 1270/3295 [13:22<21:19,  1.58it/s]
Epoch 0:  39%|███▊      | 1270/3295 [13:22<21:19,  1.58it/s]
Epoch 0:  39%|███▊      | 1271/3295 [13:24<21:21,  1.58it/s]
Epoch 0:  39%|███▊      | 1271/3295 [13:24<21:21,  1.58it/s]
Epoch 0:  39%|███▊      | 1272/3295 [13:24<21:19,  1.58it/s]
Epoch 0:  39%|███▊      | 1272/3295 [13:24<21:19,  1.58it/s]
Epoch 0:  39%|███▊      | 1273/3295 [13:24<21:18,  1.58it/s]
Epoch 0:  39%|███▊      | 1273/3295 [13:24<21:18,  1.58it/s]
Epoch 0:  39%|███▊      | 1274/3295 [13:26<21:19,  1.58it/s]
Epoch 0:  39%|███▊      | 1274/3295 [13:26<21:19,  1.58it/s]
Epoch 0:  39%|███▊      | 1275/3295 [13:26<21:17,  1.58it/s]
Epoch 0:  39%|███▊      | 1275/3295 [13:26<21:17,  1.58it/s]
Epoch 0:  39%|███▊      | 1276/3295 [13:26<21:16,  1.58it/s]
Epoch 0:  39%|███▊      | 1276/3295 [13:26<21:16,  1.58it/s]
Epoch 0:  39%|███▉      | 1277/3295 [13:28<21:17,  1.58it/s]
Epoch 0:  39%|███▉      | 1277/3295 [13:28<21:17,  1.58it/s]
Epoch 0:  39%|███▉      | 1278/3295 [13:28<21:16,  1.58it/s]
Epoch 0:  39%|███▉      | 1278/3295 [13:28<21:16,  1.58it/s]
Epoch 0:  39%|███▉      | 1279/3295 [13:28<21:14,  1.58it/s]
Epoch 0:  39%|███▉      | 1279/3295 [13:28<21:14,  1.58it/s]
Epoch 0:  39%|███▉      | 1280/3295 [13:30<21:15,  1.58it/s]
Epoch 0:  39%|███▉      | 1280/3295 [13:30<21:15,  1.58it/s]
Epoch 0:  39%|███▉      | 1281/3295 [13:30<21:14,  1.58it/s]
Epoch 0:  39%|███▉      | 1281/3295 [13:30<21:14,  1.58it/s]
Epoch 0:  39%|███▉      | 1282/3295 [13:30<21:12,  1.58it/s]
Epoch 0:  39%|███▉      | 1282/3295 [13:30<21:12,  1.58it/s]
Epoch 0:  39%|███▉      | 1283/3295 [13:32<21:13,  1.58it/s]
Epoch 0:  39%|███▉      | 1283/3295 [13:32<21:13,  1.58it/s]
Epoch 0:  39%|███▉      | 1284/3295 [13:32<21:12,  1.58it/s]
Epoch 0:  39%|███▉      | 1284/3295 [13:32<21:12,  1.58it/s]
Epoch 0:  39%|███▉      | 1285/3295 [13:32<21:10,  1.58it/s]
Epoch 0:  39%|███▉      | 1285/3295 [13:32<21:10,  1.58it/s]
Epoch 0:  39%|███▉      | 1286/3295 [13:34<21:12,  1.58it/s]
Epoch 0:  39%|███▉      | 1286/3295 [13:34<21:12,  1.58it/s]
Epoch 0:  39%|███▉      | 1287/3295 [13:34<21:11,  1.58it/s]
Epoch 0:  39%|███▉      | 1287/3295 [13:34<21:11,  1.58it/s]
Epoch 0:  39%|███▉      | 1288/3295 [13:34<21:09,  1.58it/s]
Epoch 0:  39%|███▉      | 1288/3295 [13:34<21:09,  1.58it/s]
Epoch 0:  39%|███▉      | 1289/3295 [13:37<21:11,  1.58it/s]
Epoch 0:  39%|███▉      | 1289/3295 [13:37<21:11,  1.58it/s]
Epoch 0:  39%|███▉      | 1290/3295 [13:37<21:10,  1.58it/s]
Epoch 0:  39%|███▉      | 1290/3295 [13:37<21:10,  1.58it/s]
Epoch 0:  39%|███▉      | 1291/3295 [13:37<21:08,  1.58it/s]
Epoch 0:  39%|███▉      | 1291/3295 [13:37<21:08,  1.58it/s]
Epoch 0:  39%|███▉      | 1292/3295 [13:39<21:10,  1.58it/s]
Epoch 0:  39%|███▉      | 1292/3295 [13:39<21:10,  1.58it/s]
Epoch 0:  39%|███▉      | 1293/3295 [13:39<21:08,  1.58it/s]
Epoch 0:  39%|███▉      | 1293/3295 [13:39<21:08,  1.58it/s]
Epoch 0:  39%|███▉      | 1294/3295 [13:39<21:07,  1.58it/s]
Epoch 0:  39%|███▉      | 1294/3295 [13:39<21:07,  1.58it/s]
Epoch 0:  39%|███▉      | 1295/3295 [13:41<21:08,  1.58it/s]
Epoch 0:  39%|███▉      | 1295/3295 [13:41<21:08,  1.58it/s]
Epoch 0:  39%|███▉      | 1296/3295 [13:41<21:06,  1.58it/s]
Epoch 0:  39%|███▉      | 1296/3295 [13:41<21:06,  1.58it/s]
Epoch 0:  39%|███▉      | 1297/3295 [13:41<21:05,  1.58it/s]
Epoch 0:  39%|███▉      | 1297/3295 [13:41<21:05,  1.58it/s]
Epoch 0:  39%|███▉      | 1298/3295 [13:43<21:06,  1.58it/s]
Epoch 0:  39%|███▉      | 1298/3295 [13:43<21:06,  1.58it/s]
Epoch 0:  39%|███▉      | 1299/3295 [13:43<21:04,  1.58it/s]
Epoch 0:  39%|███▉      | 1299/3295 [13:43<21:04,  1.58it/s]
Epoch 0:  39%|███▉      | 1300/3295 [13:43<21:03,  1.58it/s]
Epoch 0:  39%|███▉      | 1300/3295 [13:43<21:03,  1.58it/s]
Epoch 0:  39%|███▉      | 1301/3295 [13:44<21:04,  1.58it/s]
Epoch 0:  39%|███▉      | 1301/3295 [13:44<21:04,  1.58it/s]
Epoch 0:  40%|███▉      | 1302/3295 [13:44<21:02,  1.58it/s]
Epoch 0:  40%|███▉      | 1302/3295 [13:44<21:02,  1.58it/s]
Epoch 0:  40%|███▉      | 1303/3295 [13:44<21:01,  1.58it/s]
Epoch 0:  40%|███▉      | 1303/3295 [13:44<21:01,  1.58it/s]
Epoch 0:  40%|███▉      | 1304/3295 [13:46<21:02,  1.58it/s]
Epoch 0:  40%|███▉      | 1304/3295 [13:46<21:02,  1.58it/s]
Epoch 0:  40%|███▉      | 1305/3295 [13:46<21:00,  1.58it/s]
Epoch 0:  40%|███▉      | 1305/3295 [13:46<21:00,  1.58it/s]
Epoch 0:  40%|███▉      | 1306/3295 [13:46<20:59,  1.58it/s]
Epoch 0:  40%|███▉      | 1306/3295 [13:46<20:59,  1.58it/s]
Epoch 0:  40%|███▉      | 1307/3295 [13:48<21:00,  1.58it/s]
Epoch 0:  40%|███▉      | 1307/3295 [13:48<21:00,  1.58it/s]
Epoch 0:  40%|███▉      | 1308/3295 [13:48<20:58,  1.58it/s]
Epoch 0:  40%|███▉      | 1308/3295 [13:48<20:58,  1.58it/s]
Epoch 0:  40%|███▉      | 1309/3295 [13:48<20:57,  1.58it/s]
Epoch 0:  40%|███▉      | 1309/3295 [13:48<20:57,  1.58it/s]
Epoch 0:  40%|███▉      | 1310/3295 [13:50<20:58,  1.58it/s]
Epoch 0:  40%|███▉      | 1310/3295 [13:50<20:58,  1.58it/s]
Epoch 0:  40%|███▉      | 1311/3295 [13:50<20:56,  1.58it/s]
Epoch 0:  40%|███▉      | 1311/3295 [13:50<20:56,  1.58it/s]
Epoch 0:  40%|███▉      | 1312/3295 [13:50<20:55,  1.58it/s]
Epoch 0:  40%|███▉      | 1312/3295 [13:50<20:55,  1.58it/s]
Epoch 0:  40%|███▉      | 1313/3295 [13:52<20:56,  1.58it/s]
Epoch 0:  40%|███▉      | 1313/3295 [13:52<20:56,  1.58it/s]
Epoch 0:  40%|███▉      | 1314/3295 [13:52<20:55,  1.58it/s]
Epoch 0:  40%|███▉      | 1314/3295 [13:52<20:55,  1.58it/s]
Epoch 0:  40%|███▉      | 1315/3295 [13:52<20:53,  1.58it/s]
Epoch 0:  40%|███▉      | 1315/3295 [13:52<20:53,  1.58it/s]
Epoch 0:  40%|███▉      | 1316/3295 [13:54<20:55,  1.58it/s]
Epoch 0:  40%|███▉      | 1316/3295 [13:54<20:55,  1.58it/s]
Epoch 0:  40%|███▉      | 1317/3295 [13:54<20:53,  1.58it/s]
Epoch 0:  40%|███▉      | 1317/3295 [13:54<20:53,  1.58it/s]
Epoch 0:  40%|████      | 1318/3295 [13:54<20:51,  1.58it/s]
Epoch 0:  40%|████      | 1318/3295 [13:54<20:51,  1.58it/s]
Epoch 0:  40%|████      | 1319/3295 [13:56<20:52,  1.58it/s]
Epoch 0:  40%|████      | 1319/3295 [13:56<20:52,  1.58it/s]
Epoch 0:  40%|████      | 1320/3295 [13:56<20:51,  1.58it/s]
Epoch 0:  40%|████      | 1320/3295 [13:56<20:51,  1.58it/s]
Epoch 0:  40%|████      | 1321/3295 [13:56<20:49,  1.58it/s]
Epoch 0:  40%|████      | 1321/3295 [13:56<20:49,  1.58it/s]
Epoch 0:  40%|████      | 1322/3295 [13:58<20:51,  1.58it/s]
Epoch 0:  40%|████      | 1322/3295 [13:58<20:51,  1.58it/s]
Epoch 0:  40%|████      | 1323/3295 [13:58<20:49,  1.58it/s]
Epoch 0:  40%|████      | 1323/3295 [13:58<20:49,  1.58it/s]
Epoch 0:  40%|████      | 1324/3295 [13:58<20:48,  1.58it/s]
Epoch 0:  40%|████      | 1324/3295 [13:58<20:48,  1.58it/s]
Epoch 0:  40%|████      | 1325/3295 [14:00<20:49,  1.58it/s]
Epoch 0:  40%|████      | 1325/3295 [14:00<20:49,  1.58it/s]
Epoch 0:  40%|████      | 1326/3295 [14:00<20:47,  1.58it/s]
Epoch 0:  40%|████      | 1326/3295 [14:00<20:47,  1.58it/s]
Epoch 0:  40%|████      | 1327/3295 [14:00<20:46,  1.58it/s]
Epoch 0:  40%|████      | 1327/3295 [14:00<20:46,  1.58it/s]
Epoch 0:  40%|████      | 1328/3295 [14:01<20:47,  1.58it/s]
Epoch 0:  40%|████      | 1328/3295 [14:01<20:47,  1.58it/s]
Epoch 0:  40%|████      | 1329/3295 [14:01<20:45,  1.58it/s]
Epoch 0:  40%|████      | 1329/3295 [14:01<20:45,  1.58it/s]
Epoch 0:  40%|████      | 1330/3295 [14:02<20:44,  1.58it/s]
Epoch 0:  40%|████      | 1330/3295 [14:02<20:44,  1.58it/s]
Epoch 0:  40%|████      | 1331/3295 [14:03<20:45,  1.58it/s]
Epoch 0:  40%|████      | 1331/3295 [14:03<20:45,  1.58it/s]
Epoch 0:  40%|████      | 1332/3295 [14:03<20:43,  1.58it/s]
Epoch 0:  40%|████      | 1332/3295 [14:03<20:43,  1.58it/s]
Epoch 0:  40%|████      | 1333/3295 [14:03<20:42,  1.58it/s]
Epoch 0:  40%|████      | 1333/3295 [14:03<20:42,  1.58it/s]
Epoch 0:  40%|████      | 1334/3295 [14:05<20:43,  1.58it/s]
Epoch 0:  40%|████      | 1334/3295 [14:05<20:43,  1.58it/s]
Epoch 0:  41%|████      | 1335/3295 [14:05<20:41,  1.58it/s]
Epoch 0:  41%|████      | 1335/3295 [14:05<20:41,  1.58it/s]
Epoch 0:  41%|████      | 1336/3295 [14:05<20:40,  1.58it/s]
Epoch 0:  41%|████      | 1336/3295 [14:05<20:40,  1.58it/s]
Epoch 0:  41%|████      | 1337/3295 [14:07<20:41,  1.58it/s]
Epoch 0:  41%|████      | 1337/3295 [14:07<20:41,  1.58it/s]
Epoch 0:  41%|████      | 1338/3295 [14:07<20:39,  1.58it/s]
Epoch 0:  41%|████      | 1338/3295 [14:07<20:39,  1.58it/s]
Epoch 0:  41%|████      | 1339/3295 [14:07<20:38,  1.58it/s]
Epoch 0:  41%|████      | 1339/3295 [14:07<20:38,  1.58it/s]
Epoch 0:  41%|████      | 1340/3295 [14:09<20:39,  1.58it/s]
Epoch 0:  41%|████      | 1340/3295 [14:09<20:39,  1.58it/s]
Epoch 0:  41%|████      | 1341/3295 [14:09<20:37,  1.58it/s]
Epoch 0:  41%|████      | 1341/3295 [14:09<20:37,  1.58it/s]
Epoch 0:  41%|████      | 1342/3295 [14:09<20:36,  1.58it/s]
Epoch 0:  41%|████      | 1342/3295 [14:09<20:36,  1.58it/s]
Epoch 0:  41%|████      | 1343/3295 [14:11<20:37,  1.58it/s]
Epoch 0:  41%|████      | 1343/3295 [14:11<20:37,  1.58it/s]
Epoch 0:  41%|████      | 1344/3295 [14:11<20:35,  1.58it/s]
Epoch 0:  41%|████      | 1344/3295 [14:11<20:35,  1.58it/s]
Epoch 0:  41%|████      | 1345/3295 [14:11<20:34,  1.58it/s]
Epoch 0:  41%|████      | 1345/3295 [14:11<20:34,  1.58it/s]
Epoch 0:  41%|████      | 1346/3295 [14:13<20:35,  1.58it/s]
Epoch 0:  41%|████      | 1346/3295 [14:13<20:35,  1.58it/s]
Epoch 0:  41%|████      | 1347/3295 [14:13<20:33,  1.58it/s]
Epoch 0:  41%|████      | 1347/3295 [14:13<20:33,  1.58it/s]
Epoch 0:  41%|████      | 1348/3295 [14:13<20:32,  1.58it/s]
Epoch 0:  41%|████      | 1348/3295 [14:13<20:32,  1.58it/s]
Epoch 0:  41%|████      | 1349/3295 [14:15<20:33,  1.58it/s]
Epoch 0:  41%|████      | 1349/3295 [14:15<20:33,  1.58it/s]
Epoch 0:  41%|████      | 1350/3295 [14:15<20:31,  1.58it/s]
Epoch 0:  41%|████      | 1350/3295 [14:15<20:31,  1.58it/s]
Epoch 0:  41%|████      | 1351/3295 [14:15<20:30,  1.58it/s]
Epoch 0:  41%|████      | 1351/3295 [14:15<20:30,  1.58it/s]
Epoch 0:  41%|████      | 1352/3295 [14:16<20:31,  1.58it/s]
Epoch 0:  41%|████      | 1352/3295 [14:16<20:31,  1.58it/s]
Epoch 0:  41%|████      | 1353/3295 [14:16<20:29,  1.58it/s]
Epoch 0:  41%|████      | 1353/3295 [14:16<20:29,  1.58it/s]
Epoch 0:  41%|████      | 1354/3295 [14:16<20:28,  1.58it/s]
Epoch 0:  41%|████      | 1354/3295 [14:16<20:28,  1.58it/s]
Epoch 0:  41%|████      | 1355/3295 [14:18<20:29,  1.58it/s]
Epoch 0:  41%|████      | 1355/3295 [14:18<20:29,  1.58it/s]
Epoch 0:  41%|████      | 1356/3295 [14:18<20:27,  1.58it/s]
Epoch 0:  41%|████      | 1356/3295 [14:18<20:27,  1.58it/s]
Epoch 0:  41%|████      | 1357/3295 [14:18<20:26,  1.58it/s]
Epoch 0:  41%|████      | 1357/3295 [14:18<20:26,  1.58it/s]
Epoch 0:  41%|████      | 1358/3295 [14:20<20:27,  1.58it/s]
Epoch 0:  41%|████      | 1358/3295 [14:20<20:27,  1.58it/s]
Epoch 0:  41%|████      | 1359/3295 [14:20<20:25,  1.58it/s]
Epoch 0:  41%|████      | 1359/3295 [14:20<20:25,  1.58it/s]
Epoch 0:  41%|████▏     | 1360/3295 [14:20<20:24,  1.58it/s]
Epoch 0:  41%|████▏     | 1360/3295 [14:20<20:24,  1.58it/s]
Epoch 0:  41%|████▏     | 1361/3295 [14:22<20:25,  1.58it/s]
Epoch 0:  41%|████▏     | 1361/3295 [14:22<20:25,  1.58it/s]
Epoch 0:  41%|████▏     | 1362/3295 [14:22<20:24,  1.58it/s]
Epoch 0:  41%|████▏     | 1362/3295 [14:22<20:24,  1.58it/s]
Epoch 0:  41%|████▏     | 1363/3295 [14:22<20:22,  1.58it/s]
Epoch 0:  41%|████▏     | 1363/3295 [14:22<20:22,  1.58it/s]
Epoch 0:  41%|████▏     | 1364/3295 [14:24<20:23,  1.58it/s]
Epoch 0:  41%|████▏     | 1364/3295 [14:24<20:23,  1.58it/s]
Epoch 0:  41%|████▏     | 1365/3295 [14:24<20:22,  1.58it/s]
Epoch 0:  41%|████▏     | 1365/3295 [14:24<20:22,  1.58it/s]
Epoch 0:  41%|████▏     | 1366/3295 [14:24<20:20,  1.58it/s]
Epoch 0:  41%|████▏     | 1366/3295 [14:24<20:20,  1.58it/s]
Epoch 0:  41%|████▏     | 1367/3295 [14:26<20:21,  1.58it/s]
Epoch 0:  41%|████▏     | 1367/3295 [14:26<20:21,  1.58it/s]
Epoch 0:  42%|████▏     | 1368/3295 [14:26<20:20,  1.58it/s]
Epoch 0:  42%|████▏     | 1368/3295 [14:26<20:20,  1.58it/s]
Epoch 0:  42%|████▏     | 1369/3295 [14:26<20:18,  1.58it/s]
Epoch 0:  42%|████▏     | 1369/3295 [14:26<20:18,  1.58it/s]
Epoch 0:  42%|████▏     | 1370/3295 [14:28<20:20,  1.58it/s]
Epoch 0:  42%|████▏     | 1370/3295 [14:28<20:20,  1.58it/s]
Epoch 0:  42%|████▏     | 1371/3295 [14:28<20:18,  1.58it/s]
Epoch 0:  42%|████▏     | 1371/3295 [14:28<20:18,  1.58it/s]
Epoch 0:  42%|████▏     | 1372/3295 [14:28<20:17,  1.58it/s]
Epoch 0:  42%|████▏     | 1372/3295 [14:28<20:17,  1.58it/s]
Epoch 0:  42%|████▏     | 1373/3295 [14:30<20:18,  1.58it/s]
Epoch 0:  42%|████▏     | 1373/3295 [14:30<20:18,  1.58it/s]
Epoch 0:  42%|████▏     | 1374/3295 [14:30<20:16,  1.58it/s]
Epoch 0:  42%|████▏     | 1374/3295 [14:30<20:16,  1.58it/s]
Epoch 0:  42%|████▏     | 1375/3295 [14:30<20:15,  1.58it/s]
Epoch 0:  42%|████▏     | 1375/3295 [14:30<20:15,  1.58it/s]
Epoch 0:  42%|████▏     | 1376/3295 [14:32<20:16,  1.58it/s]
Epoch 0:  42%|████▏     | 1376/3295 [14:32<20:16,  1.58it/s]
Epoch 0:  42%|████▏     | 1377/3295 [14:32<20:14,  1.58it/s]
Epoch 0:  42%|████▏     | 1377/3295 [14:32<20:14,  1.58it/s]
Epoch 0:  42%|████▏     | 1378/3295 [14:32<20:13,  1.58it/s]
Epoch 0:  42%|████▏     | 1378/3295 [14:32<20:13,  1.58it/s]
Epoch 0:  42%|████▏     | 1379/3295 [14:34<20:14,  1.58it/s]
Epoch 0:  42%|████▏     | 1379/3295 [14:34<20:14,  1.58it/s]
Epoch 0:  42%|████▏     | 1380/3295 [14:34<20:12,  1.58it/s]
Epoch 0:  42%|████▏     | 1380/3295 [14:34<20:12,  1.58it/s]
Epoch 0:  42%|████▏     | 1381/3295 [14:34<20:11,  1.58it/s]
Epoch 0:  42%|████▏     | 1381/3295 [14:34<20:11,  1.58it/s]
Epoch 0:  42%|████▏     | 1382/3295 [14:35<20:12,  1.58it/s]
Epoch 0:  42%|████▏     | 1382/3295 [14:35<20:12,  1.58it/s]
Epoch 0:  42%|████▏     | 1383/3295 [14:36<20:11,  1.58it/s]
Epoch 0:  42%|████▏     | 1383/3295 [14:36<20:11,  1.58it/s]
Epoch 0:  42%|████▏     | 1384/3295 [14:36<20:09,  1.58it/s]
Epoch 0:  42%|████▏     | 1384/3295 [14:36<20:09,  1.58it/s]
Epoch 0:  42%|████▏     | 1385/3295 [14:37<20:10,  1.58it/s]
Epoch 0:  42%|████▏     | 1385/3295 [14:37<20:10,  1.58it/s]
Epoch 0:  42%|████▏     | 1386/3295 [14:37<20:09,  1.58it/s]
Epoch 0:  42%|████▏     | 1386/3295 [14:37<20:09,  1.58it/s]
Epoch 0:  42%|████▏     | 1387/3295 [14:37<20:07,  1.58it/s]
Epoch 0:  42%|████▏     | 1387/3295 [14:37<20:07,  1.58it/s]
Epoch 0:  42%|████▏     | 1388/3295 [14:39<20:08,  1.58it/s]
Epoch 0:  42%|████▏     | 1388/3295 [14:39<20:08,  1.58it/s]
Epoch 0:  42%|████▏     | 1389/3295 [14:39<20:07,  1.58it/s]
Epoch 0:  42%|████▏     | 1389/3295 [14:39<20:07,  1.58it/s]
Epoch 0:  42%|████▏     | 1390/3295 [14:39<20:05,  1.58it/s]
Epoch 0:  42%|████▏     | 1390/3295 [14:39<20:05,  1.58it/s]
Epoch 0:  42%|████▏     | 1391/3295 [14:41<20:06,  1.58it/s]
Epoch 0:  42%|████▏     | 1391/3295 [14:41<20:06,  1.58it/s]
Epoch 0:  42%|████▏     | 1392/3295 [14:41<20:05,  1.58it/s]
Epoch 0:  42%|████▏     | 1392/3295 [14:41<20:05,  1.58it/s]
Epoch 0:  42%|████▏     | 1393/3295 [14:41<20:03,  1.58it/s]
Epoch 0:  42%|████▏     | 1393/3295 [14:41<20:03,  1.58it/s]
Epoch 0:  42%|████▏     | 1394/3295 [14:43<20:04,  1.58it/s]
Epoch 0:  42%|████▏     | 1394/3295 [14:43<20:04,  1.58it/s]
Epoch 0:  42%|████▏     | 1395/3295 [14:43<20:03,  1.58it/s]
Epoch 0:  42%|████▏     | 1395/3295 [14:43<20:03,  1.58it/s]
Epoch 0:  42%|████▏     | 1396/3295 [14:43<20:02,  1.58it/s]
Epoch 0:  42%|████▏     | 1396/3295 [14:43<20:02,  1.58it/s]
Epoch 0:  42%|████▏     | 1397/3295 [14:45<20:03,  1.58it/s]
Epoch 0:  42%|████▏     | 1397/3295 [14:45<20:03,  1.58it/s]
Epoch 0:  42%|████▏     | 1398/3295 [14:45<20:01,  1.58it/s]
Epoch 0:  42%|████▏     | 1398/3295 [14:45<20:01,  1.58it/s]
Epoch 0:  42%|████▏     | 1399/3295 [14:45<20:00,  1.58it/s]
Epoch 0:  42%|████▏     | 1399/3295 [14:45<20:00,  1.58it/s]
Epoch 0:  42%|████▏     | 1400/3295 [14:47<20:01,  1.58it/s]
Epoch 0:  42%|████▏     | 1400/3295 [14:47<20:01,  1.58it/s]
Epoch 0:  43%|████▎     | 1401/3295 [14:47<19:59,  1.58it/s]
Epoch 0:  43%|████▎     | 1401/3295 [14:47<19:59,  1.58it/s]
Epoch 0:  43%|████▎     | 1402/3295 [14:47<19:58,  1.58it/s]
Epoch 0:  43%|████▎     | 1402/3295 [14:47<19:58,  1.58it/s]
Epoch 0:  43%|████▎     | 1403/3295 [14:49<19:59,  1.58it/s]
Epoch 0:  43%|████▎     | 1403/3295 [14:49<19:59,  1.58it/s]
Epoch 0:  43%|████▎     | 1404/3295 [14:49<19:57,  1.58it/s]
Epoch 0:  43%|████▎     | 1404/3295 [14:49<19:57,  1.58it/s]
Epoch 0:  43%|████▎     | 1405/3295 [14:49<19:56,  1.58it/s]
Epoch 0:  43%|████▎     | 1405/3295 [14:49<19:56,  1.58it/s]
Epoch 0:  43%|████▎     | 1406/3295 [14:51<19:57,  1.58it/s]
Epoch 0:  43%|████▎     | 1406/3295 [14:51<19:57,  1.58it/s]
Epoch 0:  43%|████▎     | 1407/3295 [14:51<19:55,  1.58it/s]
Epoch 0:  43%|████▎     | 1407/3295 [14:51<19:55,  1.58it/s]
Epoch 0:  43%|████▎     | 1408/3295 [14:51<19:54,  1.58it/s]
Epoch 0:  43%|████▎     | 1408/3295 [14:51<19:54,  1.58it/s]
Epoch 0:  43%|████▎     | 1409/3295 [14:53<19:55,  1.58it/s]
Epoch 0:  43%|████▎     | 1409/3295 [14:53<19:55,  1.58it/s]
Epoch 0:  43%|████▎     | 1410/3295 [14:53<19:53,  1.58it/s]
Epoch 0:  43%|████▎     | 1410/3295 [14:53<19:53,  1.58it/s]
Epoch 0:  43%|████▎     | 1411/3295 [14:53<19:52,  1.58it/s]
Epoch 0:  43%|████▎     | 1411/3295 [14:53<19:52,  1.58it/s]
Epoch 0:  43%|████▎     | 1412/3295 [14:54<19:53,  1.58it/s]
Epoch 0:  43%|████▎     | 1412/3295 [14:54<19:53,  1.58it/s]
Epoch 0:  43%|████▎     | 1413/3295 [14:54<19:51,  1.58it/s]
Epoch 0:  43%|████▎     | 1413/3295 [14:54<19:51,  1.58it/s]
Epoch 0:  43%|████▎     | 1414/3295 [14:54<19:50,  1.58it/s]
Epoch 0:  43%|████▎     | 1414/3295 [14:54<19:50,  1.58it/s]
Epoch 0:  43%|████▎     | 1415/3295 [14:56<19:51,  1.58it/s]
Epoch 0:  43%|████▎     | 1415/3295 [14:56<19:51,  1.58it/s]
Epoch 0:  43%|████▎     | 1416/3295 [14:56<19:49,  1.58it/s]
Epoch 0:  43%|████▎     | 1416/3295 [14:56<19:49,  1.58it/s]
Epoch 0:  43%|████▎     | 1417/3295 [14:56<19:48,  1.58it/s]
Epoch 0:  43%|████▎     | 1417/3295 [14:56<19:48,  1.58it/s]
Epoch 0:  43%|████▎     | 1418/3295 [14:58<19:49,  1.58it/s]
Epoch 0:  43%|████▎     | 1418/3295 [14:58<19:49,  1.58it/s]
Epoch 0:  43%|████▎     | 1419/3295 [14:58<19:47,  1.58it/s]
Epoch 0:  43%|████▎     | 1419/3295 [14:58<19:47,  1.58it/s]
Epoch 0:  43%|████▎     | 1420/3295 [14:58<19:46,  1.58it/s]
Epoch 0:  43%|████▎     | 1420/3295 [14:58<19:46,  1.58it/s]
Epoch 0:  43%|████▎     | 1421/3295 [15:00<19:47,  1.58it/s]
Epoch 0:  43%|████▎     | 1421/3295 [15:00<19:47,  1.58it/s]
Epoch 0:  43%|████▎     | 1422/3295 [15:00<19:46,  1.58it/s]
Epoch 0:  43%|████▎     | 1422/3295 [15:00<19:46,  1.58it/s]
Epoch 0:  43%|████▎     | 1423/3295 [15:00<19:44,  1.58it/s]
Epoch 0:  43%|████▎     | 1423/3295 [15:00<19:44,  1.58it/s]
Epoch 0:  43%|████▎     | 1424/3295 [15:02<19:45,  1.58it/s]
Epoch 0:  43%|████▎     | 1424/3295 [15:02<19:45,  1.58it/s]
Epoch 0:  43%|████▎     | 1425/3295 [15:02<19:44,  1.58it/s]
Epoch 0:  43%|████▎     | 1425/3295 [15:02<19:44,  1.58it/s]
Epoch 0:  43%|████▎     | 1426/3295 [15:02<19:42,  1.58it/s]
Epoch 0:  43%|████▎     | 1426/3295 [15:02<19:42,  1.58it/s]
Epoch 0:  43%|████▎     | 1427/3295 [15:04<19:43,  1.58it/s]
Epoch 0:  43%|████▎     | 1427/3295 [15:04<19:43,  1.58it/s]
Epoch 0:  43%|████▎     | 1428/3295 [15:04<19:41,  1.58it/s]
Epoch 0:  43%|████▎     | 1428/3295 [15:04<19:41,  1.58it/s]
Epoch 0:  43%|████▎     | 1429/3295 [15:04<19:40,  1.58it/s]
Epoch 0:  43%|████▎     | 1429/3295 [15:04<19:40,  1.58it/s]
Epoch 0:  43%|████▎     | 1430/3295 [15:05<19:41,  1.58it/s]
Epoch 0:  43%|████▎     | 1430/3295 [15:05<19:41,  1.58it/s]
Epoch 0:  43%|████▎     | 1431/3295 [15:05<19:39,  1.58it/s]
Epoch 0:  43%|████▎     | 1431/3295 [15:05<19:39,  1.58it/s]
Epoch 0:  43%|████▎     | 1432/3295 [15:05<19:38,  1.58it/s]
Epoch 0:  43%|████▎     | 1432/3295 [15:05<19:38,  1.58it/s]
Epoch 0:  43%|████▎     | 1433/3295 [15:07<19:39,  1.58it/s]
Epoch 0:  43%|████▎     | 1433/3295 [15:07<19:39,  1.58it/s]
Epoch 0:  44%|████▎     | 1434/3295 [15:07<19:37,  1.58it/s]
Epoch 0:  44%|████▎     | 1434/3295 [15:07<19:37,  1.58it/s]
Epoch 0:  44%|████▎     | 1435/3295 [15:07<19:36,  1.58it/s]
Epoch 0:  44%|████▎     | 1435/3295 [15:07<19:36,  1.58it/s]
Epoch 0:  44%|████▎     | 1436/3295 [15:09<19:37,  1.58it/s]
Epoch 0:  44%|████▎     | 1436/3295 [15:09<19:37,  1.58it/s]
Epoch 0:  44%|████▎     | 1437/3295 [15:09<19:35,  1.58it/s]
Epoch 0:  44%|████▎     | 1437/3295 [15:09<19:35,  1.58it/s]
Epoch 0:  44%|████▎     | 1438/3295 [15:09<19:34,  1.58it/s]
Epoch 0:  44%|████▎     | 1438/3295 [15:09<19:34,  1.58it/s]
Epoch 0:  44%|████▎     | 1439/3295 [15:11<19:35,  1.58it/s]
Epoch 0:  44%|████▎     | 1439/3295 [15:11<19:35,  1.58it/s]
Epoch 0:  44%|████▎     | 1440/3295 [15:11<19:34,  1.58it/s]
Epoch 0:  44%|████▎     | 1440/3295 [15:11<19:34,  1.58it/s]
Epoch 0:  44%|████▎     | 1441/3295 [15:11<19:32,  1.58it/s]
Epoch 0:  44%|████▎     | 1441/3295 [15:11<19:32,  1.58it/s]
Epoch 0:  44%|████▍     | 1442/3295 [15:13<19:33,  1.58it/s]
Epoch 0:  44%|████▍     | 1442/3295 [15:13<19:33,  1.58it/s]
Epoch 0:  44%|████▍     | 1443/3295 [15:13<19:32,  1.58it/s]
Epoch 0:  44%|████▍     | 1443/3295 [15:13<19:32,  1.58it/s]
Epoch 0:  44%|████▍     | 1444/3295 [15:13<19:30,  1.58it/s]
Epoch 0:  44%|████▍     | 1444/3295 [15:13<19:30,  1.58it/s]
Epoch 0:  44%|████▍     | 1445/3295 [15:15<19:31,  1.58it/s]
Epoch 0:  44%|████▍     | 1445/3295 [15:15<19:31,  1.58it/s]
Epoch 0:  44%|████▍     | 1446/3295 [15:15<19:30,  1.58it/s]
Epoch 0:  44%|████▍     | 1446/3295 [15:15<19:30,  1.58it/s]
Epoch 0:  44%|████▍     | 1447/3295 [15:15<19:28,  1.58it/s]
Epoch 0:  44%|████▍     | 1447/3295 [15:15<19:28,  1.58it/s]
Epoch 0:  44%|████▍     | 1448/3295 [15:16<19:29,  1.58it/s]
Epoch 0:  44%|████▍     | 1448/3295 [15:16<19:29,  1.58it/s]
Epoch 0:  44%|████▍     | 1449/3295 [15:16<19:28,  1.58it/s]
Epoch 0:  44%|████▍     | 1449/3295 [15:16<19:28,  1.58it/s]
Epoch 0:  44%|████▍     | 1450/3295 [15:16<19:26,  1.58it/s]
Epoch 0:  44%|████▍     | 1450/3295 [15:16<19:26,  1.58it/s]
Epoch 0:  44%|████▍     | 1451/3295 [15:18<19:27,  1.58it/s]
Epoch 0:  44%|████▍     | 1451/3295 [15:18<19:27,  1.58it/s]
Epoch 0:  44%|████▍     | 1452/3295 [15:18<19:26,  1.58it/s]
Epoch 0:  44%|████▍     | 1452/3295 [15:18<19:26,  1.58it/s]
Epoch 0:  44%|████▍     | 1453/3295 [15:18<19:24,  1.58it/s]
Epoch 0:  44%|████▍     | 1453/3295 [15:18<19:24,  1.58it/s]
Epoch 0:  44%|████▍     | 1454/3295 [15:20<19:25,  1.58it/s]
Epoch 0:  44%|████▍     | 1454/3295 [15:20<19:25,  1.58it/s]
Epoch 0:  44%|████▍     | 1455/3295 [15:20<19:24,  1.58it/s]
Epoch 0:  44%|████▍     | 1455/3295 [15:20<19:24,  1.58it/s]
Epoch 0:  44%|████▍     | 1456/3295 [15:20<19:22,  1.58it/s]
Epoch 0:  44%|████▍     | 1456/3295 [15:20<19:22,  1.58it/s]
Epoch 0:  44%|████▍     | 1457/3295 [15:22<19:23,  1.58it/s]
Epoch 0:  44%|████▍     | 1457/3295 [15:22<19:23,  1.58it/s]
Epoch 0:  44%|████▍     | 1458/3295 [15:22<19:22,  1.58it/s]
Epoch 0:  44%|████▍     | 1458/3295 [15:22<19:22,  1.58it/s]
Epoch 0:  44%|████▍     | 1459/3295 [15:22<19:20,  1.58it/s]
Epoch 0:  44%|████▍     | 1459/3295 [15:22<19:20,  1.58it/s]
Epoch 0:  44%|████▍     | 1460/3295 [15:24<19:21,  1.58it/s]
Epoch 0:  44%|████▍     | 1460/3295 [15:24<19:21,  1.58it/s]
Epoch 0:  44%|████▍     | 1461/3295 [15:24<19:20,  1.58it/s]
Epoch 0:  44%|████▍     | 1461/3295 [15:24<19:20,  1.58it/s]
Epoch 0:  44%|████▍     | 1462/3295 [15:24<19:18,  1.58it/s]
Epoch 0:  44%|████▍     | 1462/3295 [15:24<19:18,  1.58it/s]
Epoch 0:  44%|████▍     | 1463/3295 [15:26<19:19,  1.58it/s]
Epoch 0:  44%|████▍     | 1463/3295 [15:26<19:19,  1.58it/s]
Epoch 0:  44%|████▍     | 1464/3295 [15:26<19:18,  1.58it/s]
Epoch 0:  44%|████▍     | 1464/3295 [15:26<19:18,  1.58it/s]
Epoch 0:  44%|████▍     | 1465/3295 [15:26<19:16,  1.58it/s]
Epoch 0:  44%|████▍     | 1465/3295 [15:26<19:16,  1.58it/s]
Epoch 0:  44%|████▍     | 1466/3295 [15:27<19:17,  1.58it/s]
Epoch 0:  44%|████▍     | 1466/3295 [15:27<19:17,  1.58it/s]
Epoch 0:  45%|████▍     | 1467/3295 [15:28<19:16,  1.58it/s]
Epoch 0:  45%|████▍     | 1467/3295 [15:28<19:16,  1.58it/s]
Epoch 0:  45%|████▍     | 1468/3295 [15:28<19:14,  1.58it/s]
Epoch 0:  45%|████▍     | 1468/3295 [15:28<19:14,  1.58it/s]
Epoch 0:  45%|████▍     | 1469/3295 [15:29<19:15,  1.58it/s]
Epoch 0:  45%|████▍     | 1469/3295 [15:29<19:15,  1.58it/s]
Epoch 0:  45%|████▍     | 1470/3295 [15:29<19:14,  1.58it/s]
Epoch 0:  45%|████▍     | 1470/3295 [15:29<19:14,  1.58it/s]
Epoch 0:  45%|████▍     | 1471/3295 [15:29<19:13,  1.58it/s]
Epoch 0:  45%|████▍     | 1471/3295 [15:29<19:13,  1.58it/s]
Epoch 0:  45%|████▍     | 1472/3295 [15:31<19:14,  1.58it/s]
Epoch 0:  45%|████▍     | 1472/3295 [15:31<19:14,  1.58it/s]
Epoch 0:  45%|████▍     | 1473/3295 [15:31<19:12,  1.58it/s]
Epoch 0:  45%|████▍     | 1473/3295 [15:31<19:12,  1.58it/s]
Epoch 0:  45%|████▍     | 1474/3295 [15:32<19:11,  1.58it/s]
Epoch 0:  45%|████▍     | 1474/3295 [15:32<19:11,  1.58it/s]
Epoch 0:  45%|████▍     | 1475/3295 [15:33<19:12,  1.58it/s]
Epoch 0:  45%|████▍     | 1475/3295 [15:33<19:12,  1.58it/s]
Epoch 0:  45%|████▍     | 1476/3295 [15:33<19:10,  1.58it/s]
Epoch 0:  45%|████▍     | 1476/3295 [15:33<19:10,  1.58it/s]
Epoch 0:  45%|████▍     | 1477/3295 [15:33<19:09,  1.58it/s]
Epoch 0:  45%|████▍     | 1477/3295 [15:33<19:09,  1.58it/s]
Epoch 0:  45%|████▍     | 1478/3295 [15:35<19:10,  1.58it/s]
Epoch 0:  45%|████▍     | 1478/3295 [15:35<19:10,  1.58it/s]
Epoch 0:  45%|████▍     | 1479/3295 [15:35<19:08,  1.58it/s]
Epoch 0:  45%|████▍     | 1479/3295 [15:35<19:08,  1.58it/s]
Epoch 0:  45%|████▍     | 1480/3295 [15:35<19:07,  1.58it/s]
Epoch 0:  45%|████▍     | 1480/3295 [15:35<19:07,  1.58it/s]
Epoch 0:  45%|████▍     | 1481/3295 [15:37<19:08,  1.58it/s]
Epoch 0:  45%|████▍     | 1481/3295 [15:37<19:08,  1.58it/s]
Epoch 0:  45%|████▍     | 1482/3295 [15:37<19:06,  1.58it/s]
Epoch 0:  45%|████▍     | 1482/3295 [15:37<19:06,  1.58it/s]
Epoch 0:  45%|████▌     | 1483/3295 [15:37<19:05,  1.58it/s]
Epoch 0:  45%|████▌     | 1483/3295 [15:37<19:05,  1.58it/s]
Epoch 0:  45%|████▌     | 1484/3295 [15:39<19:06,  1.58it/s]
Epoch 0:  45%|████▌     | 1484/3295 [15:39<19:06,  1.58it/s]
Epoch 0:  45%|████▌     | 1485/3295 [15:39<19:04,  1.58it/s]
Epoch 0:  45%|████▌     | 1485/3295 [15:39<19:04,  1.58it/s]
Epoch 0:  45%|████▌     | 1486/3295 [15:39<19:03,  1.58it/s]
Epoch 0:  45%|████▌     | 1486/3295 [15:39<19:03,  1.58it/s]
Epoch 0:  45%|████▌     | 1487/3295 [15:41<19:04,  1.58it/s]
Epoch 0:  45%|████▌     | 1487/3295 [15:41<19:04,  1.58it/s]
Epoch 0:  45%|████▌     | 1488/3295 [15:41<19:02,  1.58it/s]
Epoch 0:  45%|████▌     | 1488/3295 [15:41<19:02,  1.58it/s]
Epoch 0:  45%|████▌     | 1489/3295 [15:41<19:01,  1.58it/s]
Epoch 0:  45%|████▌     | 1489/3295 [15:41<19:01,  1.58it/s]
Epoch 0:  45%|████▌     | 1490/3295 [15:42<19:02,  1.58it/s]
Epoch 0:  45%|████▌     | 1490/3295 [15:42<19:02,  1.58it/s]
Epoch 0:  45%|████▌     | 1491/3295 [15:42<19:00,  1.58it/s]
Epoch 0:  45%|████▌     | 1491/3295 [15:42<19:00,  1.58it/s]
Epoch 0:  45%|████▌     | 1492/3295 [15:42<18:59,  1.58it/s]
Epoch 0:  45%|████▌     | 1492/3295 [15:42<18:59,  1.58it/s]
Epoch 0:  45%|████▌     | 1493/3295 [15:44<18:59,  1.58it/s]
Epoch 0:  45%|████▌     | 1493/3295 [15:44<18:59,  1.58it/s]
Epoch 0:  45%|████▌     | 1494/3295 [15:44<18:58,  1.58it/s]
Epoch 0:  45%|████▌     | 1494/3295 [15:44<18:58,  1.58it/s]
Epoch 0:  45%|████▌     | 1495/3295 [15:44<18:57,  1.58it/s]
Epoch 0:  45%|████▌     | 1495/3295 [15:44<18:57,  1.58it/s]
Epoch 0:  45%|████▌     | 1496/3295 [15:46<18:58,  1.58it/s]
Epoch 0:  45%|████▌     | 1496/3295 [15:46<18:58,  1.58it/s]
Epoch 0:  45%|████▌     | 1497/3295 [15:46<18:56,  1.58it/s]
Epoch 0:  45%|████▌     | 1497/3295 [15:46<18:56,  1.58it/s]
Epoch 0:  45%|████▌     | 1498/3295 [15:46<18:55,  1.58it/s]
Epoch 0:  45%|████▌     | 1498/3295 [15:46<18:55,  1.58it/s]
Epoch 0:  45%|████▌     | 1499/3295 [15:48<18:56,  1.58it/s]
Epoch 0:  45%|████▌     | 1499/3295 [15:48<18:56,  1.58it/s]
Epoch 0:  46%|████▌     | 1500/3295 [15:48<18:54,  1.58it/s]
Epoch 0:  46%|████▌     | 1500/3295 [15:48<18:54,  1.58it/s]
Epoch 0:  46%|████▌     | 1501/3295 [15:48<18:53,  1.58it/s]
Epoch 0:  46%|████▌     | 1501/3295 [15:48<18:53,  1.58it/s]
Epoch 0:  46%|████▌     | 1502/3295 [15:50<18:54,  1.58it/s]
Epoch 0:  46%|████▌     | 1502/3295 [15:50<18:54,  1.58it/s]
Epoch 0:  46%|████▌     | 1503/3295 [15:50<18:52,  1.58it/s]
Epoch 0:  46%|████▌     | 1503/3295 [15:50<18:52,  1.58it/s]
Epoch 0:  46%|████▌     | 1504/3295 [15:50<18:51,  1.58it/s]
Epoch 0:  46%|████▌     | 1504/3295 [15:50<18:51,  1.58it/s]
Epoch 0:  46%|████▌     | 1505/3295 [15:51<18:52,  1.58it/s]
Epoch 0:  46%|████▌     | 1505/3295 [15:51<18:52,  1.58it/s]
Epoch 0:  46%|████▌     | 1506/3295 [15:51<18:50,  1.58it/s]
Epoch 0:  46%|████▌     | 1506/3295 [15:51<18:50,  1.58it/s]
Epoch 0:  46%|████▌     | 1507/3295 [15:51<18:49,  1.58it/s]
Epoch 0:  46%|████▌     | 1507/3295 [15:51<18:49,  1.58it/s]
Epoch 0:  46%|████▌     | 1508/3295 [15:53<18:50,  1.58it/s]
Epoch 0:  46%|████▌     | 1508/3295 [15:53<18:50,  1.58it/s]
Epoch 0:  46%|████▌     | 1509/3295 [15:53<18:49,  1.58it/s]
Epoch 0:  46%|████▌     | 1509/3295 [15:53<18:49,  1.58it/s]
Epoch 0:  46%|████▌     | 1510/3295 [15:53<18:47,  1.58it/s]
Epoch 0:  46%|████▌     | 1510/3295 [15:53<18:47,  1.58it/s]
Epoch 0:  46%|████▌     | 1511/3295 [15:55<18:48,  1.58it/s]
Epoch 0:  46%|████▌     | 1511/3295 [15:55<18:48,  1.58it/s]
Epoch 0:  46%|████▌     | 1512/3295 [15:55<18:47,  1.58it/s]
Epoch 0:  46%|████▌     | 1512/3295 [15:55<18:47,  1.58it/s]
Epoch 0:  46%|████▌     | 1513/3295 [15:55<18:45,  1.58it/s]
Epoch 0:  46%|████▌     | 1513/3295 [15:55<18:45,  1.58it/s]
Epoch 0:  46%|████▌     | 1514/3295 [15:57<18:46,  1.58it/s]
Epoch 0:  46%|████▌     | 1514/3295 [15:57<18:46,  1.58it/s]
Epoch 0:  46%|████▌     | 1515/3295 [15:57<18:45,  1.58it/s]
Epoch 0:  46%|████▌     | 1515/3295 [15:57<18:45,  1.58it/s]
Epoch 0:  46%|████▌     | 1516/3295 [15:57<18:43,  1.58it/s]
Epoch 0:  46%|████▌     | 1516/3295 [15:57<18:43,  1.58it/s]
Epoch 0:  46%|████▌     | 1517/3295 [15:59<18:44,  1.58it/s]
Epoch 0:  46%|████▌     | 1517/3295 [15:59<18:44,  1.58it/s]
Epoch 0:  46%|████▌     | 1518/3295 [15:59<18:43,  1.58it/s]
Epoch 0:  46%|████▌     | 1518/3295 [15:59<18:43,  1.58it/s]
Epoch 0:  46%|████▌     | 1519/3295 [15:59<18:41,  1.58it/s]
Epoch 0:  46%|████▌     | 1519/3295 [15:59<18:41,  1.58it/s]
Epoch 0:  46%|████▌     | 1520/3295 [16:01<18:42,  1.58it/s]
Epoch 0:  46%|████▌     | 1520/3295 [16:01<18:42,  1.58it/s]
Epoch 0:  46%|████▌     | 1521/3295 [16:01<18:41,  1.58it/s]
Epoch 0:  46%|████▌     | 1521/3295 [16:01<18:41,  1.58it/s]
Epoch 0:  46%|████▌     | 1522/3295 [16:01<18:40,  1.58it/s]
Epoch 0:  46%|████▌     | 1522/3295 [16:01<18:40,  1.58it/s]
Epoch 0:  46%|████▌     | 1523/3295 [16:03<18:40,  1.58it/s]
Epoch 0:  46%|████▌     | 1523/3295 [16:03<18:40,  1.58it/s]
Epoch 0:  46%|████▋     | 1524/3295 [16:03<18:39,  1.58it/s]
Epoch 0:  46%|████▋     | 1524/3295 [16:03<18:39,  1.58it/s]
Epoch 0:  46%|████▋     | 1525/3295 [16:03<18:38,  1.58it/s]
Epoch 0:  46%|████▋     | 1525/3295 [16:03<18:38,  1.58it/s]
Epoch 0:  46%|████▋     | 1526/3295 [16:05<18:38,  1.58it/s]
Epoch 0:  46%|████▋     | 1526/3295 [16:05<18:38,  1.58it/s]
Epoch 0:  46%|████▋     | 1527/3295 [16:05<18:37,  1.58it/s]
Epoch 0:  46%|████▋     | 1527/3295 [16:05<18:37,  1.58it/s]
Epoch 0:  46%|████▋     | 1528/3295 [16:05<18:36,  1.58it/s]
Epoch 0:  46%|████▋     | 1528/3295 [16:05<18:36,  1.58it/s]
Epoch 0:  46%|████▋     | 1529/3295 [16:07<18:37,  1.58it/s]
Epoch 0:  46%|████▋     | 1529/3295 [16:07<18:37,  1.58it/s]
Epoch 0:  46%|████▋     | 1530/3295 [16:07<18:35,  1.58it/s]
Epoch 0:  46%|████▋     | 1530/3295 [16:07<18:35,  1.58it/s]
Epoch 0:  46%|████▋     | 1531/3295 [16:07<18:34,  1.58it/s]
Epoch 0:  46%|████▋     | 1531/3295 [16:07<18:34,  1.58it/s]
Epoch 0:  46%|████▋     | 1532/3295 [16:09<18:35,  1.58it/s]
Epoch 0:  46%|████▋     | 1532/3295 [16:09<18:35,  1.58it/s]
Epoch 0:  47%|████▋     | 1533/3295 [16:09<18:33,  1.58it/s]
Epoch 0:  47%|████▋     | 1533/3295 [16:09<18:33,  1.58it/s]
Epoch 0:  47%|████▋     | 1534/3295 [16:09<18:32,  1.58it/s]
Epoch 0:  47%|████▋     | 1534/3295 [16:09<18:32,  1.58it/s]
Epoch 0:  47%|████▋     | 1535/3295 [16:10<18:33,  1.58it/s]
Epoch 0:  47%|████▋     | 1535/3295 [16:10<18:33,  1.58it/s]
Epoch 0:  47%|████▋     | 1536/3295 [16:10<18:31,  1.58it/s]
Epoch 0:  47%|████▋     | 1536/3295 [16:10<18:31,  1.58it/s]
Epoch 0:  47%|████▋     | 1537/3295 [16:11<18:30,  1.58it/s]
Epoch 0:  47%|████▋     | 1537/3295 [16:11<18:30,  1.58it/s]
Epoch 0:  47%|████▋     | 1538/3295 [16:12<18:31,  1.58it/s]
Epoch 0:  47%|████▋     | 1538/3295 [16:12<18:31,  1.58it/s]
Epoch 0:  47%|████▋     | 1539/3295 [16:12<18:30,  1.58it/s]
Epoch 0:  47%|████▋     | 1539/3295 [16:12<18:30,  1.58it/s]
Epoch 0:  47%|████▋     | 1540/3295 [16:12<18:28,  1.58it/s]
Epoch 0:  47%|████▋     | 1540/3295 [16:12<18:28,  1.58it/s]
Epoch 0:  47%|████▋     | 1541/3295 [16:14<18:29,  1.58it/s]
Epoch 0:  47%|████▋     | 1541/3295 [16:14<18:29,  1.58it/s]
Epoch 0:  47%|████▋     | 1542/3295 [16:14<18:28,  1.58it/s]
Epoch 0:  47%|████▋     | 1542/3295 [16:14<18:28,  1.58it/s]
Epoch 0:  47%|████▋     | 1543/3295 [16:14<18:26,  1.58it/s]
Epoch 0:  47%|████▋     | 1543/3295 [16:14<18:26,  1.58it/s]
Epoch 0:  47%|████▋     | 1544/3295 [16:16<18:27,  1.58it/s]
Epoch 0:  47%|████▋     | 1544/3295 [16:16<18:27,  1.58it/s]
Epoch 0:  47%|████▋     | 1545/3295 [16:16<18:26,  1.58it/s]
Epoch 0:  47%|████▋     | 1545/3295 [16:16<18:26,  1.58it/s]
Epoch 0:  47%|████▋     | 1546/3295 [16:16<18:24,  1.58it/s]
Epoch 0:  47%|████▋     | 1546/3295 [16:16<18:24,  1.58it/s]
Epoch 0:  47%|████▋     | 1547/3295 [16:18<18:25,  1.58it/s]
Epoch 0:  47%|████▋     | 1547/3295 [16:18<18:25,  1.58it/s]
Epoch 0:  47%|████▋     | 1548/3295 [16:18<18:24,  1.58it/s]
Epoch 0:  47%|████▋     | 1548/3295 [16:18<18:24,  1.58it/s]
Epoch 0:  47%|████▋     | 1549/3295 [16:18<18:22,  1.58it/s]
Epoch 0:  47%|████▋     | 1549/3295 [16:18<18:22,  1.58it/s]
Epoch 0:  47%|████▋     | 1550/3295 [16:20<18:23,  1.58it/s]
Epoch 0:  47%|████▋     | 1550/3295 [16:20<18:23,  1.58it/s]
Epoch 0:  47%|████▋     | 1551/3295 [16:20<18:22,  1.58it/s]
Epoch 0:  47%|████▋     | 1551/3295 [16:20<18:22,  1.58it/s]
Epoch 0:  47%|████▋     | 1552/3295 [16:20<18:20,  1.58it/s]
Epoch 0:  47%|████▋     | 1552/3295 [16:20<18:20,  1.58it/s]
Epoch 0:  47%|████▋     | 1553/3295 [16:22<18:21,  1.58it/s]
Epoch 0:  47%|████▋     | 1553/3295 [16:22<18:21,  1.58it/s]
Epoch 0:  47%|████▋     | 1554/3295 [16:22<18:20,  1.58it/s]
Epoch 0:  47%|████▋     | 1554/3295 [16:22<18:20,  1.58it/s]
Epoch 0:  47%|████▋     | 1555/3295 [16:22<18:18,  1.58it/s]
Epoch 0:  47%|████▋     | 1555/3295 [16:22<18:18,  1.58it/s]
Epoch 0:  47%|████▋     | 1556/3295 [16:23<18:19,  1.58it/s]
Epoch 0:  47%|████▋     | 1556/3295 [16:23<18:19,  1.58it/s]
Epoch 0:  47%|████▋     | 1557/3295 [16:23<18:18,  1.58it/s]
Epoch 0:  47%|████▋     | 1557/3295 [16:23<18:18,  1.58it/s]
Epoch 0:  47%|████▋     | 1558/3295 [16:23<18:17,  1.58it/s]
Epoch 0:  47%|████▋     | 1558/3295 [16:23<18:17,  1.58it/s]
Epoch 0:  47%|████▋     | 1559/3295 [16:25<18:17,  1.58it/s]
Epoch 0:  47%|████▋     | 1559/3295 [16:25<18:17,  1.58it/s]
Epoch 0:  47%|████▋     | 1560/3295 [16:25<18:16,  1.58it/s]
Epoch 0:  47%|████▋     | 1560/3295 [16:25<18:16,  1.58it/s]
Epoch 0:  47%|████▋     | 1561/3295 [16:25<18:15,  1.58it/s]
Epoch 0:  47%|████▋     | 1561/3295 [16:25<18:15,  1.58it/s]
Epoch 0:  47%|████▋     | 1562/3295 [16:27<18:15,  1.58it/s]
Epoch 0:  47%|████▋     | 1562/3295 [16:27<18:15,  1.58it/s]
Epoch 0:  47%|████▋     | 1563/3295 [16:27<18:14,  1.58it/s]
Epoch 0:  47%|████▋     | 1563/3295 [16:27<18:14,  1.58it/s]
Epoch 0:  47%|████▋     | 1564/3295 [16:27<18:12,  1.58it/s]
Epoch 0:  47%|████▋     | 1564/3295 [16:27<18:12,  1.58it/s]
Epoch 0:  47%|████▋     | 1565/3295 [16:29<18:13,  1.58it/s]
Epoch 0:  47%|████▋     | 1565/3295 [16:29<18:13,  1.58it/s]
Epoch 0:  48%|████▊     | 1566/3295 [16:29<18:12,  1.58it/s]
Epoch 0:  48%|████▊     | 1566/3295 [16:29<18:12,  1.58it/s]
Epoch 0:  48%|████▊     | 1567/3295 [16:29<18:10,  1.58it/s]
Epoch 0:  48%|████▊     | 1567/3295 [16:29<18:10,  1.58it/s]
Epoch 0:  48%|████▊     | 1568/3295 [16:31<18:11,  1.58it/s]
Epoch 0:  48%|████▊     | 1568/3295 [16:31<18:11,  1.58it/s]
Epoch 0:  48%|████▊     | 1569/3295 [16:31<18:10,  1.58it/s]
Epoch 0:  48%|████▊     | 1569/3295 [16:31<18:10,  1.58it/s]
Epoch 0:  48%|████▊     | 1570/3295 [16:31<18:09,  1.58it/s]
Epoch 0:  48%|████▊     | 1570/3295 [16:31<18:09,  1.58it/s]
Epoch 0:  48%|████▊     | 1571/3295 [16:33<18:09,  1.58it/s]
Epoch 0:  48%|████▊     | 1571/3295 [16:33<18:09,  1.58it/s]
Epoch 0:  48%|████▊     | 1572/3295 [16:33<18:08,  1.58it/s]
Epoch 0:  48%|████▊     | 1572/3295 [16:33<18:08,  1.58it/s]
Epoch 0:  48%|████▊     | 1573/3295 [16:33<18:07,  1.58it/s]
Epoch 0:  48%|████▊     | 1573/3295 [16:33<18:07,  1.58it/s]
Epoch 0:  48%|████▊     | 1574/3295 [16:34<18:07,  1.58it/s]
Epoch 0:  48%|████▊     | 1574/3295 [16:34<18:07,  1.58it/s]
Epoch 0:  48%|████▊     | 1575/3295 [16:34<18:06,  1.58it/s]
Epoch 0:  48%|████▊     | 1575/3295 [16:34<18:06,  1.58it/s]
Epoch 0:  48%|████▊     | 1576/3295 [16:34<18:05,  1.58it/s]
Epoch 0:  48%|████▊     | 1576/3295 [16:34<18:05,  1.58it/s]
Epoch 0:  48%|████▊     | 1577/3295 [16:36<18:05,  1.58it/s]
Epoch 0:  48%|████▊     | 1577/3295 [16:36<18:05,  1.58it/s]
Epoch 0:  48%|████▊     | 1578/3295 [16:36<18:04,  1.58it/s]
Epoch 0:  48%|████▊     | 1578/3295 [16:36<18:04,  1.58it/s]
Epoch 0:  48%|████▊     | 1579/3295 [16:36<18:03,  1.58it/s]
Epoch 0:  48%|████▊     | 1579/3295 [16:36<18:03,  1.58it/s]
Epoch 0:  48%|████▊     | 1580/3295 [16:38<18:03,  1.58it/s]
Epoch 0:  48%|████▊     | 1580/3295 [16:38<18:03,  1.58it/s]
Epoch 0:  48%|████▊     | 1581/3295 [16:38<18:02,  1.58it/s]
Epoch 0:  48%|████▊     | 1581/3295 [16:38<18:02,  1.58it/s]
Epoch 0:  48%|████▊     | 1582/3295 [16:38<18:01,  1.58it/s]
Epoch 0:  48%|████▊     | 1582/3295 [16:38<18:01,  1.58it/s]
Epoch 0:  48%|████▊     | 1583/3295 [16:40<18:02,  1.58it/s]
Epoch 0:  48%|████▊     | 1583/3295 [16:40<18:02,  1.58it/s]
Epoch 0:  48%|████▊     | 1584/3295 [16:40<18:00,  1.58it/s]
Epoch 0:  48%|████▊     | 1584/3295 [16:40<18:00,  1.58it/s]
Epoch 0:  48%|████▊     | 1585/3295 [16:40<17:59,  1.58it/s]
Epoch 0:  48%|████▊     | 1585/3295 [16:40<17:59,  1.58it/s]
Epoch 0:  48%|████▊     | 1586/3295 [16:42<18:00,  1.58it/s]
Epoch 0:  48%|████▊     | 1586/3295 [16:42<18:00,  1.58it/s]
Epoch 0:  48%|████▊     | 1587/3295 [16:42<17:58,  1.58it/s]
Epoch 0:  48%|████▊     | 1587/3295 [16:42<17:58,  1.58it/s]
Epoch 0:  48%|████▊     | 1588/3295 [16:42<17:57,  1.58it/s]
Epoch 0:  48%|████▊     | 1588/3295 [16:42<17:57,  1.58it/s]
Epoch 0:  48%|████▊     | 1589/3295 [16:44<17:58,  1.58it/s]
Epoch 0:  48%|████▊     | 1589/3295 [16:44<17:58,  1.58it/s]
Epoch 0:  48%|████▊     | 1590/3295 [16:44<17:56,  1.58it/s]
Epoch 0:  48%|████▊     | 1590/3295 [16:44<17:56,  1.58it/s]
Epoch 0:  48%|████▊     | 1591/3295 [16:44<17:55,  1.58it/s]
Epoch 0:  48%|████▊     | 1591/3295 [16:44<17:55,  1.58it/s]
Epoch 0:  48%|████▊     | 1592/3295 [16:46<17:56,  1.58it/s]
Epoch 0:  48%|████▊     | 1592/3295 [16:46<17:56,  1.58it/s]
Epoch 0:  48%|████▊     | 1593/3295 [16:46<17:54,  1.58it/s]
Epoch 0:  48%|████▊     | 1593/3295 [16:46<17:54,  1.58it/s]
Epoch 0:  48%|████▊     | 1594/3295 [16:46<17:53,  1.58it/s]
Epoch 0:  48%|████▊     | 1594/3295 [16:46<17:53,  1.58it/s]
Epoch 0:  48%|████▊     | 1595/3295 [16:47<17:54,  1.58it/s]
Epoch 0:  48%|████▊     | 1595/3295 [16:47<17:54,  1.58it/s]
Epoch 0:  48%|████▊     | 1596/3295 [16:47<17:53,  1.58it/s]
Epoch 0:  48%|████▊     | 1596/3295 [16:48<17:53,  1.58it/s]
Epoch 0:  48%|████▊     | 1597/3295 [16:48<17:51,  1.58it/s]
Epoch 0:  48%|████▊     | 1597/3295 [16:48<17:51,  1.58it/s]
Epoch 0:  48%|████▊     | 1598/3295 [16:49<17:52,  1.58it/s]
Epoch 0:  48%|████▊     | 1598/3295 [16:49<17:52,  1.58it/s]
Epoch 0:  49%|████▊     | 1599/3295 [16:49<17:51,  1.58it/s]
Epoch 0:  49%|████▊     | 1599/3295 [16:49<17:51,  1.58it/s]
Epoch 0:  49%|████▊     | 1600/3295 [16:49<17:49,  1.58it/s]
Epoch 0:  49%|████▊     | 1600/3295 [16:49<17:49,  1.58it/s]
Epoch 0:  49%|████▊     | 1601/3295 [16:51<17:50,  1.58it/s]
Epoch 0:  49%|████▊     | 1601/3295 [16:51<17:50,  1.58it/s]
Epoch 0:  49%|████▊     | 1602/3295 [16:51<17:48,  1.58it/s]
Epoch 0:  49%|████▊     | 1602/3295 [16:51<17:48,  1.58it/s]
Epoch 0:  49%|████▊     | 1603/3295 [16:51<17:47,  1.58it/s]
Epoch 0:  49%|████▊     | 1603/3295 [16:51<17:47,  1.58it/s]
Epoch 0:  49%|████▊     | 1604/3295 [16:53<17:48,  1.58it/s]
Epoch 0:  49%|████▊     | 1604/3295 [16:53<17:48,  1.58it/s]
Epoch 0:  49%|████▊     | 1605/3295 [16:53<17:47,  1.58it/s]
Epoch 0:  49%|████▊     | 1605/3295 [16:53<17:47,  1.58it/s]
Epoch 0:  49%|████▊     | 1606/3295 [16:53<17:45,  1.58it/s]
Epoch 0:  49%|████▊     | 1606/3295 [16:53<17:45,  1.58it/s]
Epoch 0:  49%|████▉     | 1607/3295 [16:55<17:46,  1.58it/s]
Epoch 0:  49%|████▉     | 1607/3295 [16:55<17:46,  1.58it/s]
Epoch 0:  49%|████▉     | 1608/3295 [16:55<17:45,  1.58it/s]
Epoch 0:  49%|████▉     | 1608/3295 [16:55<17:45,  1.58it/s]
Epoch 0:  49%|████▉     | 1609/3295 [16:55<17:43,  1.58it/s]
Epoch 0:  49%|████▉     | 1609/3295 [16:55<17:43,  1.58it/s]
Epoch 0:  49%|████▉     | 1610/3295 [16:57<17:44,  1.58it/s]
Epoch 0:  49%|████▉     | 1610/3295 [16:57<17:44,  1.58it/s]
Epoch 0:  49%|████▉     | 1611/3295 [16:57<17:43,  1.58it/s]
Epoch 0:  49%|████▉     | 1611/3295 [16:57<17:43,  1.58it/s]
Epoch 0:  49%|████▉     | 1612/3295 [16:57<17:42,  1.58it/s]
Epoch 0:  49%|████▉     | 1612/3295 [16:57<17:42,  1.58it/s]
Epoch 0:  49%|████▉     | 1613/3295 [16:59<17:42,  1.58it/s]
Epoch 0:  49%|████▉     | 1613/3295 [16:59<17:42,  1.58it/s]
Epoch 0:  49%|████▉     | 1614/3295 [16:59<17:41,  1.58it/s]
Epoch 0:  49%|████▉     | 1614/3295 [16:59<17:41,  1.58it/s]
Epoch 0:  49%|████▉     | 1615/3295 [16:59<17:40,  1.58it/s]
Epoch 0:  49%|████▉     | 1615/3295 [16:59<17:40,  1.58it/s]
Epoch 0:  49%|████▉     | 1616/3295 [17:01<17:40,  1.58it/s]
Epoch 0:  49%|████▉     | 1616/3295 [17:01<17:40,  1.58it/s]
Epoch 0:  49%|████▉     | 1617/3295 [17:01<17:39,  1.58it/s]
Epoch 0:  49%|████▉     | 1617/3295 [17:01<17:39,  1.58it/s]
Epoch 0:  49%|████▉     | 1618/3295 [17:01<17:38,  1.58it/s]
Epoch 0:  49%|████▉     | 1618/3295 [17:01<17:38,  1.58it/s]
Epoch 0:  49%|████▉     | 1619/3295 [17:02<17:38,  1.58it/s]
Epoch 0:  49%|████▉     | 1619/3295 [17:02<17:38,  1.58it/s]
Epoch 0:  49%|████▉     | 1620/3295 [17:02<17:37,  1.58it/s]
Epoch 0:  49%|████▉     | 1620/3295 [17:02<17:37,  1.58it/s]
Epoch 0:  49%|████▉     | 1621/3295 [17:02<17:36,  1.58it/s]
Epoch 0:  49%|████▉     | 1621/3295 [17:02<17:36,  1.58it/s]
Epoch 0:  49%|████▉     | 1622/3295 [17:04<17:37,  1.58it/s]
Epoch 0:  49%|████▉     | 1622/3295 [17:04<17:37,  1.58it/s]
Epoch 0:  49%|████▉     | 1623/3295 [17:04<17:35,  1.58it/s]
Epoch 0:  49%|████▉     | 1623/3295 [17:04<17:35,  1.58it/s]
Epoch 0:  49%|████▉     | 1624/3295 [17:04<17:34,  1.58it/s]
Epoch 0:  49%|████▉     | 1624/3295 [17:04<17:34,  1.58it/s]
Epoch 0:  49%|████▉     | 1625/3295 [17:06<17:35,  1.58it/s]
Epoch 0:  49%|████▉     | 1625/3295 [17:06<17:35,  1.58it/s]
Epoch 0:  49%|████▉     | 1626/3295 [17:06<17:33,  1.58it/s]
Epoch 0:  49%|████▉     | 1626/3295 [17:06<17:33,  1.58it/s]
Epoch 0:  49%|████▉     | 1627/3295 [17:06<17:32,  1.58it/s]
Epoch 0:  49%|████▉     | 1627/3295 [17:06<17:32,  1.58it/s]
Epoch 0:  49%|████▉     | 1628/3295 [17:08<17:33,  1.58it/s]
Epoch 0:  49%|████▉     | 1628/3295 [17:08<17:33,  1.58it/s]
Epoch 0:  49%|████▉     | 1629/3295 [17:08<17:31,  1.58it/s]
Epoch 0:  49%|████▉     | 1629/3295 [17:08<17:31,  1.58it/s]
Epoch 0:  49%|████▉     | 1630/3295 [17:08<17:30,  1.58it/s]
Epoch 0:  49%|████▉     | 1630/3295 [17:08<17:30,  1.58it/s]
Epoch 0:  49%|████▉     | 1631/3295 [17:10<17:31,  1.58it/s]
Epoch 0:  49%|████▉     | 1631/3295 [17:10<17:31,  1.58it/s]
Epoch 0:  50%|████▉     | 1632/3295 [17:10<17:29,  1.58it/s]
Epoch 0:  50%|████▉     | 1632/3295 [17:10<17:29,  1.58it/s]
Epoch 0:  50%|████▉     | 1633/3295 [17:10<17:28,  1.59it/s]
Epoch 0:  50%|████▉     | 1633/3295 [17:10<17:28,  1.59it/s]
Epoch 0:  50%|████▉     | 1634/3295 [17:12<17:29,  1.58it/s]
Epoch 0:  50%|████▉     | 1634/3295 [17:12<17:29,  1.58it/s]
Epoch 0:  50%|████▉     | 1635/3295 [17:12<17:28,  1.58it/s]
Epoch 0:  50%|████▉     | 1635/3295 [17:12<17:28,  1.58it/s]
Epoch 0:  50%|████▉     | 1636/3295 [17:12<17:26,  1.58it/s]
Epoch 0:  50%|████▉     | 1636/3295 [17:12<17:26,  1.58it/s]
Epoch 0:  50%|████▉     | 1637/3295 [17:14<17:27,  1.58it/s]
Epoch 0:  50%|████▉     | 1637/3295 [17:14<17:27,  1.58it/s]
Epoch 0:  50%|████▉     | 1638/3295 [17:14<17:26,  1.58it/s]
Epoch 0:  50%|████▉     | 1638/3295 [17:14<17:26,  1.58it/s]
Epoch 0:  50%|████▉     | 1639/3295 [17:14<17:24,  1.58it/s]
Epoch 0:  50%|████▉     | 1639/3295 [17:14<17:24,  1.58it/s]
Epoch 0:  50%|████▉     | 1640/3295 [17:15<17:25,  1.58it/s]
Epoch 0:  50%|████▉     | 1640/3295 [17:15<17:25,  1.58it/s]
Epoch 0:  50%|████▉     | 1641/3295 [17:15<17:24,  1.58it/s]
Epoch 0:  50%|████▉     | 1641/3295 [17:15<17:24,  1.58it/s]
Epoch 0:  50%|████▉     | 1642/3295 [17:15<17:22,  1.58it/s]
Epoch 0:  50%|████▉     | 1642/3295 [17:15<17:22,  1.58it/s]
Epoch 0:  50%|████▉     | 1643/3295 [17:17<17:23,  1.58it/s]
Epoch 0:  50%|████▉     | 1643/3295 [17:17<17:23,  1.58it/s]
Epoch 0:  50%|████▉     | 1644/3295 [17:17<17:22,  1.58it/s]
Epoch 0:  50%|████▉     | 1644/3295 [17:17<17:22,  1.58it/s]
Epoch 0:  50%|████▉     | 1645/3295 [17:18<17:21,  1.58it/s]
Epoch 0:  50%|████▉     | 1645/3295 [17:18<17:21,  1.58it/s]
Epoch 0:  50%|████▉     | 1646/3295 [17:19<17:21,  1.58it/s]
Epoch 0:  50%|████▉     | 1646/3295 [17:19<17:21,  1.58it/s]
Epoch 0:  50%|████▉     | 1647/3295 [17:19<17:20,  1.58it/s]
Epoch 0:  50%|████▉     | 1647/3295 [17:19<17:20,  1.58it/s]
Epoch 0:  50%|█████     | 1648/3295 [17:19<17:19,  1.58it/s]
Epoch 0:  50%|█████     | 1648/3295 [17:19<17:19,  1.58it/s]
Epoch 0:  50%|█████     | 1649/3295 [17:21<17:19,  1.58it/s]
Epoch 0:  50%|█████     | 1649/3295 [17:21<17:19,  1.58it/s]
Epoch 0:  50%|█████     | 1650/3295 [17:21<17:18,  1.58it/s]
Epoch 0:  50%|█████     | 1650/3295 [17:21<17:18,  1.58it/s]
Epoch 0:  50%|█████     | 1651/3295 [17:21<17:17,  1.58it/s]
Epoch 0:  50%|█████     | 1651/3295 [17:21<17:17,  1.58it/s]
Epoch 0:  50%|█████     | 1652/3295 [17:23<17:18,  1.58it/s]
Epoch 0:  50%|█████     | 1652/3295 [17:23<17:18,  1.58it/s]
Epoch 0:  50%|█████     | 1653/3295 [17:23<17:16,  1.58it/s]
Epoch 0:  50%|█████     | 1653/3295 [17:23<17:16,  1.58it/s]
Epoch 0:  50%|█████     | 1654/3295 [17:23<17:15,  1.58it/s]
Epoch 0:  50%|█████     | 1654/3295 [17:23<17:15,  1.58it/s]
Epoch 0:  50%|█████     | 1655/3295 [17:25<17:16,  1.58it/s]
Epoch 0:  50%|█████     | 1655/3295 [17:25<17:16,  1.58it/s]
Epoch 0:  50%|█████     | 1656/3295 [17:25<17:14,  1.58it/s]
Epoch 0:  50%|█████     | 1656/3295 [17:25<17:14,  1.58it/s]
Epoch 0:  50%|█████     | 1657/3295 [17:25<17:13,  1.58it/s]
Epoch 0:  50%|█████     | 1657/3295 [17:25<17:13,  1.58it/s]
Epoch 0:  50%|█████     | 1658/3295 [17:27<17:14,  1.58it/s]
Epoch 0:  50%|█████     | 1658/3295 [17:27<17:14,  1.58it/s]
Epoch 0:  50%|█████     | 1659/3295 [17:27<17:12,  1.58it/s]
Epoch 0:  50%|█████     | 1659/3295 [17:27<17:12,  1.58it/s]
Epoch 0:  50%|█████     | 1660/3295 [17:27<17:11,  1.58it/s]
Epoch 0:  50%|█████     | 1660/3295 [17:27<17:11,  1.58it/s]
Epoch 0:  50%|█████     | 1661/3295 [17:29<17:12,  1.58it/s]
Epoch 0:  50%|█████     | 1661/3295 [17:29<17:12,  1.58it/s]
Epoch 0:  50%|█████     | 1662/3295 [17:29<17:10,  1.58it/s]
Epoch 0:  50%|█████     | 1662/3295 [17:29<17:10,  1.58it/s]
Epoch 0:  50%|█████     | 1663/3295 [17:29<17:09,  1.59it/s]
Epoch 0:  50%|█████     | 1663/3295 [17:29<17:09,  1.59it/s]
Epoch 0:  51%|█████     | 1664/3295 [17:31<17:10,  1.58it/s]
Epoch 0:  51%|█████     | 1664/3295 [17:31<17:10,  1.58it/s]
Epoch 0:  51%|█████     | 1665/3295 [17:31<17:08,  1.58it/s]
Epoch 0:  51%|█████     | 1665/3295 [17:31<17:08,  1.58it/s]
Epoch 0:  51%|█████     | 1666/3295 [17:31<17:07,  1.59it/s]
Epoch 0:  51%|█████     | 1666/3295 [17:31<17:07,  1.59it/s]
Epoch 0:  51%|█████     | 1667/3295 [17:32<17:08,  1.58it/s]
Epoch 0:  51%|█████     | 1667/3295 [17:32<17:08,  1.58it/s]
Epoch 0:  51%|█████     | 1668/3295 [17:32<17:07,  1.58it/s]
Epoch 0:  51%|█████     | 1668/3295 [17:32<17:07,  1.58it/s]
Epoch 0:  51%|█████     | 1669/3295 [17:32<17:05,  1.59it/s]
Epoch 0:  51%|█████     | 1669/3295 [17:32<17:05,  1.59it/s]
Epoch 0:  51%|█████     | 1670/3295 [17:34<17:06,  1.58it/s]
Epoch 0:  51%|█████     | 1670/3295 [17:34<17:06,  1.58it/s]
Epoch 0:  51%|█████     | 1671/3295 [17:34<17:05,  1.58it/s]
Epoch 0:  51%|█████     | 1671/3295 [17:34<17:05,  1.58it/s]
Epoch 0:  51%|█████     | 1672/3295 [17:34<17:03,  1.59it/s]
Epoch 0:  51%|█████     | 1672/3295 [17:34<17:03,  1.59it/s]
Epoch 0:  51%|█████     | 1673/3295 [17:36<17:04,  1.58it/s]
Epoch 0:  51%|█████     | 1673/3295 [17:36<17:04,  1.58it/s]
Epoch 0:  51%|█████     | 1674/3295 [17:36<17:03,  1.58it/s]
Epoch 0:  51%|█████     | 1674/3295 [17:36<17:03,  1.58it/s]
Epoch 0:  51%|█████     | 1675/3295 [17:36<17:01,  1.59it/s]
Epoch 0:  51%|█████     | 1675/3295 [17:36<17:01,  1.59it/s]
Epoch 0:  51%|█████     | 1676/3295 [17:38<17:02,  1.58it/s]
Epoch 0:  51%|█████     | 1676/3295 [17:38<17:02,  1.58it/s]
Epoch 0:  51%|█████     | 1677/3295 [17:38<17:01,  1.58it/s]
Epoch 0:  51%|█████     | 1677/3295 [17:38<17:01,  1.58it/s]
Epoch 0:  51%|█████     | 1678/3295 [17:38<17:00,  1.59it/s]
Epoch 0:  51%|█████     | 1678/3295 [17:38<17:00,  1.59it/s]
Epoch 0:  51%|█████     | 1679/3295 [17:40<17:00,  1.58it/s]
Epoch 0:  51%|█████     | 1679/3295 [17:40<17:00,  1.58it/s]
Epoch 0:  51%|█████     | 1680/3295 [17:40<16:59,  1.58it/s]
Epoch 0:  51%|█████     | 1680/3295 [17:40<16:59,  1.58it/s]
Epoch 0:  51%|█████     | 1681/3295 [17:40<16:58,  1.59it/s]
Epoch 0:  51%|█████     | 1681/3295 [17:40<16:58,  1.59it/s]
Epoch 0:  51%|█████     | 1682/3295 [17:42<16:58,  1.58it/s]
Epoch 0:  51%|█████     | 1682/3295 [17:42<16:58,  1.58it/s]
Epoch 0:  51%|█████     | 1683/3295 [17:42<16:57,  1.58it/s]
Epoch 0:  51%|█████     | 1683/3295 [17:42<16:57,  1.58it/s]
Epoch 0:  51%|█████     | 1684/3295 [17:42<16:56,  1.59it/s]
Epoch 0:  51%|█████     | 1684/3295 [17:42<16:56,  1.59it/s]
Epoch 0:  51%|█████     | 1685/3295 [17:44<16:56,  1.58it/s]
Epoch 0:  51%|█████     | 1685/3295 [17:44<16:56,  1.58it/s]
Epoch 0:  51%|█████     | 1686/3295 [17:44<16:55,  1.58it/s]
Epoch 0:  51%|█████     | 1686/3295 [17:44<16:55,  1.58it/s]
Epoch 0:  51%|█████     | 1687/3295 [17:44<16:54,  1.59it/s]
Epoch 0:  51%|█████     | 1687/3295 [17:44<16:54,  1.59it/s]
Epoch 0:  51%|█████     | 1688/3295 [17:45<16:54,  1.58it/s]
Epoch 0:  51%|█████     | 1688/3295 [17:45<16:54,  1.58it/s]
Epoch 0:  51%|█████▏    | 1689/3295 [17:45<16:53,  1.58it/s]
Epoch 0:  51%|█████▏    | 1689/3295 [17:45<16:53,  1.58it/s]
Epoch 0:  51%|█████▏    | 1690/3295 [17:45<16:52,  1.59it/s]
Epoch 0:  51%|█████▏    | 1690/3295 [17:45<16:52,  1.59it/s]
Epoch 0:  51%|█████▏    | 1691/3295 [17:47<16:52,  1.58it/s]
Epoch 0:  51%|█████▏    | 1691/3295 [17:47<16:52,  1.58it/s]
Epoch 0:  51%|█████▏    | 1692/3295 [17:47<16:51,  1.58it/s]
Epoch 0:  51%|█████▏    | 1692/3295 [17:47<16:51,  1.58it/s]
Epoch 0:  51%|█████▏    | 1693/3295 [17:47<16:50,  1.59it/s]
Epoch 0:  51%|█████▏    | 1693/3295 [17:47<16:50,  1.59it/s]
Epoch 0:  51%|█████▏    | 1694/3295 [17:49<16:50,  1.58it/s]
Epoch 0:  51%|█████▏    | 1694/3295 [17:49<16:50,  1.58it/s]
Epoch 0:  51%|█████▏    | 1695/3295 [17:49<16:49,  1.58it/s]
Epoch 0:  51%|█████▏    | 1695/3295 [17:49<16:49,  1.58it/s]
Epoch 0:  51%|█████▏    | 1696/3295 [17:49<16:48,  1.59it/s]
Epoch 0:  51%|█████▏    | 1696/3295 [17:49<16:48,  1.59it/s]
Epoch 0:  52%|█████▏    | 1697/3295 [17:51<16:49,  1.58it/s]
Epoch 0:  52%|█████▏    | 1697/3295 [17:51<16:49,  1.58it/s]
Epoch 0:  52%|█████▏    | 1698/3295 [17:51<16:47,  1.58it/s]
Epoch 0:  52%|█████▏    | 1698/3295 [17:51<16:47,  1.58it/s]
Epoch 0:  52%|█████▏    | 1699/3295 [17:51<16:46,  1.59it/s]
Epoch 0:  52%|█████▏    | 1699/3295 [17:51<16:46,  1.59it/s]
Epoch 0:  52%|█████▏    | 1700/3295 [17:53<16:47,  1.58it/s]
Epoch 0:  52%|█████▏    | 1700/3295 [17:53<16:47,  1.58it/s]
Epoch 0:  52%|█████▏    | 1701/3295 [17:53<16:45,  1.58it/s]
Epoch 0:  52%|█████▏    | 1701/3295 [17:53<16:45,  1.58it/s]
Epoch 0:  52%|█████▏    | 1702/3295 [17:53<16:44,  1.59it/s]
Epoch 0:  52%|█████▏    | 1702/3295 [17:53<16:44,  1.59it/s]
Epoch 0:  52%|█████▏    | 1703/3295 [17:55<16:45,  1.58it/s]
Epoch 0:  52%|█████▏    | 1703/3295 [17:55<16:45,  1.58it/s]
Epoch 0:  52%|█████▏    | 1704/3295 [17:55<16:43,  1.58it/s]
Epoch 0:  52%|█████▏    | 1704/3295 [17:55<16:43,  1.58it/s]
Epoch 0:  52%|█████▏    | 1705/3295 [17:55<16:42,  1.59it/s]
Epoch 0:  52%|█████▏    | 1705/3295 [17:55<16:42,  1.59it/s]
Epoch 0:  52%|█████▏    | 1706/3295 [17:57<16:43,  1.58it/s]
Epoch 0:  52%|█████▏    | 1706/3295 [17:57<16:43,  1.58it/s]
Epoch 0:  52%|█████▏    | 1707/3295 [17:57<16:42,  1.58it/s]
Epoch 0:  52%|█████▏    | 1707/3295 [17:57<16:42,  1.58it/s]
Epoch 0:  52%|█████▏    | 1708/3295 [17:57<16:40,  1.59it/s]
Epoch 0:  52%|█████▏    | 1708/3295 [17:57<16:40,  1.59it/s]
Epoch 0:  52%|█████▏    | 1709/3295 [17:58<16:41,  1.58it/s]
Epoch 0:  52%|█████▏    | 1709/3295 [17:58<16:41,  1.58it/s]
Epoch 0:  52%|█████▏    | 1710/3295 [17:58<16:40,  1.58it/s]
Epoch 0:  52%|█████▏    | 1710/3295 [17:58<16:40,  1.58it/s]
Epoch 0:  52%|█████▏    | 1711/3295 [17:59<16:38,  1.59it/s]
Epoch 0:  52%|█████▏    | 1711/3295 [17:59<16:38,  1.59it/s]
Epoch 0:  52%|█████▏    | 1712/3295 [18:00<16:39,  1.58it/s]
Epoch 0:  52%|█████▏    | 1712/3295 [18:00<16:39,  1.58it/s]
Epoch 0:  52%|█████▏    | 1713/3295 [18:00<16:38,  1.58it/s]
Epoch 0:  52%|█████▏    | 1713/3295 [18:00<16:38,  1.58it/s]
Epoch 0:  52%|█████▏    | 1714/3295 [18:00<16:37,  1.59it/s]
Epoch 0:  52%|█████▏    | 1714/3295 [18:00<16:37,  1.59it/s]
Epoch 0:  52%|█████▏    | 1715/3295 [18:02<16:37,  1.58it/s]
Epoch 0:  52%|█████▏    | 1715/3295 [18:02<16:37,  1.58it/s]
Epoch 0:  52%|█████▏    | 1716/3295 [18:02<16:36,  1.58it/s]
Epoch 0:  52%|█████▏    | 1716/3295 [18:02<16:36,  1.58it/s]
Epoch 0:  52%|█████▏    | 1717/3295 [18:02<16:35,  1.59it/s]
Epoch 0:  52%|█████▏    | 1717/3295 [18:02<16:35,  1.59it/s]
Epoch 0:  52%|█████▏    | 1718/3295 [18:04<16:35,  1.58it/s]
Epoch 0:  52%|█████▏    | 1718/3295 [18:04<16:35,  1.58it/s]
Epoch 0:  52%|█████▏    | 1719/3295 [18:04<16:34,  1.58it/s]
Epoch 0:  52%|█████▏    | 1719/3295 [18:04<16:34,  1.58it/s]
Epoch 0:  52%|█████▏    | 1720/3295 [18:04<16:33,  1.59it/s]
Epoch 0:  52%|█████▏    | 1720/3295 [18:04<16:33,  1.59it/s]
Epoch 0:  52%|█████▏    | 1721/3295 [18:06<16:33,  1.58it/s]
Epoch 0:  52%|█████▏    | 1721/3295 [18:06<16:33,  1.58it/s]
Epoch 0:  52%|█████▏    | 1722/3295 [18:06<16:32,  1.59it/s]
Epoch 0:  52%|█████▏    | 1722/3295 [18:06<16:32,  1.59it/s]
Epoch 0:  52%|█████▏    | 1723/3295 [18:06<16:31,  1.59it/s]
Epoch 0:  52%|█████▏    | 1723/3295 [18:06<16:31,  1.59it/s]
Epoch 0:  52%|█████▏    | 1724/3295 [18:08<16:31,  1.58it/s]
Epoch 0:  52%|█████▏    | 1724/3295 [18:08<16:31,  1.58it/s]
Epoch 0:  52%|█████▏    | 1725/3295 [18:08<16:30,  1.58it/s]
Epoch 0:  52%|█████▏    | 1725/3295 [18:08<16:30,  1.58it/s]
Epoch 0:  52%|█████▏    | 1726/3295 [18:08<16:29,  1.59it/s]
Epoch 0:  52%|█████▏    | 1726/3295 [18:08<16:29,  1.59it/s]
Epoch 0:  52%|█████▏    | 1727/3295 [18:10<16:29,  1.58it/s]
Epoch 0:  52%|█████▏    | 1727/3295 [18:10<16:29,  1.58it/s]
Epoch 0:  52%|█████▏    | 1728/3295 [18:10<16:28,  1.58it/s]
Epoch 0:  52%|█████▏    | 1728/3295 [18:10<16:28,  1.58it/s]
Epoch 0:  52%|█████▏    | 1729/3295 [18:10<16:27,  1.59it/s]
Epoch 0:  52%|█████▏    | 1729/3295 [18:10<16:27,  1.59it/s]
Epoch 0:  53%|█████▎    | 1730/3295 [18:11<16:27,  1.58it/s]
Epoch 0:  53%|█████▎    | 1730/3295 [18:11<16:27,  1.58it/s]
Epoch 0:  53%|█████▎    | 1731/3295 [18:11<16:26,  1.59it/s]
Epoch 0:  53%|█████▎    | 1731/3295 [18:11<16:26,  1.59it/s]
Epoch 0:  53%|█████▎    | 1732/3295 [18:12<16:25,  1.59it/s]
Epoch 0:  53%|█████▎    | 1732/3295 [18:12<16:25,  1.59it/s]
Epoch 0:  53%|█████▎    | 1733/3295 [18:13<16:25,  1.58it/s]
Epoch 0:  53%|█████▎    | 1733/3295 [18:13<16:25,  1.58it/s]
Epoch 0:  53%|█████▎    | 1734/3295 [18:13<16:24,  1.59it/s]
Epoch 0:  53%|█████▎    | 1734/3295 [18:13<16:24,  1.59it/s]
Epoch 0:  53%|█████▎    | 1735/3295 [18:13<16:23,  1.59it/s]
Epoch 0:  53%|█████▎    | 1735/3295 [18:13<16:23,  1.59it/s]
Epoch 0:  53%|█████▎    | 1736/3295 [18:15<16:24,  1.58it/s]
Epoch 0:  53%|█████▎    | 1736/3295 [18:15<16:24,  1.58it/s]
Epoch 0:  53%|█████▎    | 1737/3295 [18:15<16:22,  1.59it/s]
Epoch 0:  53%|█████▎    | 1737/3295 [18:15<16:22,  1.59it/s]
Epoch 0:  53%|█████▎    | 1738/3295 [18:15<16:21,  1.59it/s]
Epoch 0:  53%|█████▎    | 1738/3295 [18:15<16:21,  1.59it/s]
Epoch 0:  53%|█████▎    | 1739/3295 [18:17<16:21,  1.58it/s]
Epoch 0:  53%|█████▎    | 1739/3295 [18:17<16:21,  1.58it/s]
Epoch 0:  53%|█████▎    | 1740/3295 [18:17<16:20,  1.59it/s]
Epoch 0:  53%|█████▎    | 1740/3295 [18:17<16:20,  1.59it/s]
Epoch 0:  53%|█████▎    | 1741/3295 [18:17<16:19,  1.59it/s]
Epoch 0:  53%|█████▎    | 1741/3295 [18:17<16:19,  1.59it/s]
Epoch 0:  53%|█████▎    | 1742/3295 [18:19<16:20,  1.58it/s]
Epoch 0:  53%|█████▎    | 1742/3295 [18:19<16:20,  1.58it/s]
Epoch 0:  53%|█████▎    | 1743/3295 [18:19<16:18,  1.59it/s]
Epoch 0:  53%|█████▎    | 1743/3295 [18:19<16:18,  1.59it/s]
Epoch 0:  53%|█████▎    | 1744/3295 [18:19<16:17,  1.59it/s]
Epoch 0:  53%|█████▎    | 1744/3295 [18:19<16:17,  1.59it/s]
Epoch 0:  53%|█████▎    | 1745/3295 [18:21<16:18,  1.58it/s]
Epoch 0:  53%|█████▎    | 1745/3295 [18:21<16:18,  1.58it/s]
Epoch 0:  53%|█████▎    | 1746/3295 [18:21<16:16,  1.59it/s]
Epoch 0:  53%|█████▎    | 1746/3295 [18:21<16:16,  1.59it/s]
Epoch 0:  53%|█████▎    | 1747/3295 [18:21<16:15,  1.59it/s]
Epoch 0:  53%|█████▎    | 1747/3295 [18:21<16:15,  1.59it/s]
Epoch 0:  53%|█████▎    | 1748/3295 [18:22<16:16,  1.58it/s]
Epoch 0:  53%|█████▎    | 1748/3295 [18:22<16:16,  1.58it/s]
Epoch 0:  53%|█████▎    | 1749/3295 [18:22<16:14,  1.59it/s]
Epoch 0:  53%|█████▎    | 1749/3295 [18:22<16:14,  1.59it/s]
Epoch 0:  53%|█████▎    | 1750/3295 [18:22<16:13,  1.59it/s]
Epoch 0:  53%|█████▎    | 1750/3295 [18:22<16:13,  1.59it/s]
Epoch 0:  53%|█████▎    | 1751/3295 [18:24<16:14,  1.58it/s]
Epoch 0:  53%|█████▎    | 1751/3295 [18:24<16:14,  1.58it/s]
Epoch 0:  53%|█████▎    | 1752/3295 [18:24<16:13,  1.59it/s]
Epoch 0:  53%|█████▎    | 1752/3295 [18:24<16:13,  1.59it/s]
Epoch 0:  53%|█████▎    | 1753/3295 [18:24<16:11,  1.59it/s]
Epoch 0:  53%|█████▎    | 1753/3295 [18:24<16:11,  1.59it/s]
Epoch 0:  53%|█████▎    | 1754/3295 [18:26<16:12,  1.58it/s]
Epoch 0:  53%|█████▎    | 1754/3295 [18:26<16:12,  1.58it/s]
Epoch 0:  53%|█████▎    | 1755/3295 [18:26<16:11,  1.59it/s]
Epoch 0:  53%|█████▎    | 1755/3295 [18:26<16:11,  1.59it/s]
Epoch 0:  53%|█████▎    | 1756/3295 [18:26<16:09,  1.59it/s]
Epoch 0:  53%|█████▎    | 1756/3295 [18:26<16:09,  1.59it/s]
Epoch 0:  53%|█████▎    | 1757/3295 [18:28<16:10,  1.58it/s]
Epoch 0:  53%|█████▎    | 1757/3295 [18:28<16:10,  1.58it/s]
Epoch 0:  53%|█████▎    | 1758/3295 [18:28<16:09,  1.59it/s]
Epoch 0:  53%|█████▎    | 1758/3295 [18:28<16:09,  1.59it/s]
Epoch 0:  53%|█████▎    | 1759/3295 [18:28<16:08,  1.59it/s]
Epoch 0:  53%|█████▎    | 1759/3295 [18:28<16:08,  1.59it/s]
Epoch 0:  53%|█████▎    | 1760/3295 [18:30<16:08,  1.58it/s]
Epoch 0:  53%|█████▎    | 1760/3295 [18:30<16:08,  1.58it/s]
Epoch 0:  53%|█████▎    | 1761/3295 [18:30<16:07,  1.59it/s]
Epoch 0:  53%|█████▎    | 1761/3295 [18:30<16:07,  1.59it/s]
Epoch 0:  53%|█████▎    | 1762/3295 [18:30<16:06,  1.59it/s]
Epoch 0:  53%|█████▎    | 1762/3295 [18:30<16:06,  1.59it/s]
Epoch 0:  54%|█████▎    | 1763/3295 [18:32<16:06,  1.58it/s]
Epoch 0:  54%|█████▎    | 1763/3295 [18:32<16:06,  1.58it/s]
Epoch 0:  54%|█████▎    | 1764/3295 [18:32<16:05,  1.59it/s]
Epoch 0:  54%|█████▎    | 1764/3295 [18:32<16:05,  1.59it/s]
Epoch 0:  54%|█████▎    | 1765/3295 [18:32<16:04,  1.59it/s]
Epoch 0:  54%|█████▎    | 1765/3295 [18:32<16:04,  1.59it/s]
Epoch 0:  54%|█████▎    | 1766/3295 [18:34<16:04,  1.58it/s]
Epoch 0:  54%|█████▎    | 1766/3295 [18:34<16:04,  1.58it/s]
Epoch 0:  54%|█████▎    | 1767/3295 [18:34<16:03,  1.59it/s]
Epoch 0:  54%|█████▎    | 1767/3295 [18:34<16:03,  1.59it/s]
Epoch 0:  54%|█████▎    | 1768/3295 [18:34<16:02,  1.59it/s]
Epoch 0:  54%|█████▎    | 1768/3295 [18:34<16:02,  1.59it/s]
Epoch 0:  54%|█████▎    | 1769/3295 [18:36<16:03,  1.58it/s]
Epoch 0:  54%|█████▎    | 1769/3295 [18:36<16:03,  1.58it/s]
Epoch 0:  54%|█████▎    | 1770/3295 [18:36<16:01,  1.59it/s]
Epoch 0:  54%|█████▎    | 1770/3295 [18:36<16:01,  1.59it/s]
Epoch 0:  54%|█████▎    | 1771/3295 [18:36<16:00,  1.59it/s]
Epoch 0:  54%|█████▎    | 1771/3295 [18:36<16:00,  1.59it/s]
Epoch 0:  54%|█████▍    | 1772/3295 [18:38<16:01,  1.58it/s]
Epoch 0:  54%|█████▍    | 1772/3295 [18:38<16:01,  1.58it/s]
Epoch 0:  54%|█████▍    | 1773/3295 [18:38<15:59,  1.59it/s]
Epoch 0:  54%|█████▍    | 1773/3295 [18:38<15:59,  1.59it/s]
Epoch 0:  54%|█████▍    | 1774/3295 [18:38<15:58,  1.59it/s]
Epoch 0:  54%|█████▍    | 1774/3295 [18:38<15:58,  1.59it/s]
Epoch 0:  54%|█████▍    | 1775/3295 [18:39<15:59,  1.58it/s]
Epoch 0:  54%|█████▍    | 1775/3295 [18:39<15:59,  1.58it/s]
Epoch 0:  54%|█████▍    | 1776/3295 [18:39<15:57,  1.59it/s]
Epoch 0:  54%|█████▍    | 1776/3295 [18:39<15:57,  1.59it/s]
Epoch 0:  54%|█████▍    | 1777/3295 [18:39<15:56,  1.59it/s]
Epoch 0:  54%|█████▍    | 1777/3295 [18:39<15:56,  1.59it/s]
Epoch 0:  54%|█████▍    | 1778/3295 [18:41<15:56,  1.59it/s]
Epoch 0:  54%|█████▍    | 1778/3295 [18:41<15:56,  1.59it/s]
Epoch 0:  54%|█████▍    | 1779/3295 [18:41<15:55,  1.59it/s]
Epoch 0:  54%|█████▍    | 1779/3295 [18:41<15:55,  1.59it/s]
Epoch 0:  54%|█████▍    | 1780/3295 [18:41<15:54,  1.59it/s]
Epoch 0:  54%|█████▍    | 1780/3295 [18:41<15:54,  1.59it/s]
Epoch 0:  54%|█████▍    | 1781/3295 [18:43<15:55,  1.59it/s]
Epoch 0:  54%|█████▍    | 1781/3295 [18:43<15:55,  1.59it/s]
Epoch 0:  54%|█████▍    | 1782/3295 [18:43<15:54,  1.59it/s]
Epoch 0:  54%|█████▍    | 1782/3295 [18:43<15:54,  1.59it/s]
Epoch 0:  54%|█████▍    | 1783/3295 [18:43<15:52,  1.59it/s]
Epoch 0:  54%|█████▍    | 1783/3295 [18:43<15:52,  1.59it/s]
Epoch 0:  54%|█████▍    | 1784/3295 [18:45<15:53,  1.59it/s]
Epoch 0:  54%|█████▍    | 1784/3295 [18:45<15:53,  1.59it/s]
Epoch 0:  54%|█████▍    | 1785/3295 [18:45<15:52,  1.59it/s]
Epoch 0:  54%|█████▍    | 1785/3295 [18:45<15:52,  1.59it/s]
Epoch 0:  54%|█████▍    | 1786/3295 [18:45<15:51,  1.59it/s]
Epoch 0:  54%|█████▍    | 1786/3295 [18:45<15:51,  1.59it/s]
Epoch 0:  54%|█████▍    | 1787/3295 [18:47<15:51,  1.59it/s]
Epoch 0:  54%|█████▍    | 1787/3295 [18:47<15:51,  1.59it/s]
Epoch 0:  54%|█████▍    | 1788/3295 [18:47<15:50,  1.59it/s]
Epoch 0:  54%|█████▍    | 1788/3295 [18:47<15:50,  1.59it/s]
Epoch 0:  54%|█████▍    | 1789/3295 [18:47<15:49,  1.59it/s]
Epoch 0:  54%|█████▍    | 1789/3295 [18:47<15:49,  1.59it/s]
Epoch 0:  54%|█████▍    | 1790/3295 [18:49<15:49,  1.58it/s]
Epoch 0:  54%|█████▍    | 1790/3295 [18:49<15:49,  1.58it/s]
Epoch 0:  54%|█████▍    | 1791/3295 [18:49<15:48,  1.59it/s]
Epoch 0:  54%|█████▍    | 1791/3295 [18:49<15:48,  1.59it/s]
Epoch 0:  54%|█████▍    | 1792/3295 [18:49<15:47,  1.59it/s]
Epoch 0:  54%|█████▍    | 1792/3295 [18:49<15:47,  1.59it/s]
Epoch 0:  54%|█████▍    | 1793/3295 [18:51<15:47,  1.58it/s]
Epoch 0:  54%|█████▍    | 1793/3295 [18:51<15:47,  1.58it/s]
Epoch 0:  54%|█████▍    | 1794/3295 [18:51<15:46,  1.59it/s]
Epoch 0:  54%|█████▍    | 1794/3295 [18:51<15:46,  1.59it/s]
Epoch 0:  54%|█████▍    | 1795/3295 [18:51<15:45,  1.59it/s]
Epoch 0:  54%|█████▍    | 1795/3295 [18:51<15:45,  1.59it/s]
Epoch 0:  55%|█████▍    | 1796/3295 [18:53<15:45,  1.58it/s]
Epoch 0:  55%|█████▍    | 1796/3295 [18:53<15:45,  1.58it/s]
Epoch 0:  55%|█████▍    | 1797/3295 [18:53<15:44,  1.59it/s]
Epoch 0:  55%|█████▍    | 1797/3295 [18:53<15:44,  1.59it/s]
Epoch 0:  55%|█████▍    | 1798/3295 [18:53<15:43,  1.59it/s]
Epoch 0:  55%|█████▍    | 1798/3295 [18:53<15:43,  1.59it/s]
Epoch 0:  55%|█████▍    | 1799/3295 [18:55<15:43,  1.58it/s]
Epoch 0:  55%|█████▍    | 1799/3295 [18:55<15:43,  1.58it/s]
Epoch 0:  55%|█████▍    | 1800/3295 [18:55<15:42,  1.59it/s]
Epoch 0:  55%|█████▍    | 1800/3295 [18:55<15:42,  1.59it/s]
Epoch 0:  55%|█████▍    | 1801/3295 [18:55<15:41,  1.59it/s]
Epoch 0:  55%|█████▍    | 1801/3295 [18:55<15:41,  1.59it/s]
Epoch 0:  55%|█████▍    | 1802/3295 [18:56<15:41,  1.59it/s]
Epoch 0:  55%|█████▍    | 1802/3295 [18:56<15:41,  1.59it/s]
Epoch 0:  55%|█████▍    | 1803/3295 [18:56<15:40,  1.59it/s]
Epoch 0:  55%|█████▍    | 1803/3295 [18:56<15:40,  1.59it/s]
Epoch 0:  55%|█████▍    | 1804/3295 [18:56<15:39,  1.59it/s]
Epoch 0:  55%|█████▍    | 1804/3295 [18:56<15:39,  1.59it/s]
Epoch 0:  55%|█████▍    | 1805/3295 [18:58<15:39,  1.59it/s]
Epoch 0:  55%|█████▍    | 1805/3295 [18:58<15:39,  1.59it/s]
Epoch 0:  55%|█████▍    | 1806/3295 [18:58<15:38,  1.59it/s]
Epoch 0:  55%|█████▍    | 1806/3295 [18:58<15:38,  1.59it/s]
Epoch 0:  55%|█████▍    | 1807/3295 [18:58<15:37,  1.59it/s]
Epoch 0:  55%|█████▍    | 1807/3295 [18:58<15:37,  1.59it/s]
Epoch 0:  55%|█████▍    | 1808/3295 [19:00<15:38,  1.59it/s]
Epoch 0:  55%|█████▍    | 1808/3295 [19:00<15:38,  1.59it/s]
Epoch 0:  55%|█████▍    | 1809/3295 [19:00<15:36,  1.59it/s]
Epoch 0:  55%|█████▍    | 1809/3295 [19:00<15:36,  1.59it/s]
Epoch 0:  55%|█████▍    | 1810/3295 [19:00<15:35,  1.59it/s]
Epoch 0:  55%|█████▍    | 1810/3295 [19:00<15:35,  1.59it/s]
Epoch 0:  55%|█████▍    | 1811/3295 [19:02<15:36,  1.59it/s]
Epoch 0:  55%|█████▍    | 1811/3295 [19:02<15:36,  1.59it/s]
Epoch 0:  55%|█████▍    | 1812/3295 [19:02<15:35,  1.59it/s]
Epoch 0:  55%|█████▍    | 1812/3295 [19:02<15:35,  1.59it/s]
Epoch 0:  55%|█████▌    | 1813/3295 [19:02<15:34,  1.59it/s]
Epoch 0:  55%|█████▌    | 1813/3295 [19:02<15:34,  1.59it/s]
Epoch 0:  55%|█████▌    | 1814/3295 [19:04<15:34,  1.59it/s]
Epoch 0:  55%|█████▌    | 1814/3295 [19:04<15:34,  1.59it/s]
Epoch 0:  55%|█████▌    | 1815/3295 [19:04<15:33,  1.59it/s]
Epoch 0:  55%|█████▌    | 1815/3295 [19:04<15:33,  1.59it/s]
Epoch 0:  55%|█████▌    | 1816/3295 [19:04<15:32,  1.59it/s]
Epoch 0:  55%|█████▌    | 1816/3295 [19:04<15:32,  1.59it/s]
Epoch 0:  55%|█████▌    | 1817/3295 [19:06<15:32,  1.59it/s]
Epoch 0:  55%|█████▌    | 1817/3295 [19:06<15:32,  1.59it/s]
Epoch 0:  55%|█████▌    | 1818/3295 [19:06<15:31,  1.59it/s]
Epoch 0:  55%|█████▌    | 1818/3295 [19:06<15:31,  1.59it/s]
Epoch 0:  55%|█████▌    | 1819/3295 [19:06<15:30,  1.59it/s]
Epoch 0:  55%|█████▌    | 1819/3295 [19:06<15:30,  1.59it/s]
Epoch 0:  55%|█████▌    | 1820/3295 [19:08<15:30,  1.59it/s]
Epoch 0:  55%|█████▌    | 1820/3295 [19:08<15:30,  1.59it/s]
Epoch 0:  55%|█████▌    | 1821/3295 [19:08<15:29,  1.59it/s]
Epoch 0:  55%|█████▌    | 1821/3295 [19:08<15:29,  1.59it/s]
Epoch 0:  55%|█████▌    | 1822/3295 [19:08<15:28,  1.59it/s]
Epoch 0:  55%|█████▌    | 1822/3295 [19:08<15:28,  1.59it/s]
Epoch 0:  55%|█████▌    | 1823/3295 [19:09<15:28,  1.59it/s]
Epoch 0:  55%|█████▌    | 1823/3295 [19:09<15:28,  1.59it/s]
Epoch 0:  55%|█████▌    | 1824/3295 [19:09<15:27,  1.59it/s]
Epoch 0:  55%|█████▌    | 1824/3295 [19:09<15:27,  1.59it/s]
Epoch 0:  55%|█████▌    | 1825/3295 [19:09<15:26,  1.59it/s]
Epoch 0:  55%|█████▌    | 1825/3295 [19:09<15:26,  1.59it/s]
Epoch 0:  55%|█████▌    | 1826/3295 [19:11<15:26,  1.59it/s]
Epoch 0:  55%|█████▌    | 1826/3295 [19:11<15:26,  1.59it/s]
Epoch 0:  55%|█████▌    | 1827/3295 [19:11<15:25,  1.59it/s]
Epoch 0:  55%|█████▌    | 1827/3295 [19:11<15:25,  1.59it/s]
Epoch 0:  55%|█████▌    | 1828/3295 [19:11<15:24,  1.59it/s]
Epoch 0:  55%|█████▌    | 1828/3295 [19:11<15:24,  1.59it/s]
Epoch 0:  56%|█████▌    | 1829/3295 [19:13<15:24,  1.59it/s]
Epoch 0:  56%|█████▌    | 1829/3295 [19:13<15:24,  1.59it/s]
Epoch 0:  56%|█████▌    | 1830/3295 [19:13<15:23,  1.59it/s]
Epoch 0:  56%|█████▌    | 1830/3295 [19:13<15:23,  1.59it/s]
Epoch 0:  56%|█████▌    | 1831/3295 [19:13<15:22,  1.59it/s]
Epoch 0:  56%|█████▌    | 1831/3295 [19:13<15:22,  1.59it/s]
Epoch 0:  56%|█████▌    | 1832/3295 [19:15<15:22,  1.59it/s]
Epoch 0:  56%|█████▌    | 1832/3295 [19:15<15:22,  1.59it/s]
Epoch 0:  56%|█████▌    | 1833/3295 [19:15<15:21,  1.59it/s]
Epoch 0:  56%|█████▌    | 1833/3295 [19:15<15:21,  1.59it/s]
Epoch 0:  56%|█████▌    | 1834/3295 [19:15<15:20,  1.59it/s]
Epoch 0:  56%|█████▌    | 1834/3295 [19:15<15:20,  1.59it/s]
Epoch 0:  56%|█████▌    | 1835/3295 [19:17<15:20,  1.59it/s]
Epoch 0:  56%|█████▌    | 1835/3295 [19:17<15:20,  1.59it/s]
Epoch 0:  56%|█████▌    | 1836/3295 [19:17<15:19,  1.59it/s]
Epoch 0:  56%|█████▌    | 1836/3295 [19:17<15:19,  1.59it/s]
Epoch 0:  56%|█████▌    | 1837/3295 [19:17<15:18,  1.59it/s]
Epoch 0:  56%|█████▌    | 1837/3295 [19:17<15:18,  1.59it/s]
Epoch 0:  56%|█████▌    | 1838/3295 [19:19<15:18,  1.59it/s]
Epoch 0:  56%|█████▌    | 1838/3295 [19:19<15:18,  1.59it/s]
Epoch 0:  56%|█████▌    | 1839/3295 [19:19<15:17,  1.59it/s]
Epoch 0:  56%|█████▌    | 1839/3295 [19:19<15:17,  1.59it/s]
Epoch 0:  56%|█████▌    | 1840/3295 [19:19<15:16,  1.59it/s]
Epoch 0:  56%|█████▌    | 1840/3295 [19:19<15:16,  1.59it/s]
Epoch 0:  56%|█████▌    | 1841/3295 [19:21<15:17,  1.59it/s]
Epoch 0:  56%|█████▌    | 1841/3295 [19:21<15:17,  1.59it/s]
Epoch 0:  56%|█████▌    | 1842/3295 [19:21<15:15,  1.59it/s]
Epoch 0:  56%|█████▌    | 1842/3295 [19:21<15:15,  1.59it/s]
Epoch 0:  56%|█████▌    | 1843/3295 [19:21<15:14,  1.59it/s]
Epoch 0:  56%|█████▌    | 1843/3295 [19:21<15:14,  1.59it/s]
Epoch 0:  56%|█████▌    | 1844/3295 [19:22<15:15,  1.59it/s]
Epoch 0:  56%|█████▌    | 1844/3295 [19:22<15:15,  1.59it/s]
Epoch 0:  56%|█████▌    | 1845/3295 [19:22<15:13,  1.59it/s]
Epoch 0:  56%|█████▌    | 1845/3295 [19:22<15:13,  1.59it/s]
Epoch 0:  56%|█████▌    | 1846/3295 [19:22<15:12,  1.59it/s]
Epoch 0:  56%|█████▌    | 1846/3295 [19:22<15:12,  1.59it/s]
Epoch 0:  56%|█████▌    | 1847/3295 [19:24<15:13,  1.59it/s]
Epoch 0:  56%|█████▌    | 1847/3295 [19:24<15:13,  1.59it/s]
Epoch 0:  56%|█████▌    | 1848/3295 [19:24<15:12,  1.59it/s]
Epoch 0:  56%|█████▌    | 1848/3295 [19:24<15:12,  1.59it/s]
Epoch 0:  56%|█████▌    | 1849/3295 [19:24<15:10,  1.59it/s]
Epoch 0:  56%|█████▌    | 1849/3295 [19:24<15:10,  1.59it/s]
Epoch 0:  56%|█████▌    | 1850/3295 [19:26<15:11,  1.59it/s]
Epoch 0:  56%|█████▌    | 1850/3295 [19:26<15:11,  1.59it/s]
Epoch 0:  56%|█████▌    | 1851/3295 [19:26<15:10,  1.59it/s]
Epoch 0:  56%|█████▌    | 1851/3295 [19:26<15:10,  1.59it/s]
Epoch 0:  56%|█████▌    | 1852/3295 [19:26<15:09,  1.59it/s]
Epoch 0:  56%|█████▌    | 1852/3295 [19:26<15:09,  1.59it/s]
Epoch 0:  56%|█████▌    | 1853/3295 [19:28<15:09,  1.59it/s]
Epoch 0:  56%|█████▌    | 1853/3295 [19:28<15:09,  1.59it/s]
Epoch 0:  56%|█████▋    | 1854/3295 [19:28<15:08,  1.59it/s]
Epoch 0:  56%|█████▋    | 1854/3295 [19:28<15:08,  1.59it/s]
Epoch 0:  56%|█████▋    | 1855/3295 [19:28<15:07,  1.59it/s]
Epoch 0:  56%|█████▋    | 1855/3295 [19:28<15:07,  1.59it/s]
Epoch 0:  56%|█████▋    | 1856/3295 [19:30<15:07,  1.59it/s]
Epoch 0:  56%|█████▋    | 1856/3295 [19:30<15:07,  1.59it/s]
Epoch 0:  56%|█████▋    | 1857/3295 [19:30<15:06,  1.59it/s]
Epoch 0:  56%|█████▋    | 1857/3295 [19:30<15:06,  1.59it/s]
Epoch 0:  56%|█████▋    | 1858/3295 [19:30<15:05,  1.59it/s]
Epoch 0:  56%|█████▋    | 1858/3295 [19:30<15:05,  1.59it/s]
Epoch 0:  56%|█████▋    | 1859/3295 [19:32<15:05,  1.59it/s]
Epoch 0:  56%|█████▋    | 1859/3295 [19:32<15:05,  1.59it/s]
Epoch 0:  56%|█████▋    | 1860/3295 [19:32<15:04,  1.59it/s]
Epoch 0:  56%|█████▋    | 1860/3295 [19:32<15:04,  1.59it/s]
Epoch 0:  56%|█████▋    | 1861/3295 [19:32<15:03,  1.59it/s]
Epoch 0:  56%|█████▋    | 1861/3295 [19:32<15:03,  1.59it/s]
Epoch 0:  57%|█████▋    | 1862/3295 [19:34<15:03,  1.59it/s]
Epoch 0:  57%|█████▋    | 1862/3295 [19:34<15:03,  1.59it/s]
Epoch 0:  57%|█████▋    | 1863/3295 [19:34<15:02,  1.59it/s]
Epoch 0:  57%|█████▋    | 1863/3295 [19:34<15:02,  1.59it/s]
Epoch 0:  57%|█████▋    | 1864/3295 [19:34<15:01,  1.59it/s]
Epoch 0:  57%|█████▋    | 1864/3295 [19:34<15:01,  1.59it/s]
Epoch 0:  57%|█████▋    | 1865/3295 [19:36<15:01,  1.59it/s]
Epoch 0:  57%|█████▋    | 1865/3295 [19:36<15:01,  1.59it/s]
Epoch 0:  57%|█████▋    | 1866/3295 [19:36<15:00,  1.59it/s]
Epoch 0:  57%|█████▋    | 1866/3295 [19:36<15:00,  1.59it/s]
Epoch 0:  57%|█████▋    | 1867/3295 [19:36<14:59,  1.59it/s]
Epoch 0:  57%|█████▋    | 1867/3295 [19:36<14:59,  1.59it/s]
Epoch 0:  57%|█████▋    | 1868/3295 [19:37<14:59,  1.59it/s]
Epoch 0:  57%|█████▋    | 1868/3295 [19:37<14:59,  1.59it/s]
Epoch 0:  57%|█████▋    | 1869/3295 [19:37<14:58,  1.59it/s]
Epoch 0:  57%|█████▋    | 1869/3295 [19:37<14:58,  1.59it/s]
Epoch 0:  57%|█████▋    | 1870/3295 [19:37<14:57,  1.59it/s]
Epoch 0:  57%|█████▋    | 1870/3295 [19:37<14:57,  1.59it/s]
Epoch 0:  57%|█████▋    | 1871/3295 [19:39<14:57,  1.59it/s]
Epoch 0:  57%|█████▋    | 1871/3295 [19:39<14:57,  1.59it/s]
Epoch 0:  57%|█████▋    | 1872/3295 [19:39<14:56,  1.59it/s]
Epoch 0:  57%|█████▋    | 1872/3295 [19:39<14:56,  1.59it/s]
Epoch 0:  57%|█████▋    | 1873/3295 [19:39<14:55,  1.59it/s]
Epoch 0:  57%|█████▋    | 1873/3295 [19:39<14:55,  1.59it/s]
Epoch 0:  57%|█████▋    | 1874/3295 [19:41<14:55,  1.59it/s]
Epoch 0:  57%|█████▋    | 1874/3295 [19:41<14:55,  1.59it/s]
Epoch 0:  57%|█████▋    | 1875/3295 [19:41<14:54,  1.59it/s]
Epoch 0:  57%|█████▋    | 1875/3295 [19:41<14:54,  1.59it/s]
Epoch 0:  57%|█████▋    | 1876/3295 [19:41<14:53,  1.59it/s]
Epoch 0:  57%|█████▋    | 1876/3295 [19:41<14:53,  1.59it/s]
Epoch 0:  57%|█████▋    | 1877/3295 [19:43<14:53,  1.59it/s]
Epoch 0:  57%|█████▋    | 1877/3295 [19:43<14:53,  1.59it/s]
Epoch 0:  57%|█████▋    | 1878/3295 [19:43<14:52,  1.59it/s]
Epoch 0:  57%|█████▋    | 1878/3295 [19:43<14:52,  1.59it/s]
Epoch 0:  57%|█████▋    | 1879/3295 [19:43<14:51,  1.59it/s]
Epoch 0:  57%|█████▋    | 1879/3295 [19:43<14:51,  1.59it/s]
Epoch 0:  57%|█████▋    | 1880/3295 [19:45<14:52,  1.59it/s]
Epoch 0:  57%|█████▋    | 1880/3295 [19:45<14:52,  1.59it/s]
Epoch 0:  57%|█████▋    | 1881/3295 [19:45<14:51,  1.59it/s]
Epoch 0:  57%|█████▋    | 1881/3295 [19:45<14:51,  1.59it/s]
Epoch 0:  57%|█████▋    | 1882/3295 [19:45<14:49,  1.59it/s]
Epoch 0:  57%|█████▋    | 1882/3295 [19:45<14:49,  1.59it/s]
Epoch 0:  57%|█████▋    | 1883/3295 [19:47<14:50,  1.59it/s]
Epoch 0:  57%|█████▋    | 1883/3295 [19:47<14:50,  1.59it/s]
Epoch 0:  57%|█████▋    | 1884/3295 [19:47<14:49,  1.59it/s]
Epoch 0:  57%|█████▋    | 1884/3295 [19:47<14:49,  1.59it/s]
Epoch 0:  57%|█████▋    | 1885/3295 [19:47<14:48,  1.59it/s]
Epoch 0:  57%|█████▋    | 1885/3295 [19:47<14:48,  1.59it/s]
Epoch 0:  57%|█████▋    | 1886/3295 [19:49<14:48,  1.59it/s]
Epoch 0:  57%|█████▋    | 1886/3295 [19:49<14:48,  1.59it/s]
Epoch 0:  57%|█████▋    | 1887/3295 [19:49<14:47,  1.59it/s]
Epoch 0:  57%|█████▋    | 1887/3295 [19:49<14:47,  1.59it/s]
Epoch 0:  57%|█████▋    | 1888/3295 [19:49<14:46,  1.59it/s]
Epoch 0:  57%|█████▋    | 1888/3295 [19:49<14:46,  1.59it/s]
Epoch 0:  57%|█████▋    | 1889/3295 [19:50<14:46,  1.59it/s]
Epoch 0:  57%|█████▋    | 1889/3295 [19:50<14:46,  1.59it/s]
Epoch 0:  57%|█████▋    | 1890/3295 [19:50<14:45,  1.59it/s]
Epoch 0:  57%|█████▋    | 1890/3295 [19:50<14:45,  1.59it/s]
Epoch 0:  57%|█████▋    | 1891/3295 [19:50<14:44,  1.59it/s]
Epoch 0:  57%|█████▋    | 1891/3295 [19:50<14:44,  1.59it/s]
Epoch 0:  57%|█████▋    | 1892/3295 [19:52<14:44,  1.59it/s]
Epoch 0:  57%|█████▋    | 1892/3295 [19:52<14:44,  1.59it/s]
Epoch 0:  57%|█████▋    | 1893/3295 [19:52<14:43,  1.59it/s]
Epoch 0:  57%|█████▋    | 1893/3295 [19:52<14:43,  1.59it/s]
Epoch 0:  57%|█████▋    | 1894/3295 [19:52<14:42,  1.59it/s]
Epoch 0:  57%|█████▋    | 1894/3295 [19:52<14:42,  1.59it/s]
Epoch 0:  58%|█████▊    | 1895/3295 [19:54<14:42,  1.59it/s]
Epoch 0:  58%|█████▊    | 1895/3295 [19:54<14:42,  1.59it/s]
Epoch 0:  58%|█████▊    | 1896/3295 [19:54<14:41,  1.59it/s]
Epoch 0:  58%|█████▊    | 1896/3295 [19:54<14:41,  1.59it/s]
Epoch 0:  58%|█████▊    | 1897/3295 [19:54<14:40,  1.59it/s]
Epoch 0:  58%|█████▊    | 1897/3295 [19:54<14:40,  1.59it/s]
Epoch 0:  58%|█████▊    | 1898/3295 [19:56<14:40,  1.59it/s]
Epoch 0:  58%|█████▊    | 1898/3295 [19:56<14:40,  1.59it/s]
Epoch 0:  58%|█████▊    | 1899/3295 [19:56<14:39,  1.59it/s]
Epoch 0:  58%|█████▊    | 1899/3295 [19:56<14:39,  1.59it/s]
Epoch 0:  58%|█████▊    | 1900/3295 [19:56<14:38,  1.59it/s]
Epoch 0:  58%|█████▊    | 1900/3295 [19:56<14:38,  1.59it/s]
Epoch 0:  58%|█████▊    | 1901/3295 [19:58<14:38,  1.59it/s]
Epoch 0:  58%|█████▊    | 1901/3295 [19:58<14:38,  1.59it/s]
Epoch 0:  58%|█████▊    | 1902/3295 [19:58<14:37,  1.59it/s]
Epoch 0:  58%|█████▊    | 1902/3295 [19:58<14:37,  1.59it/s]
Epoch 0:  58%|█████▊    | 1903/3295 [19:58<14:36,  1.59it/s]
Epoch 0:  58%|█████▊    | 1903/3295 [19:58<14:36,  1.59it/s]
Epoch 0:  58%|█████▊    | 1904/3295 [20:00<14:37,  1.59it/s]
Epoch 0:  58%|█████▊    | 1904/3295 [20:00<14:37,  1.59it/s]
Epoch 0:  58%|█████▊    | 1905/3295 [20:00<14:35,  1.59it/s]
Epoch 0:  58%|█████▊    | 1905/3295 [20:00<14:35,  1.59it/s]
Epoch 0:  58%|█████▊    | 1906/3295 [20:00<14:34,  1.59it/s]
Epoch 0:  58%|█████▊    | 1906/3295 [20:00<14:34,  1.59it/s]
Epoch 0:  58%|█████▊    | 1907/3295 [20:02<14:35,  1.59it/s]
Epoch 0:  58%|█████▊    | 1907/3295 [20:02<14:35,  1.59it/s]
Epoch 0:  58%|█████▊    | 1908/3295 [20:02<14:34,  1.59it/s]
Epoch 0:  58%|█████▊    | 1908/3295 [20:02<14:34,  1.59it/s]
Epoch 0:  58%|█████▊    | 1909/3295 [20:02<14:32,  1.59it/s]
Epoch 0:  58%|█████▊    | 1909/3295 [20:02<14:32,  1.59it/s]
Epoch 0:  58%|█████▊    | 1910/3295 [20:04<14:33,  1.59it/s]
Epoch 0:  58%|█████▊    | 1910/3295 [20:04<14:33,  1.59it/s]
Epoch 0:  58%|█████▊    | 1911/3295 [20:04<14:32,  1.59it/s]
Epoch 0:  58%|█████▊    | 1911/3295 [20:04<14:32,  1.59it/s]
Epoch 0:  58%|█████▊    | 1912/3295 [20:04<14:31,  1.59it/s]
Epoch 0:  58%|█████▊    | 1912/3295 [20:04<14:31,  1.59it/s]
Epoch 0:  58%|█████▊    | 1913/3295 [20:06<14:31,  1.59it/s]
Epoch 0:  58%|█████▊    | 1913/3295 [20:06<14:31,  1.59it/s]
Epoch 0:  58%|█████▊    | 1914/3295 [20:06<14:30,  1.59it/s]
Epoch 0:  58%|█████▊    | 1914/3295 [20:06<14:30,  1.59it/s]
Epoch 0:  58%|█████▊    | 1915/3295 [20:06<14:29,  1.59it/s]
Epoch 0:  58%|█████▊    | 1915/3295 [20:06<14:29,  1.59it/s]
Epoch 0:  58%|█████▊    | 1916/3295 [20:07<14:29,  1.59it/s]
Epoch 0:  58%|█████▊    | 1916/3295 [20:07<14:29,  1.59it/s]
Epoch 0:  58%|█████▊    | 1917/3295 [20:07<14:28,  1.59it/s]
Epoch 0:  58%|█████▊    | 1917/3295 [20:07<14:28,  1.59it/s]
Epoch 0:  58%|█████▊    | 1918/3295 [20:07<14:27,  1.59it/s]
Epoch 0:  58%|█████▊    | 1918/3295 [20:07<14:27,  1.59it/s]
Epoch 0:  58%|█████▊    | 1919/3295 [20:09<14:27,  1.59it/s]
Epoch 0:  58%|█████▊    | 1919/3295 [20:09<14:27,  1.59it/s]
Epoch 0:  58%|█████▊    | 1920/3295 [20:09<14:26,  1.59it/s]
Epoch 0:  58%|█████▊    | 1920/3295 [20:09<14:26,  1.59it/s]
Epoch 0:  58%|█████▊    | 1921/3295 [20:09<14:25,  1.59it/s]
Epoch 0:  58%|█████▊    | 1921/3295 [20:09<14:25,  1.59it/s]
Epoch 0:  58%|█████▊    | 1922/3295 [20:11<14:25,  1.59it/s]
Epoch 0:  58%|█████▊    | 1922/3295 [20:11<14:25,  1.59it/s]
Epoch 0:  58%|█████▊    | 1923/3295 [20:11<14:24,  1.59it/s]
Epoch 0:  58%|█████▊    | 1923/3295 [20:11<14:24,  1.59it/s]
Epoch 0:  58%|█████▊    | 1924/3295 [20:11<14:23,  1.59it/s]
Epoch 0:  58%|█████▊    | 1924/3295 [20:11<14:23,  1.59it/s]
Epoch 0:  58%|█████▊    | 1925/3295 [20:13<14:23,  1.59it/s]
Epoch 0:  58%|█████▊    | 1925/3295 [20:13<14:23,  1.59it/s]
Epoch 0:  58%|█████▊    | 1926/3295 [20:13<14:22,  1.59it/s]
Epoch 0:  58%|█████▊    | 1926/3295 [20:13<14:22,  1.59it/s]
Epoch 0:  58%|█████▊    | 1927/3295 [20:13<14:21,  1.59it/s]
Epoch 0:  58%|█████▊    | 1927/3295 [20:13<14:21,  1.59it/s]
Epoch 0:  59%|█████▊    | 1928/3295 [20:15<14:21,  1.59it/s]
Epoch 0:  59%|█████▊    | 1928/3295 [20:15<14:21,  1.59it/s]
Epoch 0:  59%|█████▊    | 1929/3295 [20:15<14:20,  1.59it/s]
Epoch 0:  59%|█████▊    | 1929/3295 [20:15<14:20,  1.59it/s]
Epoch 0:  59%|█████▊    | 1930/3295 [20:15<14:19,  1.59it/s]
Epoch 0:  59%|█████▊    | 1930/3295 [20:15<14:19,  1.59it/s]
Epoch 0:  59%|█████▊    | 1931/3295 [20:17<14:19,  1.59it/s]
Epoch 0:  59%|█████▊    | 1931/3295 [20:17<14:19,  1.59it/s]
Epoch 0:  59%|█████▊    | 1932/3295 [20:17<14:18,  1.59it/s]
Epoch 0:  59%|█████▊    | 1932/3295 [20:17<14:18,  1.59it/s]
Epoch 0:  59%|█████▊    | 1933/3295 [20:17<14:17,  1.59it/s]
Epoch 0:  59%|█████▊    | 1933/3295 [20:17<14:17,  1.59it/s]
Epoch 0:  59%|█████▊    | 1934/3295 [20:19<14:17,  1.59it/s]
Epoch 0:  59%|█████▊    | 1934/3295 [20:19<14:17,  1.59it/s]
Epoch 0:  59%|█████▊    | 1935/3295 [20:19<14:16,  1.59it/s]
Epoch 0:  59%|█████▊    | 1935/3295 [20:19<14:16,  1.59it/s]
Epoch 0:  59%|█████▉    | 1936/3295 [20:19<14:15,  1.59it/s]
Epoch 0:  59%|█████▉    | 1936/3295 [20:19<14:15,  1.59it/s]
Epoch 0:  59%|█████▉    | 1937/3295 [20:20<14:15,  1.59it/s]
Epoch 0:  59%|█████▉    | 1937/3295 [20:20<14:15,  1.59it/s]
Epoch 0:  59%|█████▉    | 1938/3295 [20:20<14:14,  1.59it/s]
Epoch 0:  59%|█████▉    | 1938/3295 [20:20<14:14,  1.59it/s]
Epoch 0:  59%|█████▉    | 1939/3295 [20:21<14:13,  1.59it/s]
Epoch 0:  59%|█████▉    | 1939/3295 [20:21<14:13,  1.59it/s]
Epoch 0:  59%|█████▉    | 1940/3295 [20:22<14:14,  1.59it/s]
Epoch 0:  59%|█████▉    | 1940/3295 [20:22<14:14,  1.59it/s]
Epoch 0:  59%|█████▉    | 1941/3295 [20:22<14:12,  1.59it/s]
Epoch 0:  59%|█████▉    | 1941/3295 [20:22<14:12,  1.59it/s]
Epoch 0:  59%|█████▉    | 1942/3295 [20:22<14:11,  1.59it/s]
Epoch 0:  59%|█████▉    | 1942/3295 [20:22<14:11,  1.59it/s]
Epoch 0:  59%|█████▉    | 1943/3295 [20:24<14:12,  1.59it/s]
Epoch 0:  59%|█████▉    | 1943/3295 [20:24<14:12,  1.59it/s]
Epoch 0:  59%|█████▉    | 1944/3295 [20:24<14:11,  1.59it/s]
Epoch 0:  59%|█████▉    | 1944/3295 [20:24<14:11,  1.59it/s]
Epoch 0:  59%|█████▉    | 1945/3295 [20:24<14:09,  1.59it/s]
Epoch 0:  59%|█████▉    | 1945/3295 [20:24<14:09,  1.59it/s]
Epoch 0:  59%|█████▉    | 1946/3295 [20:26<14:10,  1.59it/s]
Epoch 0:  59%|█████▉    | 1946/3295 [20:26<14:10,  1.59it/s]
Epoch 0:  59%|█████▉    | 1947/3295 [20:26<14:09,  1.59it/s]
Epoch 0:  59%|█████▉    | 1947/3295 [20:26<14:09,  1.59it/s]
Epoch 0:  59%|█████▉    | 1948/3295 [20:26<14:08,  1.59it/s]
Epoch 0:  59%|█████▉    | 1948/3295 [20:26<14:08,  1.59it/s]
Epoch 0:  59%|█████▉    | 1949/3295 [20:28<14:08,  1.59it/s]
Epoch 0:  59%|█████▉    | 1949/3295 [20:28<14:08,  1.59it/s]
Epoch 0:  59%|█████▉    | 1950/3295 [20:28<14:07,  1.59it/s]
Epoch 0:  59%|█████▉    | 1950/3295 [20:28<14:07,  1.59it/s]
Epoch 0:  59%|█████▉    | 1951/3295 [20:28<14:06,  1.59it/s]
Epoch 0:  59%|█████▉    | 1951/3295 [20:28<14:06,  1.59it/s]
Epoch 0:  59%|█████▉    | 1952/3295 [20:30<14:06,  1.59it/s]
Epoch 0:  59%|█████▉    | 1952/3295 [20:30<14:06,  1.59it/s]
Epoch 0:  59%|█████▉    | 1953/3295 [20:30<14:05,  1.59it/s]
Epoch 0:  59%|█████▉    | 1953/3295 [20:30<14:05,  1.59it/s]
Epoch 0:  59%|█████▉    | 1954/3295 [20:30<14:04,  1.59it/s]
Epoch 0:  59%|█████▉    | 1954/3295 [20:30<14:04,  1.59it/s]
Epoch 0:  59%|█████▉    | 1955/3295 [20:32<14:04,  1.59it/s]
Epoch 0:  59%|█████▉    | 1955/3295 [20:32<14:04,  1.59it/s]
Epoch 0:  59%|█████▉    | 1956/3295 [20:32<14:03,  1.59it/s]
Epoch 0:  59%|█████▉    | 1956/3295 [20:32<14:03,  1.59it/s]
Epoch 0:  59%|█████▉    | 1957/3295 [20:32<14:02,  1.59it/s]
Epoch 0:  59%|█████▉    | 1957/3295 [20:32<14:02,  1.59it/s]
Epoch 0:  59%|█████▉    | 1958/3295 [20:33<14:02,  1.59it/s]
Epoch 0:  59%|█████▉    | 1958/3295 [20:33<14:02,  1.59it/s]
Epoch 0:  59%|█████▉    | 1959/3295 [20:33<14:01,  1.59it/s]
Epoch 0:  59%|█████▉    | 1959/3295 [20:33<14:01,  1.59it/s]
Epoch 0:  59%|█████▉    | 1960/3295 [20:33<14:00,  1.59it/s]
Epoch 0:  59%|█████▉    | 1960/3295 [20:33<14:00,  1.59it/s]
Epoch 0:  60%|█████▉    | 1961/3295 [20:35<14:00,  1.59it/s]
Epoch 0:  60%|█████▉    | 1961/3295 [20:35<14:00,  1.59it/s]
Epoch 0:  60%|█████▉    | 1962/3295 [20:35<13:59,  1.59it/s]
Epoch 0:  60%|█████▉    | 1962/3295 [20:35<13:59,  1.59it/s]
Epoch 0:  60%|█████▉    | 1963/3295 [20:35<13:58,  1.59it/s]
Epoch 0:  60%|█████▉    | 1963/3295 [20:35<13:58,  1.59it/s]
Epoch 0:  60%|█████▉    | 1964/3295 [20:37<13:58,  1.59it/s]
Epoch 0:  60%|█████▉    | 1964/3295 [20:37<13:58,  1.59it/s]
Epoch 0:  60%|█████▉    | 1965/3295 [20:37<13:57,  1.59it/s]
Epoch 0:  60%|█████▉    | 1965/3295 [20:37<13:57,  1.59it/s]
Epoch 0:  60%|█████▉    | 1966/3295 [20:37<13:56,  1.59it/s]
Epoch 0:  60%|█████▉    | 1966/3295 [20:37<13:56,  1.59it/s]
Epoch 0:  60%|█████▉    | 1967/3295 [20:39<13:56,  1.59it/s]
Epoch 0:  60%|█████▉    | 1967/3295 [20:39<13:56,  1.59it/s]
Epoch 0:  60%|█████▉    | 1968/3295 [20:39<13:55,  1.59it/s]
Epoch 0:  60%|█████▉    | 1968/3295 [20:39<13:55,  1.59it/s]
Epoch 0:  60%|█████▉    | 1969/3295 [20:39<13:54,  1.59it/s]
Epoch 0:  60%|█████▉    | 1969/3295 [20:39<13:54,  1.59it/s]
Epoch 0:  60%|█████▉    | 1970/3295 [20:41<13:54,  1.59it/s]
Epoch 0:  60%|█████▉    | 1970/3295 [20:41<13:54,  1.59it/s]
Epoch 0:  60%|█████▉    | 1971/3295 [20:41<13:53,  1.59it/s]
Epoch 0:  60%|█████▉    | 1971/3295 [20:41<13:53,  1.59it/s]
Epoch 0:  60%|█████▉    | 1972/3295 [20:41<13:52,  1.59it/s]
Epoch 0:  60%|█████▉    | 1972/3295 [20:41<13:52,  1.59it/s]
Epoch 0:  60%|█████▉    | 1973/3295 [20:43<13:53,  1.59it/s]
Epoch 0:  60%|█████▉    | 1973/3295 [20:43<13:53,  1.59it/s]
Epoch 0:  60%|█████▉    | 1974/3295 [20:43<13:52,  1.59it/s]
Epoch 0:  60%|█████▉    | 1974/3295 [20:43<13:52,  1.59it/s]
Epoch 0:  60%|█████▉    | 1975/3295 [20:43<13:51,  1.59it/s]
Epoch 0:  60%|█████▉    | 1975/3295 [20:43<13:51,  1.59it/s]
Epoch 0:  60%|█████▉    | 1976/3295 [20:45<13:51,  1.59it/s]
Epoch 0:  60%|█████▉    | 1976/3295 [20:45<13:51,  1.59it/s]
Epoch 0:  60%|██████    | 1977/3295 [20:45<13:50,  1.59it/s]
Epoch 0:  60%|██████    | 1977/3295 [20:45<13:50,  1.59it/s]
Epoch 0:  60%|██████    | 1978/3295 [20:45<13:49,  1.59it/s]
Epoch 0:  60%|██████    | 1978/3295 [20:45<13:49,  1.59it/s]
Epoch 0:  60%|██████    | 1979/3295 [20:47<13:49,  1.59it/s]
Epoch 0:  60%|██████    | 1979/3295 [20:47<13:49,  1.59it/s]
Epoch 0:  60%|██████    | 1980/3295 [20:47<13:48,  1.59it/s]
Epoch 0:  60%|██████    | 1980/3295 [20:47<13:48,  1.59it/s]
Epoch 0:  60%|██████    | 1981/3295 [20:47<13:47,  1.59it/s]
Epoch 0:  60%|██████    | 1981/3295 [20:47<13:47,  1.59it/s]
Epoch 0:  60%|██████    | 1982/3295 [20:48<13:47,  1.59it/s]
Epoch 0:  60%|██████    | 1982/3295 [20:48<13:47,  1.59it/s]
Epoch 0:  60%|██████    | 1983/3295 [20:49<13:46,  1.59it/s]
Epoch 0:  60%|██████    | 1983/3295 [20:49<13:46,  1.59it/s]
Epoch 0:  60%|██████    | 1984/3295 [20:49<13:45,  1.59it/s]
Epoch 0:  60%|██████    | 1984/3295 [20:49<13:45,  1.59it/s]
Epoch 0:  60%|██████    | 1985/3295 [20:50<13:45,  1.59it/s]
Epoch 0:  60%|██████    | 1985/3295 [20:50<13:45,  1.59it/s]
Epoch 0:  60%|██████    | 1986/3295 [20:50<13:44,  1.59it/s]
Epoch 0:  60%|██████    | 1986/3295 [20:50<13:44,  1.59it/s]
Epoch 0:  60%|██████    | 1987/3295 [20:50<13:43,  1.59it/s]
Epoch 0:  60%|██████    | 1987/3295 [20:50<13:43,  1.59it/s]
Epoch 0:  60%|██████    | 1988/3295 [20:52<13:43,  1.59it/s]
Epoch 0:  60%|██████    | 1988/3295 [20:52<13:43,  1.59it/s]
Epoch 0:  60%|██████    | 1989/3295 [20:52<13:42,  1.59it/s]
Epoch 0:  60%|██████    | 1989/3295 [20:52<13:42,  1.59it/s]
Epoch 0:  60%|██████    | 1990/3295 [20:52<13:41,  1.59it/s]
Epoch 0:  60%|██████    | 1990/3295 [20:52<13:41,  1.59it/s]
Epoch 0:  60%|██████    | 1991/3295 [20:54<13:41,  1.59it/s]
Epoch 0:  60%|██████    | 1991/3295 [20:54<13:41,  1.59it/s]
Epoch 0:  60%|██████    | 1992/3295 [20:54<13:40,  1.59it/s]
Epoch 0:  60%|██████    | 1992/3295 [20:54<13:40,  1.59it/s]
Epoch 0:  60%|██████    | 1993/3295 [20:54<13:39,  1.59it/s]
Epoch 0:  60%|██████    | 1993/3295 [20:54<13:39,  1.59it/s]
Epoch 0:  61%|██████    | 1994/3295 [20:56<13:39,  1.59it/s]
Epoch 0:  61%|██████    | 1994/3295 [20:56<13:39,  1.59it/s]
Epoch 0:  61%|██████    | 1995/3295 [20:56<13:38,  1.59it/s]
Epoch 0:  61%|██████    | 1995/3295 [20:56<13:38,  1.59it/s]
Epoch 0:  61%|██████    | 1996/3295 [20:56<13:37,  1.59it/s]
Epoch 0:  61%|██████    | 1996/3295 [20:56<13:37,  1.59it/s]
Epoch 0:  61%|██████    | 1997/3295 [20:58<13:37,  1.59it/s]
Epoch 0:  61%|██████    | 1997/3295 [20:58<13:37,  1.59it/s]
Epoch 0:  61%|██████    | 1998/3295 [20:58<13:36,  1.59it/s]
Epoch 0:  61%|██████    | 1998/3295 [20:58<13:36,  1.59it/s]
Epoch 0:  61%|██████    | 1999/3295 [20:58<13:35,  1.59it/s]
Epoch 0:  61%|██████    | 1999/3295 [20:58<13:35,  1.59it/s]
Epoch 0:  61%|██████    | 2000/3295 [21:00<13:35,  1.59it/s]
Epoch 0:  61%|██████    | 2000/3295 [21:00<13:35,  1.59it/s]
Epoch 0:  61%|██████    | 2001/3295 [21:00<13:34,  1.59it/s]
Epoch 0:  61%|██████    | 2001/3295 [21:00<13:34,  1.59it/s]
Epoch 0:  61%|██████    | 2002/3295 [21:00<13:33,  1.59it/s]
Epoch 0:  61%|██████    | 2002/3295 [21:00<13:33,  1.59it/s]
Epoch 0:  61%|██████    | 2003/3295 [21:01<13:33,  1.59it/s]
Epoch 0:  61%|██████    | 2003/3295 [21:01<13:33,  1.59it/s]
Epoch 0:  61%|██████    | 2004/3295 [21:01<13:32,  1.59it/s]
Epoch 0:  61%|██████    | 2004/3295 [21:01<13:32,  1.59it/s]
Epoch 0:  61%|██████    | 2005/3295 [21:01<13:31,  1.59it/s]
Epoch 0:  61%|██████    | 2005/3295 [21:01<13:31,  1.59it/s]
Epoch 0:  61%|██████    | 2006/3295 [21:03<13:32,  1.59it/s]
Epoch 0:  61%|██████    | 2006/3295 [21:03<13:32,  1.59it/s]
Epoch 0:  61%|██████    | 2007/3295 [21:03<13:30,  1.59it/s]
Epoch 0:  61%|██████    | 2007/3295 [21:03<13:30,  1.59it/s]
Epoch 0:  61%|██████    | 2008/3295 [21:03<13:29,  1.59it/s]
Epoch 0:  61%|██████    | 2008/3295 [21:03<13:29,  1.59it/s]
Epoch 0:  61%|██████    | 2009/3295 [21:05<13:30,  1.59it/s]
Epoch 0:  61%|██████    | 2009/3295 [21:05<13:30,  1.59it/s]
Epoch 0:  61%|██████    | 2010/3295 [21:05<13:29,  1.59it/s]
Epoch 0:  61%|██████    | 2010/3295 [21:05<13:29,  1.59it/s]
Epoch 0:  61%|██████    | 2011/3295 [21:05<13:28,  1.59it/s]
Epoch 0:  61%|██████    | 2011/3295 [21:05<13:28,  1.59it/s]
Epoch 0:  61%|██████    | 2012/3295 [21:07<13:28,  1.59it/s]
Epoch 0:  61%|██████    | 2012/3295 [21:07<13:28,  1.59it/s]
Epoch 0:  61%|██████    | 2013/3295 [21:07<13:27,  1.59it/s]
Epoch 0:  61%|██████    | 2013/3295 [21:07<13:27,  1.59it/s]
Epoch 0:  61%|██████    | 2014/3295 [21:07<13:26,  1.59it/s]
Epoch 0:  61%|██████    | 2014/3295 [21:07<13:26,  1.59it/s]
Epoch 0:  61%|██████    | 2015/3295 [21:09<13:26,  1.59it/s]
Epoch 0:  61%|██████    | 2015/3295 [21:09<13:26,  1.59it/s]
Epoch 0:  61%|██████    | 2016/3295 [21:09<13:25,  1.59it/s]
Epoch 0:  61%|██████    | 2016/3295 [21:09<13:25,  1.59it/s]
Epoch 0:  61%|██████    | 2017/3295 [21:09<13:24,  1.59it/s]
Epoch 0:  61%|██████    | 2017/3295 [21:09<13:24,  1.59it/s]
Epoch 0:  61%|██████    | 2018/3295 [21:11<13:24,  1.59it/s]
Epoch 0:  61%|██████    | 2018/3295 [21:11<13:24,  1.59it/s]
Epoch 0:  61%|██████▏   | 2019/3295 [21:11<13:23,  1.59it/s]
Epoch 0:  61%|██████▏   | 2019/3295 [21:11<13:23,  1.59it/s]
Epoch 0:  61%|██████▏   | 2020/3295 [21:11<13:22,  1.59it/s]
Epoch 0:  61%|██████▏   | 2020/3295 [21:11<13:22,  1.59it/s]
Epoch 0:  61%|██████▏   | 2021/3295 [21:13<13:22,  1.59it/s]
Epoch 0:  61%|██████▏   | 2021/3295 [21:13<13:22,  1.59it/s]
Epoch 0:  61%|██████▏   | 2022/3295 [21:13<13:21,  1.59it/s]
Epoch 0:  61%|██████▏   | 2022/3295 [21:13<13:21,  1.59it/s]
Epoch 0:  61%|██████▏   | 2023/3295 [21:13<13:20,  1.59it/s]
Epoch 0:  61%|██████▏   | 2023/3295 [21:13<13:20,  1.59it/s]
Epoch 0:  61%|██████▏   | 2024/3295 [21:15<13:20,  1.59it/s]
Epoch 0:  61%|██████▏   | 2024/3295 [21:15<13:20,  1.59it/s]
Epoch 0:  61%|██████▏   | 2025/3295 [21:15<13:19,  1.59it/s]
Epoch 0:  61%|██████▏   | 2025/3295 [21:15<13:19,  1.59it/s]
Epoch 0:  61%|██████▏   | 2026/3295 [21:15<13:18,  1.59it/s]
Epoch 0:  61%|██████▏   | 2026/3295 [21:15<13:18,  1.59it/s]
Epoch 0:  62%|██████▏   | 2027/3295 [21:16<13:18,  1.59it/s]
Epoch 0:  62%|██████▏   | 2027/3295 [21:16<13:18,  1.59it/s]
Epoch 0:  62%|██████▏   | 2028/3295 [21:16<13:17,  1.59it/s]
Epoch 0:  62%|██████▏   | 2028/3295 [21:16<13:17,  1.59it/s]
Epoch 0:  62%|██████▏   | 2029/3295 [21:16<13:16,  1.59it/s]
Epoch 0:  62%|██████▏   | 2029/3295 [21:16<13:16,  1.59it/s]
Epoch 0:  62%|██████▏   | 2030/3295 [21:18<13:16,  1.59it/s]
Epoch 0:  62%|██████▏   | 2030/3295 [21:18<13:16,  1.59it/s]
Epoch 0:  62%|██████▏   | 2031/3295 [21:18<13:15,  1.59it/s]
Epoch 0:  62%|██████▏   | 2031/3295 [21:18<13:15,  1.59it/s]
Epoch 0:  62%|██████▏   | 2032/3295 [21:18<13:14,  1.59it/s]
Epoch 0:  62%|██████▏   | 2032/3295 [21:18<13:14,  1.59it/s]
Epoch 0:  62%|██████▏   | 2033/3295 [21:20<13:14,  1.59it/s]
Epoch 0:  62%|██████▏   | 2033/3295 [21:20<13:14,  1.59it/s]
Epoch 0:  62%|██████▏   | 2034/3295 [21:20<13:13,  1.59it/s]
Epoch 0:  62%|██████▏   | 2034/3295 [21:20<13:13,  1.59it/s]
Epoch 0:  62%|██████▏   | 2035/3295 [21:20<13:12,  1.59it/s]
Epoch 0:  62%|██████▏   | 2035/3295 [21:20<13:12,  1.59it/s]
Epoch 0:  62%|██████▏   | 2036/3295 [21:22<13:12,  1.59it/s]
Epoch 0:  62%|██████▏   | 2036/3295 [21:22<13:12,  1.59it/s]
Epoch 0:  62%|██████▏   | 2037/3295 [21:22<13:11,  1.59it/s]
Epoch 0:  62%|██████▏   | 2037/3295 [21:22<13:11,  1.59it/s]
Epoch 0:  62%|██████▏   | 2038/3295 [21:22<13:10,  1.59it/s]
Epoch 0:  62%|██████▏   | 2038/3295 [21:22<13:10,  1.59it/s]
Epoch 0:  62%|██████▏   | 2039/3295 [21:24<13:11,  1.59it/s]
Epoch 0:  62%|██████▏   | 2039/3295 [21:24<13:11,  1.59it/s]
Epoch 0:  62%|██████▏   | 2040/3295 [21:24<13:10,  1.59it/s]
Epoch 0:  62%|██████▏   | 2040/3295 [21:24<13:10,  1.59it/s]
Epoch 0:  62%|██████▏   | 2041/3295 [21:24<13:09,  1.59it/s]
Epoch 0:  62%|██████▏   | 2041/3295 [21:24<13:09,  1.59it/s]
Epoch 0:  62%|██████▏   | 2042/3295 [21:25<13:09,  1.59it/s]
Epoch 0:  62%|██████▏   | 2042/3295 [21:25<13:09,  1.59it/s]
Epoch 0:  62%|██████▏   | 2043/3295 [21:26<13:08,  1.59it/s]
Epoch 0:  62%|██████▏   | 2043/3295 [21:26<13:08,  1.59it/s]
Epoch 0:  62%|██████▏   | 2044/3295 [21:26<13:07,  1.59it/s]
Epoch 0:  62%|██████▏   | 2044/3295 [21:26<13:07,  1.59it/s]
Epoch 0:  62%|██████▏   | 2045/3295 [21:27<13:07,  1.59it/s]
Epoch 0:  62%|██████▏   | 2045/3295 [21:27<13:07,  1.59it/s]
Epoch 0:  62%|██████▏   | 2046/3295 [21:28<13:06,  1.59it/s]
Epoch 0:  62%|██████▏   | 2046/3295 [21:28<13:06,  1.59it/s]
Epoch 0:  62%|██████▏   | 2047/3295 [21:28<13:05,  1.59it/s]
Epoch 0:  62%|██████▏   | 2047/3295 [21:28<13:05,  1.59it/s]
Epoch 0:  62%|██████▏   | 2048/3295 [21:29<13:05,  1.59it/s]
Epoch 0:  62%|██████▏   | 2048/3295 [21:29<13:05,  1.59it/s]
Epoch 0:  62%|██████▏   | 2049/3295 [21:29<13:04,  1.59it/s]
Epoch 0:  62%|██████▏   | 2049/3295 [21:29<13:04,  1.59it/s]
Epoch 0:  62%|██████▏   | 2050/3295 [21:29<13:03,  1.59it/s]
Epoch 0:  62%|██████▏   | 2050/3295 [21:29<13:03,  1.59it/s]
Epoch 0:  62%|██████▏   | 2051/3295 [21:31<13:03,  1.59it/s]
Epoch 0:  62%|██████▏   | 2051/3295 [21:31<13:03,  1.59it/s]
Epoch 0:  62%|██████▏   | 2052/3295 [21:31<13:02,  1.59it/s]
Epoch 0:  62%|██████▏   | 2052/3295 [21:31<13:02,  1.59it/s]
Epoch 0:  62%|██████▏   | 2053/3295 [21:31<13:01,  1.59it/s]
Epoch 0:  62%|██████▏   | 2053/3295 [21:31<13:01,  1.59it/s]
Epoch 0:  62%|██████▏   | 2054/3295 [21:33<13:01,  1.59it/s]
Epoch 0:  62%|██████▏   | 2054/3295 [21:33<13:01,  1.59it/s]
Epoch 0:  62%|██████▏   | 2055/3295 [21:33<13:00,  1.59it/s]
Epoch 0:  62%|██████▏   | 2055/3295 [21:33<13:00,  1.59it/s]
Epoch 0:  62%|██████▏   | 2056/3295 [21:33<12:59,  1.59it/s]
Epoch 0:  62%|██████▏   | 2056/3295 [21:33<12:59,  1.59it/s]
Epoch 0:  62%|██████▏   | 2057/3295 [21:35<12:59,  1.59it/s]
Epoch 0:  62%|██████▏   | 2057/3295 [21:35<12:59,  1.59it/s]
Epoch 0:  62%|██████▏   | 2058/3295 [21:35<12:58,  1.59it/s]
Epoch 0:  62%|██████▏   | 2058/3295 [21:35<12:58,  1.59it/s]
Epoch 0:  62%|██████▏   | 2059/3295 [21:35<12:57,  1.59it/s]
Epoch 0:  62%|██████▏   | 2059/3295 [21:35<12:57,  1.59it/s]
Epoch 0:  63%|██████▎   | 2060/3295 [21:37<12:57,  1.59it/s]
Epoch 0:  63%|██████▎   | 2060/3295 [21:37<12:57,  1.59it/s]
Epoch 0:  63%|██████▎   | 2061/3295 [21:37<12:56,  1.59it/s]
Epoch 0:  63%|██████▎   | 2061/3295 [21:37<12:56,  1.59it/s]
Epoch 0:  63%|██████▎   | 2062/3295 [21:37<12:55,  1.59it/s]
Epoch 0:  63%|██████▎   | 2062/3295 [21:37<12:55,  1.59it/s]
Epoch 0:  63%|██████▎   | 2063/3295 [21:39<12:55,  1.59it/s]
Epoch 0:  63%|██████▎   | 2063/3295 [21:39<12:55,  1.59it/s]
Epoch 0:  63%|██████▎   | 2064/3295 [21:39<12:54,  1.59it/s]
Epoch 0:  63%|██████▎   | 2064/3295 [21:39<12:54,  1.59it/s]
Epoch 0:  63%|██████▎   | 2065/3295 [21:39<12:53,  1.59it/s]
Epoch 0:  63%|██████▎   | 2065/3295 [21:39<12:53,  1.59it/s]
Epoch 0:  63%|██████▎   | 2066/3295 [21:41<12:54,  1.59it/s]
Epoch 0:  63%|██████▎   | 2066/3295 [21:41<12:54,  1.59it/s]
Epoch 0:  63%|██████▎   | 2067/3295 [21:41<12:53,  1.59it/s]
Epoch 0:  63%|██████▎   | 2067/3295 [21:41<12:53,  1.59it/s]
Epoch 0:  63%|██████▎   | 2068/3295 [21:41<12:52,  1.59it/s]
Epoch 0:  63%|██████▎   | 2068/3295 [21:41<12:52,  1.59it/s]
Epoch 0:  63%|██████▎   | 2069/3295 [21:43<12:52,  1.59it/s]
Epoch 0:  63%|██████▎   | 2069/3295 [21:43<12:52,  1.59it/s]
Epoch 0:  63%|██████▎   | 2070/3295 [21:43<12:51,  1.59it/s]
Epoch 0:  63%|██████▎   | 2070/3295 [21:43<12:51,  1.59it/s]
Epoch 0:  63%|██████▎   | 2071/3295 [21:43<12:50,  1.59it/s]
Epoch 0:  63%|██████▎   | 2071/3295 [21:43<12:50,  1.59it/s]
Epoch 0:  63%|██████▎   | 2072/3295 [21:45<12:50,  1.59it/s]
Epoch 0:  63%|██████▎   | 2072/3295 [21:45<12:50,  1.59it/s]
Epoch 0:  63%|██████▎   | 2073/3295 [21:45<12:49,  1.59it/s]
Epoch 0:  63%|██████▎   | 2073/3295 [21:45<12:49,  1.59it/s]
Epoch 0:  63%|██████▎   | 2074/3295 [21:45<12:48,  1.59it/s]
Epoch 0:  63%|██████▎   | 2074/3295 [21:45<12:48,  1.59it/s]
Epoch 0:  63%|██████▎   | 2075/3295 [21:46<12:48,  1.59it/s]
Epoch 0:  63%|██████▎   | 2075/3295 [21:46<12:48,  1.59it/s]
Epoch 0:  63%|██████▎   | 2076/3295 [21:46<12:47,  1.59it/s]
Epoch 0:  63%|██████▎   | 2076/3295 [21:46<12:47,  1.59it/s]
Epoch 0:  63%|██████▎   | 2077/3295 [21:46<12:46,  1.59it/s]
Epoch 0:  63%|██████▎   | 2077/3295 [21:46<12:46,  1.59it/s]
Epoch 0:  63%|██████▎   | 2078/3295 [21:48<12:46,  1.59it/s]
Epoch 0:  63%|██████▎   | 2078/3295 [21:48<12:46,  1.59it/s]
Epoch 0:  63%|██████▎   | 2079/3295 [21:48<12:45,  1.59it/s]
Epoch 0:  63%|██████▎   | 2079/3295 [21:48<12:45,  1.59it/s]
Epoch 0:  63%|██████▎   | 2080/3295 [21:48<12:44,  1.59it/s]
Epoch 0:  63%|██████▎   | 2080/3295 [21:48<12:44,  1.59it/s]
Epoch 0:  63%|██████▎   | 2081/3295 [21:50<12:44,  1.59it/s]
Epoch 0:  63%|██████▎   | 2081/3295 [21:50<12:44,  1.59it/s]
Epoch 0:  63%|██████▎   | 2082/3295 [21:50<12:43,  1.59it/s]
Epoch 0:  63%|██████▎   | 2082/3295 [21:50<12:43,  1.59it/s]
Epoch 0:  63%|██████▎   | 2083/3295 [21:50<12:42,  1.59it/s]
Epoch 0:  63%|██████▎   | 2083/3295 [21:50<12:42,  1.59it/s]
Epoch 0:  63%|██████▎   | 2084/3295 [21:52<12:42,  1.59it/s]
Epoch 0:  63%|██████▎   | 2084/3295 [21:52<12:42,  1.59it/s]
Epoch 0:  63%|██████▎   | 2085/3295 [21:52<12:41,  1.59it/s]
Epoch 0:  63%|██████▎   | 2085/3295 [21:52<12:41,  1.59it/s]
Epoch 0:  63%|██████▎   | 2086/3295 [21:52<12:40,  1.59it/s]
Epoch 0:  63%|██████▎   | 2086/3295 [21:52<12:40,  1.59it/s]
Epoch 0:  63%|██████▎   | 2087/3295 [21:54<12:40,  1.59it/s]
Epoch 0:  63%|██████▎   | 2087/3295 [21:54<12:40,  1.59it/s]
Epoch 0:  63%|██████▎   | 2088/3295 [21:54<12:39,  1.59it/s]
Epoch 0:  63%|██████▎   | 2088/3295 [21:54<12:39,  1.59it/s]
Epoch 0:  63%|██████▎   | 2089/3295 [21:54<12:38,  1.59it/s]
Epoch 0:  63%|██████▎   | 2089/3295 [21:54<12:38,  1.59it/s]
Epoch 0:  63%|██████▎   | 2090/3295 [21:56<12:38,  1.59it/s]
Epoch 0:  63%|██████▎   | 2090/3295 [21:56<12:38,  1.59it/s]
Epoch 0:  63%|██████▎   | 2091/3295 [21:56<12:37,  1.59it/s]
Epoch 0:  63%|██████▎   | 2091/3295 [21:56<12:37,  1.59it/s]
Epoch 0:  63%|██████▎   | 2092/3295 [21:56<12:36,  1.59it/s]
Epoch 0:  63%|██████▎   | 2092/3295 [21:56<12:36,  1.59it/s]
Epoch 0:  64%|██████▎   | 2093/3295 [21:58<12:37,  1.59it/s]
Epoch 0:  64%|██████▎   | 2093/3295 [21:58<12:37,  1.59it/s]
Epoch 0:  64%|██████▎   | 2094/3295 [21:58<12:36,  1.59it/s]
Epoch 0:  64%|██████▎   | 2094/3295 [21:58<12:36,  1.59it/s]
Epoch 0:  64%|██████▎   | 2095/3295 [21:58<12:35,  1.59it/s]
Epoch 0:  64%|██████▎   | 2095/3295 [21:58<12:35,  1.59it/s]
Epoch 0:  64%|██████▎   | 2096/3295 [22:00<12:35,  1.59it/s]
Epoch 0:  64%|██████▎   | 2096/3295 [22:00<12:35,  1.59it/s]
Epoch 0:  64%|██████▎   | 2097/3295 [22:00<12:34,  1.59it/s]
Epoch 0:  64%|██████▎   | 2097/3295 [22:00<12:34,  1.59it/s]
Epoch 0:  64%|██████▎   | 2098/3295 [22:00<12:33,  1.59it/s]
Epoch 0:  64%|██████▎   | 2098/3295 [22:00<12:33,  1.59it/s]
Epoch 0:  64%|██████▎   | 2099/3295 [22:01<12:33,  1.59it/s]
Epoch 0:  64%|██████▎   | 2099/3295 [22:01<12:33,  1.59it/s]
Epoch 0:  64%|██████▎   | 2100/3295 [22:01<12:32,  1.59it/s]
Epoch 0:  64%|██████▎   | 2100/3295 [22:01<12:32,  1.59it/s]
Epoch 0:  64%|██████▍   | 2101/3295 [22:01<12:31,  1.59it/s]
Epoch 0:  64%|██████▍   | 2101/3295 [22:01<12:31,  1.59it/s]
Epoch 0:  64%|██████▍   | 2102/3295 [22:03<12:31,  1.59it/s]
Epoch 0:  64%|██████▍   | 2102/3295 [22:03<12:31,  1.59it/s]
Epoch 0:  64%|██████▍   | 2103/3295 [22:03<12:30,  1.59it/s]
Epoch 0:  64%|██████▍   | 2103/3295 [22:03<12:30,  1.59it/s]
Epoch 0:  64%|██████▍   | 2104/3295 [22:03<12:29,  1.59it/s]
Epoch 0:  64%|██████▍   | 2104/3295 [22:03<12:29,  1.59it/s]
Epoch 0:  64%|██████▍   | 2105/3295 [22:05<12:29,  1.59it/s]
Epoch 0:  64%|██████▍   | 2105/3295 [22:05<12:29,  1.59it/s]
Epoch 0:  64%|██████▍   | 2106/3295 [22:05<12:28,  1.59it/s]
Epoch 0:  64%|██████▍   | 2106/3295 [22:05<12:28,  1.59it/s]
Epoch 0:  64%|██████▍   | 2107/3295 [22:05<12:27,  1.59it/s]
Epoch 0:  64%|██████▍   | 2107/3295 [22:05<12:27,  1.59it/s]
Epoch 0:  64%|██████▍   | 2108/3295 [22:07<12:27,  1.59it/s]
Epoch 0:  64%|██████▍   | 2108/3295 [22:07<12:27,  1.59it/s]
Epoch 0:  64%|██████▍   | 2109/3295 [22:07<12:26,  1.59it/s]
Epoch 0:  64%|██████▍   | 2109/3295 [22:07<12:26,  1.59it/s]
Epoch 0:  64%|██████▍   | 2110/3295 [22:07<12:25,  1.59it/s]
Epoch 0:  64%|██████▍   | 2110/3295 [22:07<12:25,  1.59it/s]
Epoch 0:  64%|██████▍   | 2111/3295 [22:09<12:25,  1.59it/s]
Epoch 0:  64%|██████▍   | 2111/3295 [22:09<12:25,  1.59it/s]
Epoch 0:  64%|██████▍   | 2112/3295 [22:09<12:24,  1.59it/s]
Epoch 0:  64%|██████▍   | 2112/3295 [22:09<12:24,  1.59it/s]
Epoch 0:  64%|██████▍   | 2113/3295 [22:09<12:23,  1.59it/s]
Epoch 0:  64%|██████▍   | 2113/3295 [22:09<12:23,  1.59it/s]
Epoch 0:  64%|██████▍   | 2114/3295 [22:11<12:23,  1.59it/s]
Epoch 0:  64%|██████▍   | 2114/3295 [22:11<12:23,  1.59it/s]
Epoch 0:  64%|██████▍   | 2115/3295 [22:11<12:22,  1.59it/s]
Epoch 0:  64%|██████▍   | 2115/3295 [22:11<12:22,  1.59it/s]
Epoch 0:  64%|██████▍   | 2116/3295 [22:11<12:21,  1.59it/s]
Epoch 0:  64%|██████▍   | 2116/3295 [22:11<12:21,  1.59it/s]
Epoch 0:  64%|██████▍   | 2117/3295 [22:12<12:21,  1.59it/s]
Epoch 0:  64%|██████▍   | 2117/3295 [22:12<12:21,  1.59it/s]
Epoch 0:  64%|██████▍   | 2118/3295 [22:12<12:20,  1.59it/s]
Epoch 0:  64%|██████▍   | 2118/3295 [22:12<12:20,  1.59it/s]
Epoch 0:  64%|██████▍   | 2119/3295 [22:12<12:19,  1.59it/s]
Epoch 0:  64%|██████▍   | 2119/3295 [22:12<12:19,  1.59it/s]
Epoch 0:  64%|██████▍   | 2120/3295 [22:14<12:19,  1.59it/s]
Epoch 0:  64%|██████▍   | 2120/3295 [22:14<12:19,  1.59it/s]
Epoch 0:  64%|██████▍   | 2121/3295 [22:14<12:18,  1.59it/s]
Epoch 0:  64%|██████▍   | 2121/3295 [22:14<12:18,  1.59it/s]
Epoch 0:  64%|██████▍   | 2122/3295 [22:14<12:17,  1.59it/s]
Epoch 0:  64%|██████▍   | 2122/3295 [22:14<12:17,  1.59it/s]
Epoch 0:  64%|██████▍   | 2123/3295 [22:16<12:17,  1.59it/s]
Epoch 0:  64%|██████▍   | 2123/3295 [22:16<12:17,  1.59it/s]
Epoch 0:  64%|██████▍   | 2124/3295 [22:16<12:16,  1.59it/s]
Epoch 0:  64%|██████▍   | 2124/3295 [22:16<12:16,  1.59it/s]
Epoch 0:  64%|██████▍   | 2125/3295 [22:16<12:15,  1.59it/s]
Epoch 0:  64%|██████▍   | 2125/3295 [22:16<12:15,  1.59it/s]
Epoch 0:  65%|██████▍   | 2126/3295 [22:18<12:15,  1.59it/s]
Epoch 0:  65%|██████▍   | 2126/3295 [22:18<12:15,  1.59it/s]
Epoch 0:  65%|██████▍   | 2127/3295 [22:18<12:14,  1.59it/s]
Epoch 0:  65%|██████▍   | 2127/3295 [22:18<12:14,  1.59it/s]
Epoch 0:  65%|██████▍   | 2128/3295 [22:18<12:13,  1.59it/s]
Epoch 0:  65%|██████▍   | 2128/3295 [22:18<12:13,  1.59it/s]
Epoch 0:  65%|██████▍   | 2129/3295 [22:19<12:13,  1.59it/s]
Epoch 0:  65%|██████▍   | 2129/3295 [22:19<12:13,  1.59it/s]
Epoch 0:  65%|██████▍   | 2130/3295 [22:19<12:12,  1.59it/s]
Epoch 0:  65%|██████▍   | 2130/3295 [22:19<12:12,  1.59it/s]
Epoch 0:  65%|██████▍   | 2131/3295 [22:20<12:11,  1.59it/s]
Epoch 0:  65%|██████▍   | 2131/3295 [22:20<12:11,  1.59it/s]
Epoch 0:  65%|██████▍   | 2132/3295 [22:21<12:11,  1.59it/s]
Epoch 0:  65%|██████▍   | 2132/3295 [22:21<12:11,  1.59it/s]
Epoch 0:  65%|██████▍   | 2133/3295 [22:21<12:10,  1.59it/s]
Epoch 0:  65%|██████▍   | 2133/3295 [22:21<12:10,  1.59it/s]
Epoch 0:  65%|██████▍   | 2134/3295 [22:21<12:09,  1.59it/s]
Epoch 0:  65%|██████▍   | 2134/3295 [22:21<12:09,  1.59it/s]
Epoch 0:  65%|██████▍   | 2135/3295 [22:23<12:09,  1.59it/s]
Epoch 0:  65%|██████▍   | 2135/3295 [22:23<12:09,  1.59it/s]
Epoch 0:  65%|██████▍   | 2136/3295 [22:23<12:09,  1.59it/s]
Epoch 0:  65%|██████▍   | 2136/3295 [22:23<12:09,  1.59it/s]
Epoch 0:  65%|██████▍   | 2137/3295 [22:23<12:08,  1.59it/s]
Epoch 0:  65%|██████▍   | 2137/3295 [22:23<12:08,  1.59it/s]
Epoch 0:  65%|██████▍   | 2138/3295 [22:25<12:08,  1.59it/s]
Epoch 0:  65%|██████▍   | 2138/3295 [22:25<12:08,  1.59it/s]
Epoch 0:  65%|██████▍   | 2139/3295 [22:25<12:07,  1.59it/s]
Epoch 0:  65%|██████▍   | 2139/3295 [22:25<12:07,  1.59it/s]
Epoch 0:  65%|██████▍   | 2140/3295 [22:25<12:06,  1.59it/s]
Epoch 0:  65%|██████▍   | 2140/3295 [22:25<12:06,  1.59it/s]
Epoch 0:  65%|██████▍   | 2141/3295 [22:27<12:06,  1.59it/s]
Epoch 0:  65%|██████▍   | 2141/3295 [22:27<12:06,  1.59it/s]
Epoch 0:  65%|██████▌   | 2142/3295 [22:27<12:05,  1.59it/s]
Epoch 0:  65%|██████▌   | 2142/3295 [22:27<12:05,  1.59it/s]
Epoch 0:  65%|██████▌   | 2143/3295 [22:27<12:04,  1.59it/s]
Epoch 0:  65%|██████▌   | 2143/3295 [22:27<12:04,  1.59it/s]
Epoch 0:  65%|██████▌   | 2144/3295 [22:29<12:04,  1.59it/s]
Epoch 0:  65%|██████▌   | 2144/3295 [22:29<12:04,  1.59it/s]
Epoch 0:  65%|██████▌   | 2145/3295 [22:29<12:03,  1.59it/s]
Epoch 0:  65%|██████▌   | 2145/3295 [22:29<12:03,  1.59it/s]
Epoch 0:  65%|██████▌   | 2146/3295 [22:29<12:02,  1.59it/s]
Epoch 0:  65%|██████▌   | 2146/3295 [22:29<12:02,  1.59it/s]
Epoch 0:  65%|██████▌   | 2147/3295 [22:30<12:02,  1.59it/s]
Epoch 0:  65%|██████▌   | 2147/3295 [22:30<12:02,  1.59it/s]
Epoch 0:  65%|██████▌   | 2148/3295 [22:30<12:01,  1.59it/s]
Epoch 0:  65%|██████▌   | 2148/3295 [22:30<12:01,  1.59it/s]
Epoch 0:  65%|██████▌   | 2149/3295 [22:30<12:00,  1.59it/s]
Epoch 0:  65%|██████▌   | 2149/3295 [22:30<12:00,  1.59it/s]
Epoch 0:  65%|██████▌   | 2150/3295 [22:32<12:00,  1.59it/s]
Epoch 0:  65%|██████▌   | 2150/3295 [22:32<12:00,  1.59it/s]
Epoch 0:  65%|██████▌   | 2151/3295 [22:32<11:59,  1.59it/s]
Epoch 0:  65%|██████▌   | 2151/3295 [22:32<11:59,  1.59it/s]
Epoch 0:  65%|██████▌   | 2152/3295 [22:32<11:58,  1.59it/s]
Epoch 0:  65%|██████▌   | 2152/3295 [22:32<11:58,  1.59it/s]
Epoch 0:  65%|██████▌   | 2153/3295 [22:34<11:58,  1.59it/s]
Epoch 0:  65%|██████▌   | 2153/3295 [22:34<11:58,  1.59it/s]
Epoch 0:  65%|██████▌   | 2154/3295 [22:34<11:57,  1.59it/s]
Epoch 0:  65%|██████▌   | 2154/3295 [22:34<11:57,  1.59it/s]
Epoch 0:  65%|██████▌   | 2155/3295 [22:34<11:56,  1.59it/s]
Epoch 0:  65%|██████▌   | 2155/3295 [22:34<11:56,  1.59it/s]
Epoch 0:  65%|██████▌   | 2156/3295 [22:36<11:56,  1.59it/s]
Epoch 0:  65%|██████▌   | 2156/3295 [22:36<11:56,  1.59it/s]
Epoch 0:  65%|██████▌   | 2157/3295 [22:36<11:55,  1.59it/s]
Epoch 0:  65%|██████▌   | 2157/3295 [22:36<11:55,  1.59it/s]
Epoch 0:  65%|██████▌   | 2158/3295 [22:36<11:54,  1.59it/s]
Epoch 0:  65%|██████▌   | 2158/3295 [22:36<11:54,  1.59it/s]
Epoch 0:  66%|██████▌   | 2159/3295 [22:38<11:54,  1.59it/s]
Epoch 0:  66%|██████▌   | 2159/3295 [22:38<11:54,  1.59it/s]
Epoch 0:  66%|██████▌   | 2160/3295 [22:38<11:53,  1.59it/s]
Epoch 0:  66%|██████▌   | 2160/3295 [22:38<11:53,  1.59it/s]
Epoch 0:  66%|██████▌   | 2161/3295 [22:38<11:52,  1.59it/s]
Epoch 0:  66%|██████▌   | 2161/3295 [22:38<11:52,  1.59it/s]
Epoch 0:  66%|██████▌   | 2162/3295 [22:39<11:52,  1.59it/s]
Epoch 0:  66%|██████▌   | 2162/3295 [22:39<11:52,  1.59it/s]
Epoch 0:  66%|██████▌   | 2163/3295 [22:39<11:51,  1.59it/s]
Epoch 0:  66%|██████▌   | 2163/3295 [22:39<11:51,  1.59it/s]
Epoch 0:  66%|██████▌   | 2164/3295 [22:39<11:50,  1.59it/s]
Epoch 0:  66%|██████▌   | 2164/3295 [22:39<11:50,  1.59it/s]
Epoch 0:  66%|██████▌   | 2165/3295 [22:41<11:50,  1.59it/s]
Epoch 0:  66%|██████▌   | 2165/3295 [22:41<11:50,  1.59it/s]
Epoch 0:  66%|██████▌   | 2166/3295 [22:41<11:49,  1.59it/s]
Epoch 0:  66%|██████▌   | 2166/3295 [22:41<11:49,  1.59it/s]
Epoch 0:  66%|██████▌   | 2167/3295 [22:41<11:48,  1.59it/s]
Epoch 0:  66%|██████▌   | 2167/3295 [22:41<11:48,  1.59it/s]
Epoch 0:  66%|██████▌   | 2168/3295 [22:43<11:48,  1.59it/s]
Epoch 0:  66%|██████▌   | 2168/3295 [22:43<11:48,  1.59it/s]
Epoch 0:  66%|██████▌   | 2169/3295 [22:43<11:47,  1.59it/s]
Epoch 0:  66%|██████▌   | 2169/3295 [22:43<11:47,  1.59it/s]
Epoch 0:  66%|██████▌   | 2170/3295 [22:43<11:46,  1.59it/s]
Epoch 0:  66%|██████▌   | 2170/3295 [22:43<11:46,  1.59it/s]
Epoch 0:  66%|██████▌   | 2171/3295 [22:45<11:46,  1.59it/s]
Epoch 0:  66%|██████▌   | 2171/3295 [22:45<11:46,  1.59it/s]
Epoch 0:  66%|██████▌   | 2172/3295 [22:45<11:45,  1.59it/s]
Epoch 0:  66%|██████▌   | 2172/3295 [22:45<11:45,  1.59it/s]
Epoch 0:  66%|██████▌   | 2173/3295 [22:45<11:44,  1.59it/s]
Epoch 0:  66%|██████▌   | 2173/3295 [22:45<11:44,  1.59it/s]
Epoch 0:  66%|██████▌   | 2174/3295 [22:47<11:45,  1.59it/s]
Epoch 0:  66%|██████▌   | 2174/3295 [22:47<11:45,  1.59it/s]
Epoch 0:  66%|██████▌   | 2175/3295 [22:47<11:44,  1.59it/s]
Epoch 0:  66%|██████▌   | 2175/3295 [22:47<11:44,  1.59it/s]
Epoch 0:  66%|██████▌   | 2176/3295 [22:47<11:43,  1.59it/s]
Epoch 0:  66%|██████▌   | 2176/3295 [22:47<11:43,  1.59it/s]
Epoch 0:  66%|██████▌   | 2177/3295 [22:49<11:43,  1.59it/s]
Epoch 0:  66%|██████▌   | 2177/3295 [22:49<11:43,  1.59it/s]
Epoch 0:  66%|██████▌   | 2178/3295 [22:49<11:42,  1.59it/s]
Epoch 0:  66%|██████▌   | 2178/3295 [22:49<11:42,  1.59it/s]
Epoch 0:  66%|██████▌   | 2179/3295 [22:49<11:41,  1.59it/s]
Epoch 0:  66%|██████▌   | 2179/3295 [22:49<11:41,  1.59it/s]
Epoch 0:  66%|██████▌   | 2180/3295 [22:50<11:41,  1.59it/s]
Epoch 0:  66%|██████▌   | 2180/3295 [22:50<11:41,  1.59it/s]
Epoch 0:  66%|██████▌   | 2181/3295 [22:50<11:40,  1.59it/s]
Epoch 0:  66%|██████▌   | 2181/3295 [22:50<11:40,  1.59it/s]
Epoch 0:  66%|██████▌   | 2182/3295 [22:50<11:39,  1.59it/s]
Epoch 0:  66%|██████▌   | 2182/3295 [22:50<11:39,  1.59it/s]
Epoch 0:  66%|██████▋   | 2183/3295 [22:52<11:39,  1.59it/s]
Epoch 0:  66%|██████▋   | 2183/3295 [22:52<11:39,  1.59it/s]
Epoch 0:  66%|██████▋   | 2184/3295 [22:52<11:38,  1.59it/s]
Epoch 0:  66%|██████▋   | 2184/3295 [22:52<11:38,  1.59it/s]
Epoch 0:  66%|██████▋   | 2185/3295 [22:52<11:37,  1.59it/s]
Epoch 0:  66%|██████▋   | 2185/3295 [22:52<11:37,  1.59it/s]
Epoch 0:  66%|██████▋   | 2186/3295 [22:54<11:37,  1.59it/s]
Epoch 0:  66%|██████▋   | 2186/3295 [22:54<11:37,  1.59it/s]
Epoch 0:  66%|██████▋   | 2187/3295 [22:54<11:36,  1.59it/s]
Epoch 0:  66%|██████▋   | 2187/3295 [22:54<11:36,  1.59it/s]
Epoch 0:  66%|██████▋   | 2188/3295 [22:54<11:35,  1.59it/s]
Epoch 0:  66%|██████▋   | 2188/3295 [22:54<11:35,  1.59it/s]
Epoch 0:  66%|██████▋   | 2189/3295 [22:56<11:35,  1.59it/s]
Epoch 0:  66%|██████▋   | 2189/3295 [22:56<11:35,  1.59it/s]
Epoch 0:  66%|██████▋   | 2190/3295 [22:56<11:34,  1.59it/s]
Epoch 0:  66%|██████▋   | 2190/3295 [22:56<11:34,  1.59it/s]
Epoch 0:  66%|██████▋   | 2191/3295 [22:56<11:33,  1.59it/s]
Epoch 0:  66%|██████▋   | 2191/3295 [22:56<11:33,  1.59it/s]
Epoch 0:  67%|██████▋   | 2192/3295 [22:58<11:33,  1.59it/s]
Epoch 0:  67%|██████▋   | 2192/3295 [22:58<11:33,  1.59it/s]
Epoch 0:  67%|██████▋   | 2193/3295 [22:58<11:32,  1.59it/s]
Epoch 0:  67%|██████▋   | 2193/3295 [22:58<11:32,  1.59it/s]
Epoch 0:  67%|██████▋   | 2194/3295 [22:58<11:31,  1.59it/s]
Epoch 0:  67%|██████▋   | 2194/3295 [22:58<11:31,  1.59it/s]
Epoch 0:  67%|██████▋   | 2195/3295 [23:00<11:31,  1.59it/s]
Epoch 0:  67%|██████▋   | 2195/3295 [23:00<11:31,  1.59it/s]
Epoch 0:  67%|██████▋   | 2196/3295 [23:00<11:30,  1.59it/s]
Epoch 0:  67%|██████▋   | 2196/3295 [23:00<11:30,  1.59it/s]
Epoch 0:  67%|██████▋   | 2197/3295 [23:00<11:29,  1.59it/s]
Epoch 0:  67%|██████▋   | 2197/3295 [23:00<11:29,  1.59it/s]
Epoch 0:  67%|██████▋   | 2198/3295 [23:02<11:29,  1.59it/s]
Epoch 0:  67%|██████▋   | 2198/3295 [23:02<11:29,  1.59it/s]
Epoch 0:  67%|██████▋   | 2199/3295 [23:02<11:28,  1.59it/s]
Epoch 0:  67%|██████▋   | 2199/3295 [23:02<11:28,  1.59it/s]
Epoch 0:  67%|██████▋   | 2200/3295 [23:02<11:27,  1.59it/s]
Epoch 0:  67%|██████▋   | 2200/3295 [23:02<11:27,  1.59it/s]
Epoch 0:  67%|██████▋   | 2201/3295 [23:03<11:27,  1.59it/s]
Epoch 0:  67%|██████▋   | 2201/3295 [23:03<11:27,  1.59it/s]
Epoch 0:  67%|██████▋   | 2202/3295 [23:04<11:26,  1.59it/s]
Epoch 0:  67%|██████▋   | 2202/3295 [23:04<11:26,  1.59it/s]
Epoch 0:  67%|██████▋   | 2203/3295 [23:04<11:26,  1.59it/s]
Epoch 0:  67%|██████▋   | 2203/3295 [23:04<11:26,  1.59it/s]
Epoch 0:  67%|██████▋   | 2204/3295 [23:05<11:26,  1.59it/s]
Epoch 0:  67%|██████▋   | 2204/3295 [23:05<11:26,  1.59it/s]
Epoch 0:  67%|██████▋   | 2205/3295 [23:05<11:25,  1.59it/s]
Epoch 0:  67%|██████▋   | 2205/3295 [23:05<11:25,  1.59it/s]
Epoch 0:  67%|██████▋   | 2206/3295 [23:05<11:24,  1.59it/s]
Epoch 0:  67%|██████▋   | 2206/3295 [23:05<11:24,  1.59it/s]
Epoch 0:  67%|██████▋   | 2207/3295 [23:07<11:24,  1.59it/s]
Epoch 0:  67%|██████▋   | 2207/3295 [23:07<11:24,  1.59it/s]
Epoch 0:  67%|██████▋   | 2208/3295 [23:07<11:23,  1.59it/s]
Epoch 0:  67%|██████▋   | 2208/3295 [23:07<11:23,  1.59it/s]
Epoch 0:  67%|██████▋   | 2209/3295 [23:07<11:22,  1.59it/s]
Epoch 0:  67%|██████▋   | 2209/3295 [23:07<11:22,  1.59it/s]
Epoch 0:  67%|██████▋   | 2210/3295 [23:09<11:22,  1.59it/s]
Epoch 0:  67%|██████▋   | 2210/3295 [23:09<11:22,  1.59it/s]
Epoch 0:  67%|██████▋   | 2211/3295 [23:09<11:21,  1.59it/s]
Epoch 0:  67%|██████▋   | 2211/3295 [23:09<11:21,  1.59it/s]
Epoch 0:  67%|██████▋   | 2212/3295 [23:09<11:20,  1.59it/s]
Epoch 0:  67%|██████▋   | 2212/3295 [23:09<11:20,  1.59it/s]
Epoch 0:  67%|██████▋   | 2213/3295 [23:11<11:20,  1.59it/s]
Epoch 0:  67%|██████▋   | 2213/3295 [23:11<11:20,  1.59it/s]
Epoch 0:  67%|██████▋   | 2214/3295 [23:11<11:19,  1.59it/s]
Epoch 0:  67%|██████▋   | 2214/3295 [23:11<11:19,  1.59it/s]
Epoch 0:  67%|██████▋   | 2215/3295 [23:11<11:18,  1.59it/s]
Epoch 0:  67%|██████▋   | 2215/3295 [23:11<11:18,  1.59it/s]
Epoch 0:  67%|██████▋   | 2216/3295 [23:13<11:18,  1.59it/s]
Epoch 0:  67%|██████▋   | 2216/3295 [23:13<11:18,  1.59it/s]
Epoch 0:  67%|██████▋   | 2217/3295 [23:13<11:17,  1.59it/s]
Epoch 0:  67%|██████▋   | 2217/3295 [23:13<11:17,  1.59it/s]
Epoch 0:  67%|██████▋   | 2218/3295 [23:13<11:16,  1.59it/s]
Epoch 0:  67%|██████▋   | 2218/3295 [23:13<11:16,  1.59it/s]
Epoch 0:  67%|██████▋   | 2219/3295 [23:14<11:16,  1.59it/s]
Epoch 0:  67%|██████▋   | 2219/3295 [23:14<11:16,  1.59it/s]
Epoch 0:  67%|██████▋   | 2220/3295 [23:14<11:15,  1.59it/s]
Epoch 0:  67%|██████▋   | 2220/3295 [23:14<11:15,  1.59it/s]
Epoch 0:  67%|██████▋   | 2221/3295 [23:14<11:14,  1.59it/s]
Epoch 0:  67%|██████▋   | 2221/3295 [23:14<11:14,  1.59it/s]
Epoch 0:  67%|██████▋   | 2222/3295 [23:16<11:14,  1.59it/s]
Epoch 0:  67%|██████▋   | 2222/3295 [23:16<11:14,  1.59it/s]
Epoch 0:  67%|██████▋   | 2223/3295 [23:16<11:13,  1.59it/s]
Epoch 0:  67%|██████▋   | 2223/3295 [23:16<11:13,  1.59it/s]
Epoch 0:  67%|██████▋   | 2224/3295 [23:16<11:12,  1.59it/s]
Epoch 0:  67%|██████▋   | 2224/3295 [23:16<11:12,  1.59it/s]
Epoch 0:  68%|██████▊   | 2225/3295 [23:18<11:12,  1.59it/s]
Epoch 0:  68%|██████▊   | 2225/3295 [23:18<11:12,  1.59it/s]
Epoch 0:  68%|██████▊   | 2226/3295 [23:18<11:11,  1.59it/s]
Epoch 0:  68%|██████▊   | 2226/3295 [23:18<11:11,  1.59it/s]
Epoch 0:  68%|██████▊   | 2227/3295 [23:18<11:10,  1.59it/s]
Epoch 0:  68%|██████▊   | 2227/3295 [23:18<11:10,  1.59it/s]
Epoch 0:  68%|██████▊   | 2228/3295 [23:20<11:10,  1.59it/s]
Epoch 0:  68%|██████▊   | 2228/3295 [23:20<11:10,  1.59it/s]
Epoch 0:  68%|██████▊   | 2229/3295 [23:20<11:09,  1.59it/s]
Epoch 0:  68%|██████▊   | 2229/3295 [23:20<11:09,  1.59it/s]
Epoch 0:  68%|██████▊   | 2230/3295 [23:20<11:08,  1.59it/s]
Epoch 0:  68%|██████▊   | 2230/3295 [23:20<11:08,  1.59it/s]
Epoch 0:  68%|██████▊   | 2231/3295 [23:22<11:08,  1.59it/s]
Epoch 0:  68%|██████▊   | 2231/3295 [23:22<11:08,  1.59it/s]
Epoch 0:  68%|██████▊   | 2232/3295 [23:22<11:07,  1.59it/s]
Epoch 0:  68%|██████▊   | 2232/3295 [23:22<11:07,  1.59it/s]
Epoch 0:  68%|██████▊   | 2233/3295 [23:22<11:06,  1.59it/s]
Epoch 0:  68%|██████▊   | 2233/3295 [23:22<11:06,  1.59it/s]
Epoch 0:  68%|██████▊   | 2234/3295 [23:24<11:06,  1.59it/s]
Epoch 0:  68%|██████▊   | 2234/3295 [23:24<11:06,  1.59it/s]
Epoch 0:  68%|██████▊   | 2235/3295 [23:24<11:06,  1.59it/s]
Epoch 0:  68%|██████▊   | 2235/3295 [23:24<11:06,  1.59it/s]
Epoch 0:  68%|██████▊   | 2236/3295 [23:24<11:05,  1.59it/s]
Epoch 0:  68%|██████▊   | 2236/3295 [23:24<11:05,  1.59it/s]
Epoch 0:  68%|██████▊   | 2237/3295 [23:26<11:05,  1.59it/s]
Epoch 0:  68%|██████▊   | 2237/3295 [23:26<11:05,  1.59it/s]
Epoch 0:  68%|██████▊   | 2238/3295 [23:26<11:04,  1.59it/s]
Epoch 0:  68%|██████▊   | 2238/3295 [23:26<11:04,  1.59it/s]
Epoch 0:  68%|██████▊   | 2239/3295 [23:26<11:03,  1.59it/s]
Epoch 0:  68%|██████▊   | 2239/3295 [23:26<11:03,  1.59it/s]
Epoch 0:  68%|██████▊   | 2240/3295 [23:27<11:03,  1.59it/s]
Epoch 0:  68%|██████▊   | 2240/3295 [23:27<11:03,  1.59it/s]
Epoch 0:  68%|██████▊   | 2241/3295 [23:27<11:02,  1.59it/s]
Epoch 0:  68%|██████▊   | 2241/3295 [23:27<11:02,  1.59it/s]
Epoch 0:  68%|██████▊   | 2242/3295 [23:27<11:01,  1.59it/s]
Epoch 0:  68%|██████▊   | 2242/3295 [23:27<11:01,  1.59it/s]
Epoch 0:  68%|██████▊   | 2243/3295 [23:29<11:01,  1.59it/s]
Epoch 0:  68%|██████▊   | 2243/3295 [23:29<11:01,  1.59it/s]
Epoch 0:  68%|██████▊   | 2244/3295 [23:29<11:00,  1.59it/s]
Epoch 0:  68%|██████▊   | 2244/3295 [23:29<11:00,  1.59it/s]
Epoch 0:  68%|██████▊   | 2245/3295 [23:29<10:59,  1.59it/s]
Epoch 0:  68%|██████▊   | 2245/3295 [23:29<10:59,  1.59it/s]
Epoch 0:  68%|██████▊   | 2246/3295 [23:31<10:59,  1.59it/s]
Epoch 0:  68%|██████▊   | 2246/3295 [23:31<10:59,  1.59it/s]
Epoch 0:  68%|██████▊   | 2247/3295 [23:31<10:58,  1.59it/s]
Epoch 0:  68%|██████▊   | 2247/3295 [23:31<10:58,  1.59it/s]
Epoch 0:  68%|██████▊   | 2248/3295 [23:31<10:57,  1.59it/s]
Epoch 0:  68%|██████▊   | 2248/3295 [23:31<10:57,  1.59it/s]
Epoch 0:  68%|██████▊   | 2249/3295 [23:33<10:57,  1.59it/s]
Epoch 0:  68%|██████▊   | 2249/3295 [23:33<10:57,  1.59it/s]
Epoch 0:  68%|██████▊   | 2250/3295 [23:33<10:56,  1.59it/s]
Epoch 0:  68%|██████▊   | 2250/3295 [23:33<10:56,  1.59it/s]
Epoch 0:  68%|██████▊   | 2251/3295 [23:33<10:55,  1.59it/s]
Epoch 0:  68%|██████▊   | 2251/3295 [23:33<10:55,  1.59it/s]
Epoch 0:  68%|██████▊   | 2252/3295 [23:35<10:55,  1.59it/s]
Epoch 0:  68%|██████▊   | 2252/3295 [23:35<10:55,  1.59it/s]
Epoch 0:  68%|██████▊   | 2253/3295 [23:35<10:54,  1.59it/s]
Epoch 0:  68%|██████▊   | 2253/3295 [23:35<10:54,  1.59it/s]
Epoch 0:  68%|██████▊   | 2254/3295 [23:35<10:53,  1.59it/s]
Epoch 0:  68%|██████▊   | 2254/3295 [23:35<10:53,  1.59it/s]
Epoch 0:  68%|██████▊   | 2255/3295 [23:36<10:53,  1.59it/s]
Epoch 0:  68%|██████▊   | 2255/3295 [23:36<10:53,  1.59it/s]
Epoch 0:  68%|██████▊   | 2256/3295 [23:37<10:52,  1.59it/s]
Epoch 0:  68%|██████▊   | 2256/3295 [23:37<10:52,  1.59it/s]
Epoch 0:  68%|██████▊   | 2257/3295 [23:37<10:51,  1.59it/s]
Epoch 0:  68%|██████▊   | 2257/3295 [23:37<10:51,  1.59it/s]
Epoch 0:  69%|██████▊   | 2258/3295 [23:38<10:51,  1.59it/s]
Epoch 0:  69%|██████▊   | 2258/3295 [23:38<10:51,  1.59it/s]
Epoch 0:  69%|██████▊   | 2259/3295 [23:38<10:50,  1.59it/s]
Epoch 0:  69%|██████▊   | 2259/3295 [23:38<10:50,  1.59it/s]
Epoch 0:  69%|██████▊   | 2260/3295 [23:38<10:49,  1.59it/s]
Epoch 0:  69%|██████▊   | 2260/3295 [23:38<10:49,  1.59it/s]
Epoch 0:  69%|██████▊   | 2261/3295 [23:40<10:49,  1.59it/s]
Epoch 0:  69%|██████▊   | 2261/3295 [23:40<10:49,  1.59it/s]
Epoch 0:  69%|██████▊   | 2262/3295 [23:40<10:48,  1.59it/s]
Epoch 0:  69%|██████▊   | 2262/3295 [23:40<10:48,  1.59it/s]
Epoch 0:  69%|██████▊   | 2263/3295 [23:40<10:47,  1.59it/s]
Epoch 0:  69%|██████▊   | 2263/3295 [23:40<10:47,  1.59it/s]
Epoch 0:  69%|██████▊   | 2264/3295 [23:42<10:47,  1.59it/s]
Epoch 0:  69%|██████▊   | 2264/3295 [23:42<10:47,  1.59it/s]
Epoch 0:  69%|██████▊   | 2265/3295 [23:42<10:46,  1.59it/s]
Epoch 0:  69%|██████▊   | 2265/3295 [23:42<10:46,  1.59it/s]
Epoch 0:  69%|██████▉   | 2266/3295 [23:42<10:46,  1.59it/s]
Epoch 0:  69%|██████▉   | 2266/3295 [23:42<10:46,  1.59it/s]
Epoch 0:  69%|██████▉   | 2267/3295 [23:44<10:45,  1.59it/s]
Epoch 0:  69%|██████▉   | 2267/3295 [23:44<10:45,  1.59it/s]
Epoch 0:  69%|██████▉   | 2268/3295 [23:44<10:44,  1.59it/s]
Epoch 0:  69%|██████▉   | 2268/3295 [23:44<10:44,  1.59it/s]
Epoch 0:  69%|██████▉   | 2269/3295 [23:44<10:44,  1.59it/s]
Epoch 0:  69%|██████▉   | 2269/3295 [23:44<10:44,  1.59it/s]
Epoch 0:  69%|██████▉   | 2270/3295 [23:46<10:43,  1.59it/s]
Epoch 0:  69%|██████▉   | 2270/3295 [23:46<10:43,  1.59it/s]
Epoch 0:  69%|██████▉   | 2271/3295 [23:46<10:43,  1.59it/s]
Epoch 0:  69%|██████▉   | 2271/3295 [23:46<10:43,  1.59it/s]
Epoch 0:  69%|██████▉   | 2272/3295 [23:46<10:42,  1.59it/s]
Epoch 0:  69%|██████▉   | 2272/3295 [23:46<10:42,  1.59it/s]
Epoch 0:  69%|██████▉   | 2273/3295 [23:47<10:42,  1.59it/s]
Epoch 0:  69%|██████▉   | 2273/3295 [23:47<10:42,  1.59it/s]
Epoch 0:  69%|██████▉   | 2274/3295 [23:48<10:41,  1.59it/s]
Epoch 0:  69%|██████▉   | 2274/3295 [23:48<10:41,  1.59it/s]
Epoch 0:  69%|██████▉   | 2275/3295 [23:48<10:40,  1.59it/s]
Epoch 0:  69%|██████▉   | 2275/3295 [23:48<10:40,  1.59it/s]
Epoch 0:  69%|██████▉   | 2276/3295 [23:49<10:40,  1.59it/s]
Epoch 0:  69%|██████▉   | 2276/3295 [23:49<10:40,  1.59it/s]
Epoch 0:  69%|██████▉   | 2277/3295 [23:49<10:39,  1.59it/s]
Epoch 0:  69%|██████▉   | 2277/3295 [23:49<10:39,  1.59it/s]
Epoch 0:  69%|██████▉   | 2278/3295 [23:49<10:38,  1.59it/s]
Epoch 0:  69%|██████▉   | 2278/3295 [23:49<10:38,  1.59it/s]
Epoch 0:  69%|██████▉   | 2279/3295 [23:51<10:38,  1.59it/s]
Epoch 0:  69%|██████▉   | 2279/3295 [23:51<10:38,  1.59it/s]
Epoch 0:  69%|██████▉   | 2280/3295 [23:51<10:37,  1.59it/s]
Epoch 0:  69%|██████▉   | 2280/3295 [23:51<10:37,  1.59it/s]
Epoch 0:  69%|██████▉   | 2281/3295 [23:51<10:36,  1.59it/s]
Epoch 0:  69%|██████▉   | 2281/3295 [23:51<10:36,  1.59it/s]
Epoch 0:  69%|██████▉   | 2282/3295 [23:53<10:36,  1.59it/s]
Epoch 0:  69%|██████▉   | 2282/3295 [23:53<10:36,  1.59it/s]
Epoch 0:  69%|██████▉   | 2283/3295 [23:53<10:35,  1.59it/s]
Epoch 0:  69%|██████▉   | 2283/3295 [23:53<10:35,  1.59it/s]
Epoch 0:  69%|██████▉   | 2284/3295 [23:53<10:34,  1.59it/s]
Epoch 0:  69%|██████▉   | 2284/3295 [23:53<10:34,  1.59it/s]
Epoch 0:  69%|██████▉   | 2285/3295 [23:55<10:34,  1.59it/s]
Epoch 0:  69%|██████▉   | 2285/3295 [23:55<10:34,  1.59it/s]
Epoch 0:  69%|██████▉   | 2286/3295 [23:55<10:33,  1.59it/s]
Epoch 0:  69%|██████▉   | 2286/3295 [23:55<10:33,  1.59it/s]
Epoch 0:  69%|██████▉   | 2287/3295 [23:55<10:32,  1.59it/s]
Epoch 0:  69%|██████▉   | 2287/3295 [23:55<10:32,  1.59it/s]
Epoch 0:  69%|██████▉   | 2288/3295 [23:57<10:32,  1.59it/s]
Epoch 0:  69%|██████▉   | 2288/3295 [23:57<10:32,  1.59it/s]
Epoch 0:  69%|██████▉   | 2289/3295 [23:57<10:31,  1.59it/s]
Epoch 0:  69%|██████▉   | 2289/3295 [23:57<10:31,  1.59it/s]
Epoch 0:  69%|██████▉   | 2290/3295 [23:57<10:30,  1.59it/s]
Epoch 0:  69%|██████▉   | 2290/3295 [23:57<10:30,  1.59it/s]
Epoch 0:  70%|██████▉   | 2291/3295 [23:59<10:30,  1.59it/s]
Epoch 0:  70%|██████▉   | 2291/3295 [23:59<10:30,  1.59it/s]
Epoch 0:  70%|██████▉   | 2292/3295 [23:59<10:29,  1.59it/s]
Epoch 0:  70%|██████▉   | 2292/3295 [23:59<10:29,  1.59it/s]
Epoch 0:  70%|██████▉   | 2293/3295 [23:59<10:28,  1.59it/s]
Epoch 0:  70%|██████▉   | 2293/3295 [23:59<10:28,  1.59it/s]
Epoch 0:  70%|██████▉   | 2294/3295 [24:01<10:28,  1.59it/s]
Epoch 0:  70%|██████▉   | 2294/3295 [24:01<10:28,  1.59it/s]
Epoch 0:  70%|██████▉   | 2295/3295 [24:01<10:27,  1.59it/s]
Epoch 0:  70%|██████▉   | 2295/3295 [24:01<10:27,  1.59it/s]
Epoch 0:  70%|██████▉   | 2296/3295 [24:01<10:27,  1.59it/s]
Epoch 0:  70%|██████▉   | 2296/3295 [24:01<10:27,  1.59it/s]
Epoch 0:  70%|██████▉   | 2297/3295 [24:02<10:26,  1.59it/s]
Epoch 0:  70%|██████▉   | 2297/3295 [24:02<10:26,  1.59it/s]
Epoch 0:  70%|██████▉   | 2298/3295 [24:02<10:26,  1.59it/s]
Epoch 0:  70%|██████▉   | 2298/3295 [24:02<10:26,  1.59it/s]
Epoch 0:  70%|██████▉   | 2299/3295 [24:03<10:25,  1.59it/s]
Epoch 0:  70%|██████▉   | 2299/3295 [24:03<10:25,  1.59it/s]
Epoch 0:  70%|██████▉   | 2300/3295 [24:04<10:25,  1.59it/s]
Epoch 0:  70%|██████▉   | 2300/3295 [24:04<10:25,  1.59it/s]
Epoch 0:  70%|██████▉   | 2301/3295 [24:04<10:24,  1.59it/s]
Epoch 0:  70%|██████▉   | 2301/3295 [24:04<10:24,  1.59it/s]
Epoch 0:  70%|██████▉   | 2302/3295 [24:04<10:23,  1.59it/s]
Epoch 0:  70%|██████▉   | 2302/3295 [24:04<10:23,  1.59it/s]
Epoch 0:  70%|██████▉   | 2303/3295 [24:06<10:23,  1.59it/s]
Epoch 0:  70%|██████▉   | 2303/3295 [24:06<10:23,  1.59it/s]
Epoch 0:  70%|██████▉   | 2304/3295 [24:06<10:22,  1.59it/s]
Epoch 0:  70%|██████▉   | 2304/3295 [24:06<10:22,  1.59it/s]
Epoch 0:  70%|██████▉   | 2305/3295 [24:06<10:21,  1.59it/s]
Epoch 0:  70%|██████▉   | 2305/3295 [24:06<10:21,  1.59it/s]
Epoch 0:  70%|██████▉   | 2306/3295 [24:08<10:21,  1.59it/s]
Epoch 0:  70%|██████▉   | 2306/3295 [24:08<10:21,  1.59it/s]
Epoch 0:  70%|███████   | 2307/3295 [24:08<10:20,  1.59it/s]
Epoch 0:  70%|███████   | 2307/3295 [24:08<10:20,  1.59it/s]
Epoch 0:  70%|███████   | 2308/3295 [24:08<10:19,  1.59it/s]
Epoch 0:  70%|███████   | 2308/3295 [24:08<10:19,  1.59it/s]
Epoch 0:  70%|███████   | 2309/3295 [24:10<10:19,  1.59it/s]
Epoch 0:  70%|███████   | 2309/3295 [24:10<10:19,  1.59it/s]
Epoch 0:  70%|███████   | 2310/3295 [24:10<10:18,  1.59it/s]
Epoch 0:  70%|███████   | 2310/3295 [24:10<10:18,  1.59it/s]
Epoch 0:  70%|███████   | 2311/3295 [24:10<10:17,  1.59it/s]
Epoch 0:  70%|███████   | 2311/3295 [24:10<10:17,  1.59it/s]
Epoch 0:  70%|███████   | 2312/3295 [24:12<10:17,  1.59it/s]
Epoch 0:  70%|███████   | 2312/3295 [24:12<10:17,  1.59it/s]
Epoch 0:  70%|███████   | 2313/3295 [24:12<10:16,  1.59it/s]
Epoch 0:  70%|███████   | 2313/3295 [24:12<10:16,  1.59it/s]
Epoch 0:  70%|███████   | 2314/3295 [24:12<10:15,  1.59it/s]
Epoch 0:  70%|███████   | 2314/3295 [24:12<10:15,  1.59it/s]
Epoch 0:  70%|███████   | 2315/3295 [24:14<10:15,  1.59it/s]
Epoch 0:  70%|███████   | 2315/3295 [24:14<10:15,  1.59it/s]
Epoch 0:  70%|███████   | 2316/3295 [24:14<10:14,  1.59it/s]
Epoch 0:  70%|███████   | 2316/3295 [24:14<10:14,  1.59it/s]
Epoch 0:  70%|███████   | 2317/3295 [24:14<10:13,  1.59it/s]
Epoch 0:  70%|███████   | 2317/3295 [24:14<10:13,  1.59it/s]
Epoch 0:  70%|███████   | 2318/3295 [24:15<10:13,  1.59it/s]
Epoch 0:  70%|███████   | 2318/3295 [24:15<10:13,  1.59it/s]
Epoch 0:  70%|███████   | 2319/3295 [24:15<10:12,  1.59it/s]
Epoch 0:  70%|███████   | 2319/3295 [24:15<10:12,  1.59it/s]
Epoch 0:  70%|███████   | 2320/3295 [24:16<10:11,  1.59it/s]
Epoch 0:  70%|███████   | 2320/3295 [24:16<10:11,  1.59it/s]
Epoch 0:  70%|███████   | 2321/3295 [24:17<10:11,  1.59it/s]
Epoch 0:  70%|███████   | 2321/3295 [24:17<10:11,  1.59it/s]
Epoch 0:  70%|███████   | 2322/3295 [24:17<10:10,  1.59it/s]
Epoch 0:  70%|███████   | 2322/3295 [24:17<10:10,  1.59it/s]
Epoch 0:  71%|███████   | 2323/3295 [24:17<10:10,  1.59it/s]
Epoch 0:  71%|███████   | 2323/3295 [24:17<10:10,  1.59it/s]
Epoch 0:  71%|███████   | 2324/3295 [24:19<10:09,  1.59it/s]
Epoch 0:  71%|███████   | 2324/3295 [24:19<10:09,  1.59it/s]
Epoch 0:  71%|███████   | 2325/3295 [24:19<10:09,  1.59it/s]
Epoch 0:  71%|███████   | 2325/3295 [24:19<10:09,  1.59it/s]
Epoch 0:  71%|███████   | 2326/3295 [24:19<10:08,  1.59it/s]
Epoch 0:  71%|███████   | 2326/3295 [24:19<10:08,  1.59it/s]
Epoch 0:  71%|███████   | 2327/3295 [24:21<10:08,  1.59it/s]
Epoch 0:  71%|███████   | 2327/3295 [24:21<10:08,  1.59it/s]
Epoch 0:  71%|███████   | 2328/3295 [24:21<10:07,  1.59it/s]
Epoch 0:  71%|███████   | 2328/3295 [24:21<10:07,  1.59it/s]
Epoch 0:  71%|███████   | 2329/3295 [24:21<10:06,  1.59it/s]
Epoch 0:  71%|███████   | 2329/3295 [24:21<10:06,  1.59it/s]
Epoch 0:  71%|███████   | 2330/3295 [24:23<10:06,  1.59it/s]
Epoch 0:  71%|███████   | 2330/3295 [24:23<10:06,  1.59it/s]
Epoch 0:  71%|███████   | 2331/3295 [24:23<10:05,  1.59it/s]
Epoch 0:  71%|███████   | 2331/3295 [24:23<10:05,  1.59it/s]
Epoch 0:  71%|███████   | 2332/3295 [24:23<10:04,  1.59it/s]
Epoch 0:  71%|███████   | 2332/3295 [24:23<10:04,  1.59it/s]
Epoch 0:  71%|███████   | 2333/3295 [24:25<10:04,  1.59it/s]
Epoch 0:  71%|███████   | 2333/3295 [24:25<10:04,  1.59it/s]
Epoch 0:  71%|███████   | 2334/3295 [24:25<10:03,  1.59it/s]
Epoch 0:  71%|███████   | 2334/3295 [24:25<10:03,  1.59it/s]
Epoch 0:  71%|███████   | 2335/3295 [24:25<10:02,  1.59it/s]
Epoch 0:  71%|███████   | 2335/3295 [24:25<10:02,  1.59it/s]
Epoch 0:  71%|███████   | 2336/3295 [24:27<10:02,  1.59it/s]
Epoch 0:  71%|███████   | 2336/3295 [24:27<10:02,  1.59it/s]
Epoch 0:  71%|███████   | 2337/3295 [24:27<10:01,  1.59it/s]
Epoch 0:  71%|███████   | 2337/3295 [24:27<10:01,  1.59it/s]
Epoch 0:  71%|███████   | 2338/3295 [24:27<10:00,  1.59it/s]
Epoch 0:  71%|███████   | 2338/3295 [24:27<10:00,  1.59it/s]
Epoch 0:  71%|███████   | 2339/3295 [24:29<10:00,  1.59it/s]
Epoch 0:  71%|███████   | 2339/3295 [24:29<10:00,  1.59it/s]
Epoch 0:  71%|███████   | 2340/3295 [24:29<09:59,  1.59it/s]
Epoch 0:  71%|███████   | 2340/3295 [24:29<09:59,  1.59it/s]
Epoch 0:  71%|███████   | 2341/3295 [24:29<09:58,  1.59it/s]
Epoch 0:  71%|███████   | 2341/3295 [24:29<09:58,  1.59it/s]
Epoch 0:  71%|███████   | 2342/3295 [24:30<09:58,  1.59it/s]
Epoch 0:  71%|███████   | 2342/3295 [24:30<09:58,  1.59it/s]
Epoch 0:  71%|███████   | 2343/3295 [24:30<09:57,  1.59it/s]
Epoch 0:  71%|███████   | 2343/3295 [24:30<09:57,  1.59it/s]
Epoch 0:  71%|███████   | 2344/3295 [24:30<09:56,  1.59it/s]
Epoch 0:  71%|███████   | 2344/3295 [24:30<09:56,  1.59it/s]
Epoch 0:  71%|███████   | 2345/3295 [24:32<09:56,  1.59it/s]
Epoch 0:  71%|███████   | 2345/3295 [24:32<09:56,  1.59it/s]
Epoch 0:  71%|███████   | 2346/3295 [24:32<09:55,  1.59it/s]
Epoch 0:  71%|███████   | 2346/3295 [24:32<09:55,  1.59it/s]
Epoch 0:  71%|███████   | 2347/3295 [24:32<09:54,  1.59it/s]
Epoch 0:  71%|███████   | 2347/3295 [24:32<09:54,  1.59it/s]
Epoch 0:  71%|███████▏  | 2348/3295 [24:34<09:54,  1.59it/s]
Epoch 0:  71%|███████▏  | 2348/3295 [24:34<09:54,  1.59it/s]
Epoch 0:  71%|███████▏  | 2349/3295 [24:34<09:53,  1.59it/s]
Epoch 0:  71%|███████▏  | 2349/3295 [24:34<09:53,  1.59it/s]
Epoch 0:  71%|███████▏  | 2350/3295 [24:34<09:53,  1.59it/s]
Epoch 0:  71%|███████▏  | 2350/3295 [24:34<09:53,  1.59it/s]
Epoch 0:  71%|███████▏  | 2351/3295 [24:36<09:52,  1.59it/s]
Epoch 0:  71%|███████▏  | 2351/3295 [24:36<09:52,  1.59it/s]
Epoch 0:  71%|███████▏  | 2352/3295 [24:36<09:51,  1.59it/s]
Epoch 0:  71%|███████▏  | 2352/3295 [24:36<09:51,  1.59it/s]
Epoch 0:  71%|███████▏  | 2353/3295 [24:36<09:51,  1.59it/s]
Epoch 0:  71%|███████▏  | 2353/3295 [24:36<09:51,  1.59it/s]
Epoch 0:  71%|███████▏  | 2354/3295 [24:38<09:51,  1.59it/s]
Epoch 0:  71%|███████▏  | 2354/3295 [24:38<09:51,  1.59it/s]
Epoch 0:  71%|███████▏  | 2355/3295 [24:38<09:50,  1.59it/s]
Epoch 0:  71%|███████▏  | 2355/3295 [24:38<09:50,  1.59it/s]
Epoch 0:  72%|███████▏  | 2356/3295 [24:38<09:49,  1.59it/s]
Epoch 0:  72%|███████▏  | 2356/3295 [24:38<09:49,  1.59it/s]
Epoch 0:  72%|███████▏  | 2357/3295 [24:40<09:49,  1.59it/s]
Epoch 0:  72%|███████▏  | 2357/3295 [24:40<09:49,  1.59it/s]
Epoch 0:  72%|███████▏  | 2358/3295 [24:40<09:48,  1.59it/s]
Epoch 0:  72%|███████▏  | 2358/3295 [24:40<09:48,  1.59it/s]
Epoch 0:  72%|███████▏  | 2359/3295 [24:40<09:47,  1.59it/s]
Epoch 0:  72%|███████▏  | 2359/3295 [24:40<09:47,  1.59it/s]
Epoch 0:  72%|███████▏  | 2360/3295 [24:42<09:47,  1.59it/s]
Epoch 0:  72%|███████▏  | 2360/3295 [24:42<09:47,  1.59it/s]
Epoch 0:  72%|███████▏  | 2361/3295 [24:42<09:46,  1.59it/s]
Epoch 0:  72%|███████▏  | 2361/3295 [24:42<09:46,  1.59it/s]
Epoch 0:  72%|███████▏  | 2362/3295 [24:42<09:45,  1.59it/s]
Epoch 0:  72%|███████▏  | 2362/3295 [24:42<09:45,  1.59it/s]
Epoch 0:  72%|███████▏  | 2363/3295 [24:44<09:45,  1.59it/s]
Epoch 0:  72%|███████▏  | 2363/3295 [24:44<09:45,  1.59it/s]
Epoch 0:  72%|███████▏  | 2364/3295 [24:44<09:44,  1.59it/s]
Epoch 0:  72%|███████▏  | 2364/3295 [24:44<09:44,  1.59it/s]
Epoch 0:  72%|███████▏  | 2365/3295 [24:44<09:43,  1.59it/s]
Epoch 0:  72%|███████▏  | 2365/3295 [24:44<09:43,  1.59it/s]
Epoch 0:  72%|███████▏  | 2366/3295 [24:46<09:43,  1.59it/s]
Epoch 0:  72%|███████▏  | 2366/3295 [24:46<09:43,  1.59it/s]
Epoch 0:  72%|███████▏  | 2367/3295 [24:46<09:42,  1.59it/s]
Epoch 0:  72%|███████▏  | 2367/3295 [24:46<09:42,  1.59it/s]
Epoch 0:  72%|███████▏  | 2368/3295 [24:46<09:41,  1.59it/s]
Epoch 0:  72%|███████▏  | 2368/3295 [24:46<09:41,  1.59it/s]
Epoch 0:  72%|███████▏  | 2369/3295 [24:48<09:41,  1.59it/s]
Epoch 0:  72%|███████▏  | 2369/3295 [24:48<09:41,  1.59it/s]
Epoch 0:  72%|███████▏  | 2370/3295 [24:48<09:40,  1.59it/s]
Epoch 0:  72%|███████▏  | 2370/3295 [24:48<09:40,  1.59it/s]
Epoch 0:  72%|███████▏  | 2371/3295 [24:48<09:39,  1.59it/s]
Epoch 0:  72%|███████▏  | 2371/3295 [24:48<09:39,  1.59it/s]
Epoch 0:  72%|███████▏  | 2372/3295 [24:49<09:39,  1.59it/s]
Epoch 0:  72%|███████▏  | 2372/3295 [24:49<09:39,  1.59it/s]
Epoch 0:  72%|███████▏  | 2373/3295 [24:50<09:38,  1.59it/s]
Epoch 0:  72%|███████▏  | 2373/3295 [24:50<09:38,  1.59it/s]
Epoch 0:  72%|███████▏  | 2374/3295 [24:50<09:38,  1.59it/s]
Epoch 0:  72%|███████▏  | 2374/3295 [24:50<09:38,  1.59it/s]
Epoch 0:  72%|███████▏  | 2375/3295 [24:51<09:37,  1.59it/s]
Epoch 0:  72%|███████▏  | 2375/3295 [24:51<09:37,  1.59it/s]
Epoch 0:  72%|███████▏  | 2376/3295 [24:51<09:36,  1.59it/s]
Epoch 0:  72%|███████▏  | 2376/3295 [24:51<09:36,  1.59it/s]
Epoch 0:  72%|███████▏  | 2377/3295 [24:51<09:36,  1.59it/s]
Epoch 0:  72%|███████▏  | 2377/3295 [24:51<09:36,  1.59it/s]
Epoch 0:  72%|███████▏  | 2378/3295 [24:53<09:35,  1.59it/s]
Epoch 0:  72%|███████▏  | 2378/3295 [24:53<09:35,  1.59it/s]
Epoch 0:  72%|███████▏  | 2379/3295 [24:53<09:35,  1.59it/s]
Epoch 0:  72%|███████▏  | 2379/3295 [24:53<09:35,  1.59it/s]
Epoch 0:  72%|███████▏  | 2380/3295 [24:53<09:34,  1.59it/s]
Epoch 0:  72%|███████▏  | 2380/3295 [24:53<09:34,  1.59it/s]
Epoch 0:  72%|███████▏  | 2381/3295 [24:55<09:34,  1.59it/s]
Epoch 0:  72%|███████▏  | 2381/3295 [24:55<09:34,  1.59it/s]
Epoch 0:  72%|███████▏  | 2382/3295 [24:55<09:33,  1.59it/s]
Epoch 0:  72%|███████▏  | 2382/3295 [24:55<09:33,  1.59it/s]
Epoch 0:  72%|███████▏  | 2383/3295 [24:55<09:32,  1.59it/s]
Epoch 0:  72%|███████▏  | 2383/3295 [24:55<09:32,  1.59it/s]
Epoch 0:  72%|███████▏  | 2384/3295 [24:57<09:32,  1.59it/s]
Epoch 0:  72%|███████▏  | 2384/3295 [24:57<09:32,  1.59it/s]
Epoch 0:  72%|███████▏  | 2385/3295 [24:57<09:31,  1.59it/s]
Epoch 0:  72%|███████▏  | 2385/3295 [24:57<09:31,  1.59it/s]
Epoch 0:  72%|███████▏  | 2386/3295 [24:57<09:30,  1.59it/s]
Epoch 0:  72%|███████▏  | 2386/3295 [24:57<09:30,  1.59it/s]
Epoch 0:  72%|███████▏  | 2387/3295 [24:59<09:30,  1.59it/s]
Epoch 0:  72%|███████▏  | 2387/3295 [24:59<09:30,  1.59it/s]
Epoch 0:  72%|███████▏  | 2388/3295 [24:59<09:29,  1.59it/s]
Epoch 0:  72%|███████▏  | 2388/3295 [24:59<09:29,  1.59it/s]
Epoch 0:  73%|███████▎  | 2389/3295 [24:59<09:28,  1.59it/s]
Epoch 0:  73%|███████▎  | 2389/3295 [24:59<09:28,  1.59it/s]
Epoch 0:  73%|███████▎  | 2390/3295 [25:01<09:28,  1.59it/s]
Epoch 0:  73%|███████▎  | 2390/3295 [25:01<09:28,  1.59it/s]
Epoch 0:  73%|███████▎  | 2391/3295 [25:01<09:27,  1.59it/s]
Epoch 0:  73%|███████▎  | 2391/3295 [25:01<09:27,  1.59it/s]
Epoch 0:  73%|███████▎  | 2392/3295 [25:01<09:26,  1.59it/s]
Epoch 0:  73%|███████▎  | 2392/3295 [25:01<09:26,  1.59it/s]
Epoch 0:  73%|███████▎  | 2393/3295 [25:02<09:26,  1.59it/s]
Epoch 0:  73%|███████▎  | 2393/3295 [25:02<09:26,  1.59it/s]
Epoch 0:  73%|███████▎  | 2394/3295 [25:03<09:25,  1.59it/s]
Epoch 0:  73%|███████▎  | 2394/3295 [25:03<09:25,  1.59it/s]
Epoch 0:  73%|███████▎  | 2395/3295 [25:03<09:24,  1.59it/s]
Epoch 0:  73%|███████▎  | 2395/3295 [25:03<09:24,  1.59it/s]
Epoch 0:  73%|███████▎  | 2396/3295 [25:04<09:24,  1.59it/s]
Epoch 0:  73%|███████▎  | 2396/3295 [25:04<09:24,  1.59it/s]
Epoch 0:  73%|███████▎  | 2397/3295 [25:05<09:23,  1.59it/s]
Epoch 0:  73%|███████▎  | 2397/3295 [25:05<09:23,  1.59it/s]
Epoch 0:  73%|███████▎  | 2398/3295 [25:05<09:22,  1.59it/s]
Epoch 0:  73%|███████▎  | 2398/3295 [25:05<09:22,  1.59it/s]
Epoch 0:  73%|███████▎  | 2399/3295 [25:06<09:22,  1.59it/s]
Epoch 0:  73%|███████▎  | 2399/3295 [25:06<09:22,  1.59it/s]
Epoch 0:  73%|███████▎  | 2400/3295 [25:06<09:21,  1.59it/s]
Epoch 0:  73%|███████▎  | 2400/3295 [25:06<09:21,  1.59it/s]
Epoch 0:  73%|███████▎  | 2401/3295 [25:06<09:21,  1.59it/s]
Epoch 0:  73%|███████▎  | 2401/3295 [25:06<09:21,  1.59it/s]
Epoch 0:  73%|███████▎  | 2402/3295 [25:08<09:20,  1.59it/s]
Epoch 0:  73%|███████▎  | 2402/3295 [25:08<09:20,  1.59it/s]
Epoch 0:  73%|███████▎  | 2403/3295 [25:08<09:20,  1.59it/s]
Epoch 0:  73%|███████▎  | 2403/3295 [25:08<09:20,  1.59it/s]
Epoch 0:  73%|███████▎  | 2404/3295 [25:08<09:19,  1.59it/s]
Epoch 0:  73%|███████▎  | 2404/3295 [25:08<09:19,  1.59it/s]
Epoch 0:  73%|███████▎  | 2405/3295 [25:10<09:18,  1.59it/s]
Epoch 0:  73%|███████▎  | 2405/3295 [25:10<09:18,  1.59it/s]
Epoch 0:  73%|███████▎  | 2406/3295 [25:10<09:18,  1.59it/s]
Epoch 0:  73%|███████▎  | 2406/3295 [25:10<09:18,  1.59it/s]
Epoch 0:  73%|███████▎  | 2407/3295 [25:10<09:17,  1.59it/s]
Epoch 0:  73%|███████▎  | 2407/3295 [25:10<09:17,  1.59it/s]
Epoch 0:  73%|███████▎  | 2408/3295 [25:12<09:17,  1.59it/s]
Epoch 0:  73%|███████▎  | 2408/3295 [25:12<09:17,  1.59it/s]
Epoch 0:  73%|███████▎  | 2409/3295 [25:12<09:16,  1.59it/s]
Epoch 0:  73%|███████▎  | 2409/3295 [25:12<09:16,  1.59it/s]
Epoch 0:  73%|███████▎  | 2410/3295 [25:12<09:15,  1.59it/s]
Epoch 0:  73%|███████▎  | 2410/3295 [25:12<09:15,  1.59it/s]
Epoch 0:  73%|███████▎  | 2411/3295 [25:14<09:15,  1.59it/s]
Epoch 0:  73%|███████▎  | 2411/3295 [25:14<09:15,  1.59it/s]
Epoch 0:  73%|███████▎  | 2412/3295 [25:14<09:14,  1.59it/s]
Epoch 0:  73%|███████▎  | 2412/3295 [25:14<09:14,  1.59it/s]
Epoch 0:  73%|███████▎  | 2413/3295 [25:14<09:13,  1.59it/s]
Epoch 0:  73%|███████▎  | 2413/3295 [25:14<09:13,  1.59it/s]
Epoch 0:  73%|███████▎  | 2414/3295 [25:16<09:13,  1.59it/s]
Epoch 0:  73%|███████▎  | 2414/3295 [25:16<09:13,  1.59it/s]
Epoch 0:  73%|███████▎  | 2415/3295 [25:16<09:12,  1.59it/s]
Epoch 0:  73%|███████▎  | 2415/3295 [25:16<09:12,  1.59it/s]
Epoch 0:  73%|███████▎  | 2416/3295 [25:16<09:11,  1.59it/s]
Epoch 0:  73%|███████▎  | 2416/3295 [25:16<09:11,  1.59it/s]
Epoch 0:  73%|███████▎  | 2417/3295 [25:18<09:11,  1.59it/s]
Epoch 0:  73%|███████▎  | 2417/3295 [25:18<09:11,  1.59it/s]
Epoch 0:  73%|███████▎  | 2418/3295 [25:18<09:10,  1.59it/s]
Epoch 0:  73%|███████▎  | 2418/3295 [25:18<09:10,  1.59it/s]
Epoch 0:  73%|███████▎  | 2419/3295 [25:18<09:09,  1.59it/s]
Epoch 0:  73%|███████▎  | 2419/3295 [25:18<09:09,  1.59it/s]
Epoch 0:  73%|███████▎  | 2420/3295 [25:19<09:09,  1.59it/s]
Epoch 0:  73%|███████▎  | 2420/3295 [25:19<09:09,  1.59it/s]
Epoch 0:  73%|███████▎  | 2421/3295 [25:19<09:08,  1.59it/s]
Epoch 0:  73%|███████▎  | 2421/3295 [25:19<09:08,  1.59it/s]
Epoch 0:  74%|███████▎  | 2422/3295 [25:20<09:07,  1.59it/s]
Epoch 0:  74%|███████▎  | 2422/3295 [25:20<09:07,  1.59it/s]
Epoch 0:  74%|███████▎  | 2423/3295 [25:21<09:07,  1.59it/s]
Epoch 0:  74%|███████▎  | 2423/3295 [25:21<09:07,  1.59it/s]
Epoch 0:  74%|███████▎  | 2424/3295 [25:21<09:06,  1.59it/s]
Epoch 0:  74%|███████▎  | 2424/3295 [25:21<09:06,  1.59it/s]
Epoch 0:  74%|███████▎  | 2425/3295 [25:21<09:05,  1.59it/s]
Epoch 0:  74%|███████▎  | 2425/3295 [25:21<09:05,  1.59it/s]
Epoch 0:  74%|███████▎  | 2426/3295 [25:23<09:05,  1.59it/s]
Epoch 0:  74%|███████▎  | 2426/3295 [25:23<09:05,  1.59it/s]
Epoch 0:  74%|███████▎  | 2427/3295 [25:23<09:04,  1.59it/s]
Epoch 0:  74%|███████▎  | 2427/3295 [25:23<09:04,  1.59it/s]
Epoch 0:  74%|███████▎  | 2428/3295 [25:23<09:04,  1.59it/s]
Epoch 0:  74%|███████▎  | 2428/3295 [25:23<09:04,  1.59it/s]
Epoch 0:  74%|███████▎  | 2429/3295 [25:25<09:03,  1.59it/s]
Epoch 0:  74%|███████▎  | 2429/3295 [25:25<09:03,  1.59it/s]
Epoch 0:  74%|███████▎  | 2430/3295 [25:25<09:03,  1.59it/s]
Epoch 0:  74%|███████▎  | 2430/3295 [25:25<09:03,  1.59it/s]
Epoch 0:  74%|███████▍  | 2431/3295 [25:25<09:02,  1.59it/s]
Epoch 0:  74%|███████▍  | 2431/3295 [25:25<09:02,  1.59it/s]
Epoch 0:  74%|███████▍  | 2432/3295 [25:27<09:02,  1.59it/s]
Epoch 0:  74%|███████▍  | 2432/3295 [25:27<09:02,  1.59it/s]
Epoch 0:  74%|███████▍  | 2433/3295 [25:27<09:01,  1.59it/s]
Epoch 0:  74%|███████▍  | 2433/3295 [25:27<09:01,  1.59it/s]
Epoch 0:  74%|███████▍  | 2434/3295 [25:27<09:00,  1.59it/s]
Epoch 0:  74%|███████▍  | 2434/3295 [25:27<09:00,  1.59it/s]
Epoch 0:  74%|███████▍  | 2435/3295 [25:29<09:00,  1.59it/s]
Epoch 0:  74%|███████▍  | 2435/3295 [25:29<09:00,  1.59it/s]
Epoch 0:  74%|███████▍  | 2436/3295 [25:29<08:59,  1.59it/s]
Epoch 0:  74%|███████▍  | 2436/3295 [25:29<08:59,  1.59it/s]
Epoch 0:  74%|███████▍  | 2437/3295 [25:29<08:58,  1.59it/s]
Epoch 0:  74%|███████▍  | 2437/3295 [25:29<08:58,  1.59it/s]
Epoch 0:  74%|███████▍  | 2438/3295 [25:31<08:58,  1.59it/s]
Epoch 0:  74%|███████▍  | 2438/3295 [25:31<08:58,  1.59it/s]
Epoch 0:  74%|███████▍  | 2439/3295 [25:31<08:57,  1.59it/s]
Epoch 0:  74%|███████▍  | 2439/3295 [25:31<08:57,  1.59it/s]
Epoch 0:  74%|███████▍  | 2440/3295 [25:31<08:56,  1.59it/s]
Epoch 0:  74%|███████▍  | 2440/3295 [25:31<08:56,  1.59it/s]
Epoch 0:  74%|███████▍  | 2441/3295 [25:32<08:56,  1.59it/s]
Epoch 0:  74%|███████▍  | 2441/3295 [25:32<08:56,  1.59it/s]
Epoch 0:  74%|███████▍  | 2442/3295 [25:33<08:55,  1.59it/s]
Epoch 0:  74%|███████▍  | 2442/3295 [25:33<08:55,  1.59it/s]
Epoch 0:  74%|███████▍  | 2443/3295 [25:33<08:54,  1.59it/s]
Epoch 0:  74%|███████▍  | 2443/3295 [25:33<08:54,  1.59it/s]
Epoch 0:  74%|███████▍  | 2444/3295 [25:34<08:54,  1.59it/s]
Epoch 0:  74%|███████▍  | 2444/3295 [25:34<08:54,  1.59it/s]
Epoch 0:  74%|███████▍  | 2445/3295 [25:34<08:53,  1.59it/s]
Epoch 0:  74%|███████▍  | 2445/3295 [25:34<08:53,  1.59it/s]
Epoch 0:  74%|███████▍  | 2446/3295 [25:34<08:52,  1.59it/s]
Epoch 0:  74%|███████▍  | 2446/3295 [25:34<08:52,  1.59it/s]
Epoch 0:  74%|███████▍  | 2447/3295 [25:36<08:52,  1.59it/s]
Epoch 0:  74%|███████▍  | 2447/3295 [25:36<08:52,  1.59it/s]
Epoch 0:  74%|███████▍  | 2448/3295 [25:36<08:51,  1.59it/s]
Epoch 0:  74%|███████▍  | 2448/3295 [25:36<08:51,  1.59it/s]
Epoch 0:  74%|███████▍  | 2449/3295 [25:36<08:50,  1.59it/s]
Epoch 0:  74%|███████▍  | 2449/3295 [25:36<08:50,  1.59it/s]
Epoch 0:  74%|███████▍  | 2450/3295 [25:38<08:50,  1.59it/s]
Epoch 0:  74%|███████▍  | 2450/3295 [25:38<08:50,  1.59it/s]
Epoch 0:  74%|███████▍  | 2451/3295 [25:38<08:49,  1.59it/s]
Epoch 0:  74%|███████▍  | 2451/3295 [25:38<08:49,  1.59it/s]
Epoch 0:  74%|███████▍  | 2452/3295 [25:38<08:48,  1.59it/s]
Epoch 0:  74%|███████▍  | 2452/3295 [25:38<08:48,  1.59it/s]
Epoch 0:  74%|███████▍  | 2453/3295 [25:40<08:48,  1.59it/s]
Epoch 0:  74%|███████▍  | 2453/3295 [25:40<08:48,  1.59it/s]
Epoch 0:  74%|███████▍  | 2454/3295 [25:40<08:47,  1.59it/s]
Epoch 0:  74%|███████▍  | 2454/3295 [25:40<08:47,  1.59it/s]
Epoch 0:  75%|███████▍  | 2455/3295 [25:40<08:47,  1.59it/s]
Epoch 0:  75%|███████▍  | 2455/3295 [25:40<08:47,  1.59it/s]
Epoch 0:  75%|███████▍  | 2456/3295 [25:42<08:46,  1.59it/s]
Epoch 0:  75%|███████▍  | 2456/3295 [25:42<08:46,  1.59it/s]
Epoch 0:  75%|███████▍  | 2457/3295 [25:42<08:45,  1.59it/s]
Epoch 0:  75%|███████▍  | 2457/3295 [25:42<08:45,  1.59it/s]
Epoch 0:  75%|███████▍  | 2458/3295 [25:42<08:45,  1.59it/s]
Epoch 0:  75%|███████▍  | 2458/3295 [25:42<08:45,  1.59it/s]
Epoch 0:  75%|███████▍  | 2459/3295 [25:43<08:44,  1.59it/s]
Epoch 0:  75%|███████▍  | 2459/3295 [25:43<08:44,  1.59it/s]
Epoch 0:  75%|███████▍  | 2460/3295 [25:43<08:44,  1.59it/s]
Epoch 0:  75%|███████▍  | 2460/3295 [25:43<08:44,  1.59it/s]
Epoch 0:  75%|███████▍  | 2461/3295 [25:43<08:43,  1.59it/s]
Epoch 0:  75%|███████▍  | 2461/3295 [25:43<08:43,  1.59it/s]
Epoch 0:  75%|███████▍  | 2462/3295 [25:45<08:43,  1.59it/s]
Epoch 0:  75%|███████▍  | 2462/3295 [25:45<08:43,  1.59it/s]
Epoch 0:  75%|███████▍  | 2463/3295 [25:45<08:42,  1.59it/s]
Epoch 0:  75%|███████▍  | 2463/3295 [25:45<08:42,  1.59it/s]
Epoch 0:  75%|███████▍  | 2464/3295 [25:45<08:41,  1.59it/s]
Epoch 0:  75%|███████▍  | 2464/3295 [25:45<08:41,  1.59it/s]
Epoch 0:  75%|███████▍  | 2465/3295 [25:47<08:41,  1.59it/s]
Epoch 0:  75%|███████▍  | 2465/3295 [25:47<08:41,  1.59it/s]
Epoch 0:  75%|███████▍  | 2466/3295 [25:47<08:40,  1.59it/s]
Epoch 0:  75%|███████▍  | 2466/3295 [25:47<08:40,  1.59it/s]
Epoch 0:  75%|███████▍  | 2467/3295 [25:47<08:39,  1.59it/s]
Epoch 0:  75%|███████▍  | 2467/3295 [25:47<08:39,  1.59it/s]
Epoch 0:  75%|███████▍  | 2468/3295 [25:49<08:39,  1.59it/s]
Epoch 0:  75%|███████▍  | 2468/3295 [25:49<08:39,  1.59it/s]
Epoch 0:  75%|███████▍  | 2469/3295 [25:49<08:38,  1.59it/s]
Epoch 0:  75%|███████▍  | 2469/3295 [25:49<08:38,  1.59it/s]
Epoch 0:  75%|███████▍  | 2470/3295 [25:49<08:37,  1.59it/s]
Epoch 0:  75%|███████▍  | 2470/3295 [25:49<08:37,  1.59it/s]
Epoch 0:  75%|███████▍  | 2471/3295 [25:51<08:37,  1.59it/s]
Epoch 0:  75%|███████▍  | 2471/3295 [25:51<08:37,  1.59it/s]
Epoch 0:  75%|███████▌  | 2472/3295 [25:51<08:36,  1.59it/s]
Epoch 0:  75%|███████▌  | 2472/3295 [25:51<08:36,  1.59it/s]
Epoch 0:  75%|███████▌  | 2473/3295 [25:51<08:35,  1.59it/s]
Epoch 0:  75%|███████▌  | 2473/3295 [25:51<08:35,  1.59it/s]
Epoch 0:  75%|███████▌  | 2474/3295 [25:53<08:35,  1.59it/s]
Epoch 0:  75%|███████▌  | 2474/3295 [25:53<08:35,  1.59it/s]
Epoch 0:  75%|███████▌  | 2475/3295 [25:53<08:34,  1.59it/s]
Epoch 0:  75%|███████▌  | 2475/3295 [25:53<08:34,  1.59it/s]
Epoch 0:  75%|███████▌  | 2476/3295 [25:53<08:33,  1.59it/s]
Epoch 0:  75%|███████▌  | 2476/3295 [25:53<08:33,  1.59it/s]
Epoch 0:  75%|███████▌  | 2477/3295 [25:55<08:33,  1.59it/s]
Epoch 0:  75%|███████▌  | 2477/3295 [25:55<08:33,  1.59it/s]
Epoch 0:  75%|███████▌  | 2478/3295 [25:55<08:32,  1.59it/s]
Epoch 0:  75%|███████▌  | 2478/3295 [25:55<08:32,  1.59it/s]
Epoch 0:  75%|███████▌  | 2479/3295 [25:55<08:31,  1.59it/s]
Epoch 0:  75%|███████▌  | 2479/3295 [25:55<08:31,  1.59it/s]
Epoch 0:  75%|███████▌  | 2480/3295 [25:57<08:31,  1.59it/s]
Epoch 0:  75%|███████▌  | 2480/3295 [25:57<08:31,  1.59it/s]
Epoch 0:  75%|███████▌  | 2481/3295 [25:57<08:30,  1.59it/s]
Epoch 0:  75%|███████▌  | 2481/3295 [25:57<08:30,  1.59it/s]
Epoch 0:  75%|███████▌  | 2482/3295 [25:57<08:30,  1.59it/s]
Epoch 0:  75%|███████▌  | 2482/3295 [25:57<08:30,  1.59it/s]
Epoch 0:  75%|███████▌  | 2483/3295 [25:58<08:29,  1.59it/s]
Epoch 0:  75%|███████▌  | 2483/3295 [25:58<08:29,  1.59it/s]
Epoch 0:  75%|███████▌  | 2484/3295 [25:58<08:28,  1.59it/s]
Epoch 0:  75%|███████▌  | 2484/3295 [25:58<08:28,  1.59it/s]
Epoch 0:  75%|███████▌  | 2485/3295 [25:59<08:28,  1.59it/s]
Epoch 0:  75%|███████▌  | 2485/3295 [25:59<08:28,  1.59it/s]
Epoch 0:  75%|███████▌  | 2486/3295 [26:00<08:27,  1.59it/s]
Epoch 0:  75%|███████▌  | 2486/3295 [26:00<08:27,  1.59it/s]
Epoch 0:  75%|███████▌  | 2487/3295 [26:00<08:27,  1.59it/s]
Epoch 0:  75%|███████▌  | 2487/3295 [26:00<08:27,  1.59it/s]
Epoch 0:  76%|███████▌  | 2488/3295 [26:00<08:26,  1.59it/s]
Epoch 0:  76%|███████▌  | 2488/3295 [26:00<08:26,  1.59it/s]
Epoch 0:  76%|███████▌  | 2489/3295 [26:02<08:25,  1.59it/s]
Epoch 0:  76%|███████▌  | 2489/3295 [26:02<08:25,  1.59it/s]
Epoch 0:  76%|███████▌  | 2490/3295 [26:02<08:25,  1.59it/s]
Epoch 0:  76%|███████▌  | 2490/3295 [26:02<08:25,  1.59it/s]
Epoch 0:  76%|███████▌  | 2491/3295 [26:02<08:24,  1.59it/s]
Epoch 0:  76%|███████▌  | 2491/3295 [26:02<08:24,  1.59it/s]
Epoch 0:  76%|███████▌  | 2492/3295 [26:04<08:24,  1.59it/s]
Epoch 0:  76%|███████▌  | 2492/3295 [26:04<08:24,  1.59it/s]
Epoch 0:  76%|███████▌  | 2493/3295 [26:04<08:23,  1.59it/s]
Epoch 0:  76%|███████▌  | 2493/3295 [26:04<08:23,  1.59it/s]
Epoch 0:  76%|███████▌  | 2494/3295 [26:04<08:22,  1.59it/s]
Epoch 0:  76%|███████▌  | 2494/3295 [26:04<08:22,  1.59it/s]
Epoch 0:  76%|███████▌  | 2495/3295 [26:06<08:22,  1.59it/s]
Epoch 0:  76%|███████▌  | 2495/3295 [26:06<08:22,  1.59it/s]
Epoch 0:  76%|███████▌  | 2496/3295 [26:06<08:21,  1.59it/s]
Epoch 0:  76%|███████▌  | 2496/3295 [26:06<08:21,  1.59it/s]
Epoch 0:  76%|███████▌  | 2497/3295 [26:06<08:20,  1.59it/s]
Epoch 0:  76%|███████▌  | 2497/3295 [26:06<08:20,  1.59it/s]
Epoch 0:  76%|███████▌  | 2498/3295 [26:08<08:20,  1.59it/s]
Epoch 0:  76%|███████▌  | 2498/3295 [26:08<08:20,  1.59it/s]
Epoch 0:  76%|███████▌  | 2499/3295 [26:08<08:19,  1.59it/s]
Epoch 0:  76%|███████▌  | 2499/3295 [26:08<08:19,  1.59it/s]
Epoch 0:  76%|███████▌  | 2500/3295 [26:08<08:18,  1.59it/s]
Epoch 0:  76%|███████▌  | 2500/3295 [26:08<08:18,  1.59it/s]
Epoch 0:  76%|███████▌  | 2501/3295 [26:09<08:18,  1.59it/s]
Epoch 0:  76%|███████▌  | 2501/3295 [26:09<08:18,  1.59it/s]
Epoch 0:  76%|███████▌  | 2502/3295 [26:10<08:17,  1.59it/s]
Epoch 0:  76%|███████▌  | 2502/3295 [26:10<08:17,  1.59it/s]
Epoch 0:  76%|███████▌  | 2503/3295 [26:10<08:16,  1.59it/s]
Epoch 0:  76%|███████▌  | 2503/3295 [26:10<08:16,  1.59it/s]
Epoch 0:  76%|███████▌  | 2504/3295 [26:11<08:16,  1.59it/s]
Epoch 0:  76%|███████▌  | 2504/3295 [26:11<08:16,  1.59it/s]
Epoch 0:  76%|███████▌  | 2505/3295 [26:11<08:15,  1.59it/s]
Epoch 0:  76%|███████▌  | 2505/3295 [26:11<08:15,  1.59it/s]
Epoch 0:  76%|███████▌  | 2506/3295 [26:11<08:14,  1.59it/s]
Epoch 0:  76%|███████▌  | 2506/3295 [26:11<08:14,  1.59it/s]
Epoch 0:  76%|███████▌  | 2507/3295 [26:13<08:14,  1.59it/s]
Epoch 0:  76%|███████▌  | 2507/3295 [26:13<08:14,  1.59it/s]
Epoch 0:  76%|███████▌  | 2508/3295 [26:13<08:13,  1.59it/s]
Epoch 0:  76%|███████▌  | 2508/3295 [26:13<08:13,  1.59it/s]
Epoch 0:  76%|███████▌  | 2509/3295 [26:13<08:13,  1.59it/s]
Epoch 0:  76%|███████▌  | 2509/3295 [26:13<08:13,  1.59it/s]
Epoch 0:  76%|███████▌  | 2510/3295 [26:15<08:12,  1.59it/s]
Epoch 0:  76%|███████▌  | 2510/3295 [26:15<08:12,  1.59it/s]
Epoch 0:  76%|███████▌  | 2511/3295 [26:15<08:11,  1.59it/s]
Epoch 0:  76%|███████▌  | 2511/3295 [26:15<08:11,  1.59it/s]
Epoch 0:  76%|███████▌  | 2512/3295 [26:15<08:11,  1.59it/s]
Epoch 0:  76%|███████▌  | 2512/3295 [26:15<08:11,  1.59it/s]
Epoch 0:  76%|███████▋  | 2513/3295 [26:17<08:10,  1.59it/s]
Epoch 0:  76%|███████▋  | 2513/3295 [26:17<08:10,  1.59it/s]
Epoch 0:  76%|███████▋  | 2514/3295 [26:17<08:09,  1.59it/s]
Epoch 0:  76%|███████▋  | 2514/3295 [26:17<08:09,  1.59it/s]
Epoch 0:  76%|███████▋  | 2515/3295 [26:17<08:09,  1.59it/s]
Epoch 0:  76%|███████▋  | 2515/3295 [26:17<08:09,  1.59it/s]
Epoch 0:  76%|███████▋  | 2516/3295 [26:18<08:08,  1.59it/s]
Epoch 0:  76%|███████▋  | 2516/3295 [26:18<08:08,  1.59it/s]
Epoch 0:  76%|███████▋  | 2517/3295 [26:18<08:08,  1.59it/s]
Epoch 0:  76%|███████▋  | 2517/3295 [26:18<08:08,  1.59it/s]
Epoch 0:  76%|███████▋  | 2518/3295 [26:19<08:07,  1.59it/s]
Epoch 0:  76%|███████▋  | 2518/3295 [26:19<08:07,  1.59it/s]
Epoch 0:  76%|███████▋  | 2519/3295 [26:20<08:07,  1.59it/s]
Epoch 0:  76%|███████▋  | 2519/3295 [26:20<08:07,  1.59it/s]
Epoch 0:  76%|███████▋  | 2520/3295 [26:20<08:06,  1.59it/s]
Epoch 0:  76%|███████▋  | 2520/3295 [26:20<08:06,  1.59it/s]
Epoch 0:  77%|███████▋  | 2521/3295 [26:20<08:05,  1.59it/s]
Epoch 0:  77%|███████▋  | 2521/3295 [26:20<08:05,  1.59it/s]
Epoch 0:  77%|███████▋  | 2522/3295 [26:22<08:05,  1.59it/s]
Epoch 0:  77%|███████▋  | 2522/3295 [26:22<08:05,  1.59it/s]
Epoch 0:  77%|███████▋  | 2523/3295 [26:22<08:04,  1.59it/s]
Epoch 0:  77%|███████▋  | 2523/3295 [26:22<08:04,  1.59it/s]
Epoch 0:  77%|███████▋  | 2524/3295 [26:22<08:03,  1.59it/s]
Epoch 0:  77%|███████▋  | 2524/3295 [26:22<08:03,  1.59it/s]
Epoch 0:  77%|███████▋  | 2525/3295 [26:24<08:03,  1.59it/s]
Epoch 0:  77%|███████▋  | 2525/3295 [26:24<08:03,  1.59it/s]
Epoch 0:  77%|███████▋  | 2526/3295 [26:24<08:02,  1.59it/s]
Epoch 0:  77%|███████▋  | 2526/3295 [26:24<08:02,  1.59it/s]
Epoch 0:  77%|███████▋  | 2527/3295 [26:24<08:01,  1.59it/s]
Epoch 0:  77%|███████▋  | 2527/3295 [26:24<08:01,  1.59it/s]
Epoch 0:  77%|███████▋  | 2528/3295 [26:26<08:01,  1.59it/s]
Epoch 0:  77%|███████▋  | 2528/3295 [26:26<08:01,  1.59it/s]
Epoch 0:  77%|███████▋  | 2529/3295 [26:26<08:00,  1.59it/s]
Epoch 0:  77%|███████▋  | 2529/3295 [26:26<08:00,  1.59it/s]
Epoch 0:  77%|███████▋  | 2530/3295 [26:26<07:59,  1.59it/s]
Epoch 0:  77%|███████▋  | 2530/3295 [26:26<07:59,  1.59it/s]
Epoch 0:  77%|███████▋  | 2531/3295 [26:28<07:59,  1.59it/s]
Epoch 0:  77%|███████▋  | 2531/3295 [26:28<07:59,  1.59it/s]
Epoch 0:  77%|███████▋  | 2532/3295 [26:28<07:58,  1.59it/s]
Epoch 0:  77%|███████▋  | 2532/3295 [26:28<07:58,  1.59it/s]
Epoch 0:  77%|███████▋  | 2533/3295 [26:28<07:57,  1.59it/s]
Epoch 0:  77%|███████▋  | 2533/3295 [26:28<07:57,  1.59it/s]
Epoch 0:  77%|███████▋  | 2534/3295 [26:29<07:57,  1.59it/s]
Epoch 0:  77%|███████▋  | 2534/3295 [26:29<07:57,  1.59it/s]
Epoch 0:  77%|███████▋  | 2535/3295 [26:29<07:56,  1.59it/s]
Epoch 0:  77%|███████▋  | 2535/3295 [26:29<07:56,  1.59it/s]
Epoch 0:  77%|███████▋  | 2536/3295 [26:29<07:55,  1.60it/s]
Epoch 0:  77%|███████▋  | 2536/3295 [26:29<07:55,  1.60it/s]
Epoch 0:  77%|███████▋  | 2537/3295 [26:31<07:55,  1.59it/s]
Epoch 0:  77%|███████▋  | 2537/3295 [26:31<07:55,  1.59it/s]
Epoch 0:  77%|███████▋  | 2538/3295 [26:31<07:54,  1.59it/s]
Epoch 0:  77%|███████▋  | 2538/3295 [26:31<07:54,  1.59it/s]
Epoch 0:  77%|███████▋  | 2539/3295 [26:31<07:53,  1.59it/s]
Epoch 0:  77%|███████▋  | 2539/3295 [26:31<07:53,  1.59it/s]
Epoch 0:  77%|███████▋  | 2540/3295 [26:33<07:53,  1.59it/s]
Epoch 0:  77%|███████▋  | 2540/3295 [26:33<07:53,  1.59it/s]
Epoch 0:  77%|███████▋  | 2541/3295 [26:33<07:52,  1.59it/s]
Epoch 0:  77%|███████▋  | 2541/3295 [26:33<07:52,  1.59it/s]
Epoch 0:  77%|███████▋  | 2542/3295 [26:33<07:52,  1.59it/s]
Epoch 0:  77%|███████▋  | 2542/3295 [26:33<07:52,  1.59it/s]
Epoch 0:  77%|███████▋  | 2543/3295 [26:35<07:51,  1.59it/s]
Epoch 0:  77%|███████▋  | 2543/3295 [26:35<07:51,  1.59it/s]
Epoch 0:  77%|███████▋  | 2544/3295 [26:35<07:51,  1.59it/s]
Epoch 0:  77%|███████▋  | 2544/3295 [26:35<07:51,  1.59it/s]
Epoch 0:  77%|███████▋  | 2545/3295 [26:35<07:50,  1.59it/s]
Epoch 0:  77%|███████▋  | 2545/3295 [26:35<07:50,  1.59it/s]
Epoch 0:  77%|███████▋  | 2546/3295 [26:37<07:49,  1.59it/s]
Epoch 0:  77%|███████▋  | 2546/3295 [26:37<07:49,  1.59it/s]
Epoch 0:  77%|███████▋  | 2547/3295 [26:37<07:49,  1.59it/s]
Epoch 0:  77%|███████▋  | 2547/3295 [26:37<07:49,  1.59it/s]
Epoch 0:  77%|███████▋  | 2548/3295 [26:37<07:48,  1.60it/s]
Epoch 0:  77%|███████▋  | 2548/3295 [26:37<07:48,  1.60it/s]
Epoch 0:  77%|███████▋  | 2549/3295 [26:39<07:48,  1.59it/s]
Epoch 0:  77%|███████▋  | 2549/3295 [26:39<07:48,  1.59it/s]
Epoch 0:  77%|███████▋  | 2550/3295 [26:39<07:47,  1.59it/s]
Epoch 0:  77%|███████▋  | 2550/3295 [26:39<07:47,  1.59it/s]
Epoch 0:  77%|███████▋  | 2551/3295 [26:39<07:46,  1.60it/s]
Epoch 0:  77%|███████▋  | 2551/3295 [26:39<07:46,  1.60it/s]
Epoch 0:  77%|███████▋  | 2552/3295 [26:41<07:46,  1.59it/s]
Epoch 0:  77%|███████▋  | 2552/3295 [26:41<07:46,  1.59it/s]
Epoch 0:  77%|███████▋  | 2553/3295 [26:41<07:45,  1.59it/s]
Epoch 0:  77%|███████▋  | 2553/3295 [26:41<07:45,  1.59it/s]
Epoch 0:  78%|███████▊  | 2554/3295 [26:41<07:44,  1.60it/s]
Epoch 0:  78%|███████▊  | 2554/3295 [26:41<07:44,  1.60it/s]
Epoch 0:  78%|███████▊  | 2555/3295 [26:43<07:44,  1.59it/s]
Epoch 0:  78%|███████▊  | 2555/3295 [26:43<07:44,  1.59it/s]
Epoch 0:  78%|███████▊  | 2556/3295 [26:43<07:43,  1.59it/s]
Epoch 0:  78%|███████▊  | 2556/3295 [26:43<07:43,  1.59it/s]
Epoch 0:  78%|███████▊  | 2557/3295 [26:43<07:42,  1.60it/s]
Epoch 0:  78%|███████▊  | 2557/3295 [26:43<07:42,  1.60it/s]
Epoch 0:  78%|███████▊  | 2558/3295 [26:44<07:42,  1.59it/s]
Epoch 0:  78%|███████▊  | 2558/3295 [26:44<07:42,  1.59it/s]
Epoch 0:  78%|███████▊  | 2559/3295 [26:44<07:41,  1.59it/s]
Epoch 0:  78%|███████▊  | 2559/3295 [26:44<07:41,  1.59it/s]
Epoch 0:  78%|███████▊  | 2560/3295 [26:44<07:40,  1.60it/s]
Epoch 0:  78%|███████▊  | 2560/3295 [26:44<07:40,  1.60it/s]
Epoch 0:  78%|███████▊  | 2561/3295 [26:46<07:40,  1.59it/s]
Epoch 0:  78%|███████▊  | 2561/3295 [26:46<07:40,  1.59it/s]
Epoch 0:  78%|███████▊  | 2562/3295 [26:46<07:39,  1.59it/s]
Epoch 0:  78%|███████▊  | 2562/3295 [26:46<07:39,  1.59it/s]
Epoch 0:  78%|███████▊  | 2563/3295 [26:46<07:38,  1.60it/s]
Epoch 0:  78%|███████▊  | 2563/3295 [26:46<07:38,  1.60it/s]
Epoch 0:  78%|███████▊  | 2564/3295 [26:48<07:38,  1.59it/s]
Epoch 0:  78%|███████▊  | 2564/3295 [26:48<07:38,  1.59it/s]
Epoch 0:  78%|███████▊  | 2565/3295 [26:48<07:37,  1.59it/s]
Epoch 0:  78%|███████▊  | 2565/3295 [26:48<07:37,  1.59it/s]
Epoch 0:  78%|███████▊  | 2566/3295 [26:48<07:37,  1.60it/s]
Epoch 0:  78%|███████▊  | 2566/3295 [26:48<07:37,  1.60it/s]
Epoch 0:  78%|███████▊  | 2567/3295 [26:50<07:36,  1.59it/s]
Epoch 0:  78%|███████▊  | 2567/3295 [26:50<07:36,  1.59it/s]
Epoch 0:  78%|███████▊  | 2568/3295 [26:50<07:35,  1.59it/s]
Epoch 0:  78%|███████▊  | 2568/3295 [26:50<07:35,  1.59it/s]
Epoch 0:  78%|███████▊  | 2569/3295 [26:50<07:35,  1.60it/s]
Epoch 0:  78%|███████▊  | 2569/3295 [26:50<07:35,  1.60it/s]
Epoch 0:  78%|███████▊  | 2570/3295 [26:52<07:34,  1.59it/s]
Epoch 0:  78%|███████▊  | 2570/3295 [26:52<07:34,  1.59it/s]
Epoch 0:  78%|███████▊  | 2571/3295 [26:52<07:34,  1.59it/s]
Epoch 0:  78%|███████▊  | 2571/3295 [26:52<07:34,  1.59it/s]
Epoch 0:  78%|███████▊  | 2572/3295 [26:52<07:33,  1.60it/s]
Epoch 0:  78%|███████▊  | 2572/3295 [26:52<07:33,  1.60it/s]
Epoch 0:  78%|███████▊  | 2573/3295 [26:54<07:32,  1.59it/s]
Epoch 0:  78%|███████▊  | 2573/3295 [26:54<07:32,  1.59it/s]
Epoch 0:  78%|███████▊  | 2574/3295 [26:54<07:32,  1.59it/s]
Epoch 0:  78%|███████▊  | 2574/3295 [26:54<07:32,  1.59it/s]
Epoch 0:  78%|███████▊  | 2575/3295 [26:54<07:31,  1.60it/s]
Epoch 0:  78%|███████▊  | 2575/3295 [26:54<07:31,  1.60it/s]
Epoch 0:  78%|███████▊  | 2576/3295 [26:55<07:31,  1.59it/s]
Epoch 0:  78%|███████▊  | 2576/3295 [26:55<07:31,  1.59it/s]
Epoch 0:  78%|███████▊  | 2577/3295 [26:55<07:30,  1.59it/s]
Epoch 0:  78%|███████▊  | 2577/3295 [26:55<07:30,  1.59it/s]
Epoch 0:  78%|███████▊  | 2578/3295 [26:56<07:29,  1.60it/s]
Epoch 0:  78%|███████▊  | 2578/3295 [26:56<07:29,  1.60it/s]
Epoch 0:  78%|███████▊  | 2579/3295 [26:57<07:29,  1.59it/s]
Epoch 0:  78%|███████▊  | 2579/3295 [26:57<07:29,  1.59it/s]
Epoch 0:  78%|███████▊  | 2580/3295 [26:57<07:28,  1.59it/s]
Epoch 0:  78%|███████▊  | 2580/3295 [26:57<07:28,  1.59it/s]
Epoch 0:  78%|███████▊  | 2581/3295 [26:57<07:27,  1.60it/s]
Epoch 0:  78%|███████▊  | 2581/3295 [26:57<07:27,  1.60it/s]
Epoch 0:  78%|███████▊  | 2582/3295 [26:59<07:27,  1.59it/s]
Epoch 0:  78%|███████▊  | 2582/3295 [26:59<07:27,  1.59it/s]
Epoch 0:  78%|███████▊  | 2583/3295 [26:59<07:26,  1.59it/s]
Epoch 0:  78%|███████▊  | 2583/3295 [26:59<07:26,  1.59it/s]
Epoch 0:  78%|███████▊  | 2584/3295 [26:59<07:25,  1.60it/s]
Epoch 0:  78%|███████▊  | 2584/3295 [26:59<07:25,  1.60it/s]
Epoch 0:  78%|███████▊  | 2585/3295 [27:01<07:25,  1.59it/s]
Epoch 0:  78%|███████▊  | 2585/3295 [27:01<07:25,  1.59it/s]
Epoch 0:  78%|███████▊  | 2586/3295 [27:01<07:24,  1.59it/s]
Epoch 0:  78%|███████▊  | 2586/3295 [27:01<07:24,  1.59it/s]
Epoch 0:  79%|███████▊  | 2587/3295 [27:01<07:23,  1.60it/s]
Epoch 0:  79%|███████▊  | 2587/3295 [27:01<07:23,  1.60it/s]
Epoch 0:  79%|███████▊  | 2588/3295 [27:03<07:23,  1.59it/s]
Epoch 0:  79%|███████▊  | 2588/3295 [27:03<07:23,  1.59it/s]
Epoch 0:  79%|███████▊  | 2589/3295 [27:03<07:22,  1.59it/s]
Epoch 0:  79%|███████▊  | 2589/3295 [27:03<07:22,  1.59it/s]
Epoch 0:  79%|███████▊  | 2590/3295 [27:03<07:21,  1.60it/s]
Epoch 0:  79%|███████▊  | 2590/3295 [27:03<07:21,  1.60it/s]
Epoch 0:  79%|███████▊  | 2591/3295 [27:04<07:21,  1.59it/s]
Epoch 0:  79%|███████▊  | 2591/3295 [27:04<07:21,  1.59it/s]
Epoch 0:  79%|███████▊  | 2592/3295 [27:04<07:20,  1.60it/s]
Epoch 0:  79%|███████▊  | 2592/3295 [27:04<07:20,  1.60it/s]
Epoch 0:  79%|███████▊  | 2593/3295 [27:04<07:19,  1.60it/s]
Epoch 0:  79%|███████▊  | 2593/3295 [27:04<07:19,  1.60it/s]
Epoch 0:  79%|███████▊  | 2594/3295 [27:06<07:19,  1.59it/s]
Epoch 0:  79%|███████▊  | 2594/3295 [27:06<07:19,  1.59it/s]
Epoch 0:  79%|███████▉  | 2595/3295 [27:06<07:18,  1.60it/s]
Epoch 0:  79%|███████▉  | 2595/3295 [27:06<07:18,  1.60it/s]
Epoch 0:  79%|███████▉  | 2596/3295 [27:06<07:18,  1.60it/s]
Epoch 0:  79%|███████▉  | 2596/3295 [27:06<07:18,  1.60it/s]
Epoch 0:  79%|███████▉  | 2597/3295 [27:08<07:17,  1.59it/s]
Epoch 0:  79%|███████▉  | 2597/3295 [27:08<07:17,  1.59it/s]
Epoch 0:  79%|███████▉  | 2598/3295 [27:08<07:16,  1.60it/s]
Epoch 0:  79%|███████▉  | 2598/3295 [27:08<07:16,  1.60it/s]
Epoch 0:  79%|███████▉  | 2599/3295 [27:08<07:16,  1.60it/s]
Epoch 0:  79%|███████▉  | 2599/3295 [27:08<07:16,  1.60it/s]
Epoch 0:  79%|███████▉  | 2600/3295 [27:10<07:15,  1.59it/s]
Epoch 0:  79%|███████▉  | 2600/3295 [27:10<07:15,  1.59it/s]
Epoch 0:  79%|███████▉  | 2601/3295 [27:10<07:15,  1.60it/s]
Epoch 0:  79%|███████▉  | 2601/3295 [27:10<07:15,  1.60it/s]
Epoch 0:  79%|███████▉  | 2602/3295 [27:10<07:14,  1.60it/s]
Epoch 0:  79%|███████▉  | 2602/3295 [27:10<07:14,  1.60it/s]
Epoch 0:  79%|███████▉  | 2603/3295 [27:12<07:13,  1.59it/s]
Epoch 0:  79%|███████▉  | 2603/3295 [27:12<07:13,  1.59it/s]
Epoch 0:  79%|███████▉  | 2604/3295 [27:12<07:13,  1.60it/s]
Epoch 0:  79%|███████▉  | 2604/3295 [27:12<07:13,  1.60it/s]
Epoch 0:  79%|███████▉  | 2605/3295 [27:12<07:12,  1.60it/s]
Epoch 0:  79%|███████▉  | 2605/3295 [27:12<07:12,  1.60it/s]
Epoch 0:  79%|███████▉  | 2606/3295 [27:14<07:12,  1.59it/s]
Epoch 0:  79%|███████▉  | 2606/3295 [27:14<07:12,  1.59it/s]
Epoch 0:  79%|███████▉  | 2607/3295 [27:14<07:11,  1.60it/s]
Epoch 0:  79%|███████▉  | 2607/3295 [27:14<07:11,  1.60it/s]
Epoch 0:  79%|███████▉  | 2608/3295 [27:14<07:10,  1.60it/s]
Epoch 0:  79%|███████▉  | 2608/3295 [27:14<07:10,  1.60it/s]
Epoch 0:  79%|███████▉  | 2609/3295 [27:16<07:10,  1.59it/s]
Epoch 0:  79%|███████▉  | 2609/3295 [27:16<07:10,  1.59it/s]
Epoch 0:  79%|███████▉  | 2610/3295 [27:16<07:09,  1.60it/s]
Epoch 0:  79%|███████▉  | 2610/3295 [27:16<07:09,  1.60it/s]
Epoch 0:  79%|███████▉  | 2611/3295 [27:16<07:08,  1.60it/s]
Epoch 0:  79%|███████▉  | 2611/3295 [27:16<07:08,  1.60it/s]
Epoch 0:  79%|███████▉  | 2612/3295 [27:18<07:08,  1.59it/s]
Epoch 0:  79%|███████▉  | 2612/3295 [27:18<07:08,  1.59it/s]
Epoch 0:  79%|███████▉  | 2613/3295 [27:18<07:07,  1.60it/s]
Epoch 0:  79%|███████▉  | 2613/3295 [27:18<07:07,  1.60it/s]
Epoch 0:  79%|███████▉  | 2614/3295 [27:18<07:06,  1.60it/s]
Epoch 0:  79%|███████▉  | 2614/3295 [27:18<07:06,  1.60it/s]
Epoch 0:  79%|███████▉  | 2615/3295 [27:19<07:06,  1.59it/s]
Epoch 0:  79%|███████▉  | 2615/3295 [27:19<07:06,  1.59it/s]
Epoch 0:  79%|███████▉  | 2616/3295 [27:19<07:05,  1.60it/s]
Epoch 0:  79%|███████▉  | 2616/3295 [27:19<07:05,  1.60it/s]
Epoch 0:  79%|███████▉  | 2617/3295 [27:19<07:04,  1.60it/s]
Epoch 0:  79%|███████▉  | 2617/3295 [27:19<07:04,  1.60it/s]
Epoch 0:  79%|███████▉  | 2618/3295 [27:21<07:04,  1.59it/s]
Epoch 0:  79%|███████▉  | 2618/3295 [27:21<07:04,  1.59it/s]
Epoch 0:  79%|███████▉  | 2619/3295 [27:21<07:03,  1.60it/s]
Epoch 0:  79%|███████▉  | 2619/3295 [27:21<07:03,  1.60it/s]
Epoch 0:  80%|███████▉  | 2620/3295 [27:21<07:02,  1.60it/s]
Epoch 0:  80%|███████▉  | 2620/3295 [27:21<07:02,  1.60it/s]
Epoch 0:  80%|███████▉  | 2621/3295 [27:23<07:02,  1.59it/s]
Epoch 0:  80%|███████▉  | 2621/3295 [27:23<07:02,  1.59it/s]
Epoch 0:  80%|███████▉  | 2622/3295 [27:23<07:01,  1.60it/s]
Epoch 0:  80%|███████▉  | 2622/3295 [27:23<07:01,  1.60it/s]
Epoch 0:  80%|███████▉  | 2623/3295 [27:23<07:01,  1.60it/s]
Epoch 0:  80%|███████▉  | 2623/3295 [27:23<07:01,  1.60it/s]
Epoch 0:  80%|███████▉  | 2624/3295 [27:25<07:00,  1.59it/s]
Epoch 0:  80%|███████▉  | 2624/3295 [27:25<07:00,  1.59it/s]
Epoch 0:  80%|███████▉  | 2625/3295 [27:25<06:59,  1.60it/s]
Epoch 0:  80%|███████▉  | 2625/3295 [27:25<06:59,  1.60it/s]
Epoch 0:  80%|███████▉  | 2626/3295 [27:25<06:59,  1.60it/s]
Epoch 0:  80%|███████▉  | 2626/3295 [27:25<06:59,  1.60it/s]
Epoch 0:  80%|███████▉  | 2627/3295 [27:27<06:58,  1.59it/s]
Epoch 0:  80%|███████▉  | 2627/3295 [27:27<06:58,  1.59it/s]
Epoch 0:  80%|███████▉  | 2628/3295 [27:27<06:58,  1.60it/s]
Epoch 0:  80%|███████▉  | 2628/3295 [27:27<06:58,  1.60it/s]
Epoch 0:  80%|███████▉  | 2629/3295 [27:27<06:57,  1.60it/s]
Epoch 0:  80%|███████▉  | 2629/3295 [27:27<06:57,  1.60it/s]
Epoch 0:  80%|███████▉  | 2630/3295 [27:29<06:57,  1.59it/s]
Epoch 0:  80%|███████▉  | 2630/3295 [27:29<06:57,  1.59it/s]
Epoch 0:  80%|███████▉  | 2631/3295 [27:29<06:56,  1.60it/s]
Epoch 0:  80%|███████▉  | 2631/3295 [27:29<06:56,  1.60it/s]
Epoch 0:  80%|███████▉  | 2632/3295 [27:29<06:55,  1.60it/s]
Epoch 0:  80%|███████▉  | 2632/3295 [27:29<06:55,  1.60it/s]
Epoch 0:  80%|███████▉  | 2633/3295 [27:31<06:55,  1.59it/s]
Epoch 0:  80%|███████▉  | 2633/3295 [27:31<06:55,  1.59it/s]
Epoch 0:  80%|███████▉  | 2634/3295 [27:31<06:54,  1.60it/s]
Epoch 0:  80%|███████▉  | 2634/3295 [27:31<06:54,  1.60it/s]
Epoch 0:  80%|███████▉  | 2635/3295 [27:31<06:53,  1.60it/s]
Epoch 0:  80%|███████▉  | 2635/3295 [27:31<06:53,  1.60it/s]
Epoch 0:  80%|████████  | 2636/3295 [27:32<06:53,  1.59it/s]
Epoch 0:  80%|████████  | 2636/3295 [27:32<06:53,  1.59it/s]
Epoch 0:  80%|████████  | 2637/3295 [27:32<06:52,  1.60it/s]
Epoch 0:  80%|████████  | 2637/3295 [27:32<06:52,  1.60it/s]
Epoch 0:  80%|████████  | 2638/3295 [27:33<06:51,  1.60it/s]
Epoch 0:  80%|████████  | 2638/3295 [27:33<06:51,  1.60it/s]
Epoch 0:  80%|████████  | 2639/3295 [27:34<06:51,  1.59it/s]
Epoch 0:  80%|████████  | 2639/3295 [27:34<06:51,  1.59it/s]
Epoch 0:  80%|████████  | 2640/3295 [27:34<06:50,  1.60it/s]
Epoch 0:  80%|████████  | 2640/3295 [27:34<06:50,  1.60it/s]
Epoch 0:  80%|████████  | 2641/3295 [27:34<06:49,  1.60it/s]
Epoch 0:  80%|████████  | 2641/3295 [27:34<06:49,  1.60it/s]
Epoch 0:  80%|████████  | 2642/3295 [27:36<06:49,  1.59it/s]
Epoch 0:  80%|████████  | 2642/3295 [27:36<06:49,  1.59it/s]
Epoch 0:  80%|████████  | 2643/3295 [27:36<06:48,  1.60it/s]
Epoch 0:  80%|████████  | 2643/3295 [27:36<06:48,  1.60it/s]
Epoch 0:  80%|████████  | 2644/3295 [27:36<06:47,  1.60it/s]
Epoch 0:  80%|████████  | 2644/3295 [27:36<06:47,  1.60it/s]
Epoch 0:  80%|████████  | 2645/3295 [27:38<06:47,  1.59it/s]
Epoch 0:  80%|████████  | 2645/3295 [27:38<06:47,  1.59it/s]
Epoch 0:  80%|████████  | 2646/3295 [27:38<06:46,  1.60it/s]
Epoch 0:  80%|████████  | 2646/3295 [27:38<06:46,  1.60it/s]
Epoch 0:  80%|████████  | 2647/3295 [27:38<06:46,  1.60it/s]
Epoch 0:  80%|████████  | 2647/3295 [27:38<06:46,  1.60it/s]
Epoch 0:  80%|████████  | 2648/3295 [27:40<06:45,  1.59it/s]
Epoch 0:  80%|████████  | 2648/3295 [27:40<06:45,  1.59it/s]
Epoch 0:  80%|████████  | 2649/3295 [27:40<06:44,  1.60it/s]
Epoch 0:  80%|████████  | 2649/3295 [27:40<06:44,  1.60it/s]
Epoch 0:  80%|████████  | 2650/3295 [27:40<06:44,  1.60it/s]
Epoch 0:  80%|████████  | 2650/3295 [27:40<06:44,  1.60it/s]
Epoch 0:  80%|████████  | 2651/3295 [27:42<06:43,  1.59it/s]
Epoch 0:  80%|████████  | 2651/3295 [27:42<06:43,  1.59it/s]
Epoch 0:  80%|████████  | 2652/3295 [27:42<06:43,  1.60it/s]
Epoch 0:  80%|████████  | 2652/3295 [27:42<06:43,  1.60it/s]
Epoch 0:  81%|████████  | 2653/3295 [27:42<06:42,  1.60it/s]
Epoch 0:  81%|████████  | 2653/3295 [27:42<06:42,  1.60it/s]
Epoch 0:  81%|████████  | 2654/3295 [27:44<06:41,  1.59it/s]
Epoch 0:  81%|████████  | 2654/3295 [27:44<06:41,  1.59it/s]
Epoch 0:  81%|████████  | 2655/3295 [27:44<06:41,  1.60it/s]
Epoch 0:  81%|████████  | 2655/3295 [27:44<06:41,  1.60it/s]
Epoch 0:  81%|████████  | 2656/3295 [27:44<06:40,  1.60it/s]
Epoch 0:  81%|████████  | 2656/3295 [27:44<06:40,  1.60it/s]
Epoch 0:  81%|████████  | 2657/3295 [27:45<06:40,  1.59it/s]
Epoch 0:  81%|████████  | 2657/3295 [27:45<06:40,  1.59it/s]
Epoch 0:  81%|████████  | 2658/3295 [27:45<06:39,  1.60it/s]
Epoch 0:  81%|████████  | 2658/3295 [27:45<06:39,  1.60it/s]
Epoch 0:  81%|████████  | 2659/3295 [27:45<06:38,  1.60it/s]
Epoch 0:  81%|████████  | 2659/3295 [27:45<06:38,  1.60it/s]
Epoch 0:  81%|████████  | 2660/3295 [27:47<06:38,  1.59it/s]
Epoch 0:  81%|████████  | 2660/3295 [27:47<06:38,  1.59it/s]
Epoch 0:  81%|████████  | 2661/3295 [27:47<06:37,  1.60it/s]
Epoch 0:  81%|████████  | 2661/3295 [27:47<06:37,  1.60it/s]
Epoch 0:  81%|████████  | 2662/3295 [27:47<06:36,  1.60it/s]
Epoch 0:  81%|████████  | 2662/3295 [27:47<06:36,  1.60it/s]
Epoch 0:  81%|████████  | 2663/3295 [27:49<06:36,  1.60it/s]
Epoch 0:  81%|████████  | 2663/3295 [27:49<06:36,  1.60it/s]
Epoch 0:  81%|████████  | 2664/3295 [27:49<06:35,  1.60it/s]
Epoch 0:  81%|████████  | 2664/3295 [27:49<06:35,  1.60it/s]
Epoch 0:  81%|████████  | 2665/3295 [27:49<06:34,  1.60it/s]
Epoch 0:  81%|████████  | 2665/3295 [27:49<06:34,  1.60it/s]
Epoch 0:  81%|████████  | 2666/3295 [27:51<06:34,  1.60it/s]
Epoch 0:  81%|████████  | 2666/3295 [27:51<06:34,  1.60it/s]
Epoch 0:  81%|████████  | 2667/3295 [27:51<06:33,  1.60it/s]
Epoch 0:  81%|████████  | 2667/3295 [27:51<06:33,  1.60it/s]
Epoch 0:  81%|████████  | 2668/3295 [27:51<06:32,  1.60it/s]
Epoch 0:  81%|████████  | 2668/3295 [27:51<06:32,  1.60it/s]
Epoch 0:  81%|████████  | 2669/3295 [27:53<06:32,  1.60it/s]
Epoch 0:  81%|████████  | 2669/3295 [27:53<06:32,  1.60it/s]
Epoch 0:  81%|████████  | 2670/3295 [27:53<06:31,  1.60it/s]
Epoch 0:  81%|████████  | 2670/3295 [27:53<06:31,  1.60it/s]
Epoch 0:  81%|████████  | 2671/3295 [27:53<06:30,  1.60it/s]
Epoch 0:  81%|████████  | 2671/3295 [27:53<06:30,  1.60it/s]
Epoch 0:  81%|████████  | 2672/3295 [27:54<06:30,  1.60it/s]
Epoch 0:  81%|████████  | 2672/3295 [27:54<06:30,  1.60it/s]
Epoch 0:  81%|████████  | 2673/3295 [27:54<06:29,  1.60it/s]
Epoch 0:  81%|████████  | 2673/3295 [27:54<06:29,  1.60it/s]
Epoch 0:  81%|████████  | 2674/3295 [27:54<06:28,  1.60it/s]
Epoch 0:  81%|████████  | 2674/3295 [27:54<06:28,  1.60it/s]
Epoch 0:  81%|████████  | 2675/3295 [27:56<06:28,  1.60it/s]
Epoch 0:  81%|████████  | 2675/3295 [27:56<06:28,  1.60it/s]
Epoch 0:  81%|████████  | 2676/3295 [27:56<06:27,  1.60it/s]
Epoch 0:  81%|████████  | 2676/3295 [27:56<06:27,  1.60it/s]
Epoch 0:  81%|████████  | 2677/3295 [27:56<06:27,  1.60it/s]
Epoch 0:  81%|████████  | 2677/3295 [27:56<06:27,  1.60it/s]
Epoch 0:  81%|████████▏ | 2678/3295 [27:58<06:26,  1.60it/s]
Epoch 0:  81%|████████▏ | 2678/3295 [27:58<06:26,  1.60it/s]
Epoch 0:  81%|████████▏ | 2679/3295 [27:58<06:25,  1.60it/s]
Epoch 0:  81%|████████▏ | 2679/3295 [27:58<06:25,  1.60it/s]
Epoch 0:  81%|████████▏ | 2680/3295 [27:58<06:25,  1.60it/s]
Epoch 0:  81%|████████▏ | 2680/3295 [27:58<06:25,  1.60it/s]
Epoch 0:  81%|████████▏ | 2681/3295 [28:00<06:24,  1.60it/s]
Epoch 0:  81%|████████▏ | 2681/3295 [28:00<06:24,  1.60it/s]
Epoch 0:  81%|████████▏ | 2682/3295 [28:00<06:24,  1.60it/s]
Epoch 0:  81%|████████▏ | 2682/3295 [28:00<06:24,  1.60it/s]
Epoch 0:  81%|████████▏ | 2683/3295 [28:00<06:23,  1.60it/s]
Epoch 0:  81%|████████▏ | 2683/3295 [28:00<06:23,  1.60it/s]
Epoch 0:  81%|████████▏ | 2684/3295 [28:02<06:22,  1.60it/s]
Epoch 0:  81%|████████▏ | 2684/3295 [28:02<06:22,  1.60it/s]
Epoch 0:  81%|████████▏ | 2685/3295 [28:02<06:22,  1.60it/s]
Epoch 0:  81%|████████▏ | 2685/3295 [28:02<06:22,  1.60it/s]
Epoch 0:  82%|████████▏ | 2686/3295 [28:02<06:21,  1.60it/s]
Epoch 0:  82%|████████▏ | 2686/3295 [28:02<06:21,  1.60it/s]
Epoch 0:  82%|████████▏ | 2687/3295 [28:04<06:21,  1.60it/s]
Epoch 0:  82%|████████▏ | 2687/3295 [28:04<06:21,  1.60it/s]
Epoch 0:  82%|████████▏ | 2688/3295 [28:04<06:20,  1.60it/s]
Epoch 0:  82%|████████▏ | 2688/3295 [28:04<06:20,  1.60it/s]
Epoch 0:  82%|████████▏ | 2689/3295 [28:04<06:19,  1.60it/s]
Epoch 0:  82%|████████▏ | 2689/3295 [28:04<06:19,  1.60it/s]
Epoch 0:  82%|████████▏ | 2690/3295 [28:05<06:19,  1.60it/s]
Epoch 0:  82%|████████▏ | 2690/3295 [28:05<06:19,  1.60it/s]
Epoch 0:  82%|████████▏ | 2691/3295 [28:05<06:18,  1.60it/s]
Epoch 0:  82%|████████▏ | 2691/3295 [28:05<06:18,  1.60it/s]
Epoch 0:  82%|████████▏ | 2692/3295 [28:05<06:17,  1.60it/s]
Epoch 0:  82%|████████▏ | 2692/3295 [28:05<06:17,  1.60it/s]
Epoch 0:  82%|████████▏ | 2693/3295 [28:07<06:17,  1.60it/s]
Epoch 0:  82%|████████▏ | 2693/3295 [28:07<06:17,  1.60it/s]
Epoch 0:  82%|████████▏ | 2694/3295 [28:07<06:16,  1.60it/s]
Epoch 0:  82%|████████▏ | 2694/3295 [28:07<06:16,  1.60it/s]
Epoch 0:  82%|████████▏ | 2695/3295 [28:07<06:15,  1.60it/s]
Epoch 0:  82%|████████▏ | 2695/3295 [28:07<06:15,  1.60it/s]
Epoch 0:  82%|████████▏ | 2696/3295 [28:09<06:15,  1.60it/s]
Epoch 0:  82%|████████▏ | 2696/3295 [28:09<06:15,  1.60it/s]
Epoch 0:  82%|████████▏ | 2697/3295 [28:09<06:14,  1.60it/s]
Epoch 0:  82%|████████▏ | 2697/3295 [28:09<06:14,  1.60it/s]
Epoch 0:  82%|████████▏ | 2698/3295 [28:09<06:13,  1.60it/s]
Epoch 0:  82%|████████▏ | 2698/3295 [28:09<06:13,  1.60it/s]
Epoch 0:  82%|████████▏ | 2699/3295 [28:11<06:13,  1.60it/s]
Epoch 0:  82%|████████▏ | 2699/3295 [28:11<06:13,  1.60it/s]
Epoch 0:  82%|████████▏ | 2700/3295 [28:11<06:12,  1.60it/s]
Epoch 0:  82%|████████▏ | 2700/3295 [28:11<06:12,  1.60it/s]
Epoch 0:  82%|████████▏ | 2701/3295 [28:11<06:11,  1.60it/s]
Epoch 0:  82%|████████▏ | 2701/3295 [28:11<06:11,  1.60it/s]
Epoch 0:  82%|████████▏ | 2702/3295 [28:13<06:11,  1.60it/s]
Epoch 0:  82%|████████▏ | 2702/3295 [28:13<06:11,  1.60it/s]
Epoch 0:  82%|████████▏ | 2703/3295 [28:13<06:10,  1.60it/s]
Epoch 0:  82%|████████▏ | 2703/3295 [28:13<06:10,  1.60it/s]
Epoch 0:  82%|████████▏ | 2704/3295 [28:13<06:10,  1.60it/s]
Epoch 0:  82%|████████▏ | 2704/3295 [28:13<06:10,  1.60it/s]
Epoch 0:  82%|████████▏ | 2705/3295 [28:14<06:09,  1.60it/s]
Epoch 0:  82%|████████▏ | 2705/3295 [28:14<06:09,  1.60it/s]
Epoch 0:  82%|████████▏ | 2706/3295 [28:15<06:08,  1.60it/s]
Epoch 0:  82%|████████▏ | 2706/3295 [28:15<06:08,  1.60it/s]
Epoch 0:  82%|████████▏ | 2707/3295 [28:15<06:08,  1.60it/s]
Epoch 0:  82%|████████▏ | 2707/3295 [28:15<06:08,  1.60it/s]
Epoch 0:  82%|████████▏ | 2708/3295 [28:16<06:07,  1.60it/s]
Epoch 0:  82%|████████▏ | 2708/3295 [28:16<06:07,  1.60it/s]
Epoch 0:  82%|████████▏ | 2709/3295 [28:16<06:07,  1.60it/s]
Epoch 0:  82%|████████▏ | 2709/3295 [28:16<06:07,  1.60it/s]
Epoch 0:  82%|████████▏ | 2710/3295 [28:16<06:06,  1.60it/s]
Epoch 0:  82%|████████▏ | 2710/3295 [28:16<06:06,  1.60it/s]
Epoch 0:  82%|████████▏ | 2711/3295 [28:18<06:05,  1.60it/s]
Epoch 0:  82%|████████▏ | 2711/3295 [28:18<06:05,  1.60it/s]
Epoch 0:  82%|████████▏ | 2712/3295 [28:18<06:05,  1.60it/s]
Epoch 0:  82%|████████▏ | 2712/3295 [28:18<06:05,  1.60it/s]
Epoch 0:  82%|████████▏ | 2713/3295 [28:18<06:04,  1.60it/s]
Epoch 0:  82%|████████▏ | 2713/3295 [28:18<06:04,  1.60it/s]
Epoch 0:  82%|████████▏ | 2714/3295 [28:20<06:04,  1.60it/s]
Epoch 0:  82%|████████▏ | 2714/3295 [28:20<06:04,  1.60it/s]
Epoch 0:  82%|████████▏ | 2715/3295 [28:20<06:03,  1.60it/s]
Epoch 0:  82%|████████▏ | 2715/3295 [28:20<06:03,  1.60it/s]
Epoch 0:  82%|████████▏ | 2716/3295 [28:20<06:02,  1.60it/s]
Epoch 0:  82%|████████▏ | 2716/3295 [28:20<06:02,  1.60it/s]
Epoch 0:  82%|████████▏ | 2717/3295 [28:22<06:02,  1.60it/s]
Epoch 0:  82%|████████▏ | 2717/3295 [28:22<06:02,  1.60it/s]
Epoch 0:  82%|████████▏ | 2718/3295 [28:22<06:01,  1.60it/s]
Epoch 0:  82%|████████▏ | 2718/3295 [28:22<06:01,  1.60it/s]
Epoch 0:  83%|████████▎ | 2719/3295 [28:22<06:00,  1.60it/s]
Epoch 0:  83%|████████▎ | 2719/3295 [28:22<06:00,  1.60it/s]
Epoch 0:  83%|████████▎ | 2720/3295 [28:23<06:00,  1.60it/s]
Epoch 0:  83%|████████▎ | 2720/3295 [28:23<06:00,  1.60it/s]
Epoch 0:  83%|████████▎ | 2721/3295 [28:24<05:59,  1.60it/s]
Epoch 0:  83%|████████▎ | 2721/3295 [28:24<05:59,  1.60it/s]
Epoch 0:  83%|████████▎ | 2722/3295 [28:24<05:58,  1.60it/s]
Epoch 0:  83%|████████▎ | 2722/3295 [28:24<05:58,  1.60it/s]
Epoch 0:  83%|████████▎ | 2723/3295 [28:25<05:58,  1.60it/s]
Epoch 0:  83%|████████▎ | 2723/3295 [28:25<05:58,  1.60it/s]
Epoch 0:  83%|████████▎ | 2724/3295 [28:25<05:57,  1.60it/s]
Epoch 0:  83%|████████▎ | 2724/3295 [28:25<05:57,  1.60it/s]
Epoch 0:  83%|████████▎ | 2725/3295 [28:25<05:56,  1.60it/s]
Epoch 0:  83%|████████▎ | 2725/3295 [28:25<05:56,  1.60it/s]
Epoch 0:  83%|████████▎ | 2726/3295 [28:27<05:56,  1.60it/s]
Epoch 0:  83%|████████▎ | 2726/3295 [28:27<05:56,  1.60it/s]
Epoch 0:  83%|████████▎ | 2727/3295 [28:27<05:55,  1.60it/s]
Epoch 0:  83%|████████▎ | 2727/3295 [28:27<05:55,  1.60it/s]
Epoch 0:  83%|████████▎ | 2728/3295 [28:27<05:54,  1.60it/s]
Epoch 0:  83%|████████▎ | 2728/3295 [28:27<05:54,  1.60it/s]
Epoch 0:  83%|████████▎ | 2729/3295 [28:29<05:54,  1.60it/s]
Epoch 0:  83%|████████▎ | 2729/3295 [28:29<05:54,  1.60it/s]
Epoch 0:  83%|████████▎ | 2730/3295 [28:29<05:53,  1.60it/s]
Epoch 0:  83%|████████▎ | 2730/3295 [28:29<05:53,  1.60it/s]
Epoch 0:  83%|████████▎ | 2731/3295 [28:29<05:53,  1.60it/s]
Epoch 0:  83%|████████▎ | 2731/3295 [28:29<05:53,  1.60it/s]
Epoch 0:  83%|████████▎ | 2732/3295 [28:31<05:52,  1.60it/s]
Epoch 0:  83%|████████▎ | 2732/3295 [28:31<05:52,  1.60it/s]
Epoch 0:  83%|████████▎ | 2733/3295 [28:31<05:51,  1.60it/s]
Epoch 0:  83%|████████▎ | 2733/3295 [28:31<05:51,  1.60it/s]
Epoch 0:  83%|████████▎ | 2734/3295 [28:31<05:51,  1.60it/s]
Epoch 0:  83%|████████▎ | 2734/3295 [28:31<05:51,  1.60it/s]
Epoch 0:  83%|████████▎ | 2735/3295 [28:33<05:50,  1.60it/s]
Epoch 0:  83%|████████▎ | 2735/3295 [28:33<05:50,  1.60it/s]
Epoch 0:  83%|████████▎ | 2736/3295 [28:33<05:50,  1.60it/s]
Epoch 0:  83%|████████▎ | 2736/3295 [28:33<05:50,  1.60it/s]
Epoch 0:  83%|████████▎ | 2737/3295 [28:33<05:49,  1.60it/s]
Epoch 0:  83%|████████▎ | 2737/3295 [28:33<05:49,  1.60it/s]
Epoch 0:  83%|████████▎ | 2738/3295 [28:35<05:48,  1.60it/s]
Epoch 0:  83%|████████▎ | 2738/3295 [28:35<05:48,  1.60it/s]
Epoch 0:  83%|████████▎ | 2739/3295 [28:35<05:48,  1.60it/s]
Epoch 0:  83%|████████▎ | 2739/3295 [28:35<05:48,  1.60it/s]
Epoch 0:  83%|████████▎ | 2740/3295 [28:35<05:47,  1.60it/s]
Epoch 0:  83%|████████▎ | 2740/3295 [28:35<05:47,  1.60it/s]
Epoch 0:  83%|████████▎ | 2741/3295 [28:37<05:47,  1.60it/s]
Epoch 0:  83%|████████▎ | 2741/3295 [28:37<05:47,  1.60it/s]
Epoch 0:  83%|████████▎ | 2742/3295 [28:37<05:46,  1.60it/s]
Epoch 0:  83%|████████▎ | 2742/3295 [28:37<05:46,  1.60it/s]
Epoch 0:  83%|████████▎ | 2743/3295 [28:37<05:45,  1.60it/s]
Epoch 0:  83%|████████▎ | 2743/3295 [28:37<05:45,  1.60it/s]
Epoch 0:  83%|████████▎ | 2744/3295 [28:39<05:45,  1.60it/s]
Epoch 0:  83%|████████▎ | 2744/3295 [28:39<05:45,  1.60it/s]
Epoch 0:  83%|████████▎ | 2745/3295 [28:39<05:44,  1.60it/s]
Epoch 0:  83%|████████▎ | 2745/3295 [28:39<05:44,  1.60it/s]
Epoch 0:  83%|████████▎ | 2746/3295 [28:39<05:43,  1.60it/s]
Epoch 0:  83%|████████▎ | 2746/3295 [28:39<05:43,  1.60it/s]
Epoch 0:  83%|████████▎ | 2747/3295 [28:40<05:43,  1.60it/s]
Epoch 0:  83%|████████▎ | 2747/3295 [28:40<05:43,  1.60it/s]
Epoch 0:  83%|████████▎ | 2748/3295 [28:40<05:42,  1.60it/s]
Epoch 0:  83%|████████▎ | 2748/3295 [28:40<05:42,  1.60it/s]
Epoch 0:  83%|████████▎ | 2749/3295 [28:40<05:41,  1.60it/s]
Epoch 0:  83%|████████▎ | 2749/3295 [28:40<05:41,  1.60it/s]
Epoch 0:  83%|████████▎ | 2750/3295 [28:42<05:41,  1.60it/s]
Epoch 0:  83%|████████▎ | 2750/3295 [28:42<05:41,  1.60it/s]
Epoch 0:  83%|████████▎ | 2751/3295 [28:42<05:40,  1.60it/s]
Epoch 0:  83%|████████▎ | 2751/3295 [28:42<05:40,  1.60it/s]
Epoch 0:  84%|████████▎ | 2752/3295 [28:42<05:39,  1.60it/s]
Epoch 0:  84%|████████▎ | 2752/3295 [28:42<05:39,  1.60it/s]
Epoch 0:  84%|████████▎ | 2753/3295 [28:44<05:39,  1.60it/s]
Epoch 0:  84%|████████▎ | 2753/3295 [28:44<05:39,  1.60it/s]
Epoch 0:  84%|████████▎ | 2754/3295 [28:44<05:38,  1.60it/s]
Epoch 0:  84%|████████▎ | 2754/3295 [28:44<05:38,  1.60it/s]
Epoch 0:  84%|████████▎ | 2755/3295 [28:44<05:38,  1.60it/s]
Epoch 0:  84%|████████▎ | 2755/3295 [28:44<05:38,  1.60it/s]
Epoch 0:  84%|████████▎ | 2756/3295 [28:46<05:37,  1.60it/s]
Epoch 0:  84%|████████▎ | 2756/3295 [28:46<05:37,  1.60it/s]
Epoch 0:  84%|████████▎ | 2757/3295 [28:46<05:36,  1.60it/s]
Epoch 0:  84%|████████▎ | 2757/3295 [28:46<05:36,  1.60it/s]
Epoch 0:  84%|████████▎ | 2758/3295 [28:46<05:36,  1.60it/s]
Epoch 0:  84%|████████▎ | 2758/3295 [28:46<05:36,  1.60it/s]
Epoch 0:  84%|████████▎ | 2759/3295 [28:48<05:35,  1.60it/s]
Epoch 0:  84%|████████▎ | 2759/3295 [28:48<05:35,  1.60it/s]
Epoch 0:  84%|████████▍ | 2760/3295 [28:48<05:34,  1.60it/s]
Epoch 0:  84%|████████▍ | 2760/3295 [28:48<05:34,  1.60it/s]
Epoch 0:  84%|████████▍ | 2761/3295 [28:48<05:34,  1.60it/s]
Epoch 0:  84%|████████▍ | 2761/3295 [28:48<05:34,  1.60it/s]
Epoch 0:  84%|████████▍ | 2762/3295 [28:49<05:33,  1.60it/s]
Epoch 0:  84%|████████▍ | 2762/3295 [28:49<05:33,  1.60it/s]
Epoch 0:  84%|████████▍ | 2763/3295 [28:49<05:33,  1.60it/s]
Epoch 0:  84%|████████▍ | 2763/3295 [28:49<05:33,  1.60it/s]
Epoch 0:  84%|████████▍ | 2764/3295 [28:50<05:32,  1.60it/s]
Epoch 0:  84%|████████▍ | 2764/3295 [28:50<05:32,  1.60it/s]
Epoch 0:  84%|████████▍ | 2765/3295 [28:51<05:31,  1.60it/s]
Epoch 0:  84%|████████▍ | 2765/3295 [28:51<05:31,  1.60it/s]
Epoch 0:  84%|████████▍ | 2766/3295 [28:51<05:31,  1.60it/s]
Epoch 0:  84%|████████▍ | 2766/3295 [28:51<05:31,  1.60it/s]
Epoch 0:  84%|████████▍ | 2767/3295 [28:51<05:30,  1.60it/s]
Epoch 0:  84%|████████▍ | 2767/3295 [28:51<05:30,  1.60it/s]
Epoch 0:  84%|████████▍ | 2768/3295 [28:53<05:30,  1.60it/s]
Epoch 0:  84%|████████▍ | 2768/3295 [28:53<05:30,  1.60it/s]
Epoch 0:  84%|████████▍ | 2769/3295 [28:53<05:29,  1.60it/s]
Epoch 0:  84%|████████▍ | 2769/3295 [28:53<05:29,  1.60it/s]
Epoch 0:  84%|████████▍ | 2770/3295 [28:53<05:28,  1.60it/s]
Epoch 0:  84%|████████▍ | 2770/3295 [28:53<05:28,  1.60it/s]
Epoch 0:  84%|████████▍ | 2771/3295 [28:55<05:28,  1.60it/s]
Epoch 0:  84%|████████▍ | 2771/3295 [28:55<05:28,  1.60it/s]
Epoch 0:  84%|████████▍ | 2772/3295 [28:55<05:27,  1.60it/s]
Epoch 0:  84%|████████▍ | 2772/3295 [28:55<05:27,  1.60it/s]
Epoch 0:  84%|████████▍ | 2773/3295 [28:55<05:26,  1.60it/s]
Epoch 0:  84%|████████▍ | 2773/3295 [28:55<05:26,  1.60it/s]
Epoch 0:  84%|████████▍ | 2774/3295 [28:57<05:26,  1.60it/s]
Epoch 0:  84%|████████▍ | 2774/3295 [28:57<05:26,  1.60it/s]
Epoch 0:  84%|████████▍ | 2775/3295 [28:57<05:25,  1.60it/s]
Epoch 0:  84%|████████▍ | 2775/3295 [28:57<05:25,  1.60it/s]
Epoch 0:  84%|████████▍ | 2776/3295 [28:57<05:24,  1.60it/s]
Epoch 0:  84%|████████▍ | 2776/3295 [28:57<05:24,  1.60it/s]
Epoch 0:  84%|████████▍ | 2777/3295 [28:59<05:24,  1.60it/s]
Epoch 0:  84%|████████▍ | 2777/3295 [28:59<05:24,  1.60it/s]
Epoch 0:  84%|████████▍ | 2778/3295 [28:59<05:23,  1.60it/s]
Epoch 0:  84%|████████▍ | 2778/3295 [28:59<05:23,  1.60it/s]
Epoch 0:  84%|████████▍ | 2779/3295 [28:59<05:22,  1.60it/s]
Epoch 0:  84%|████████▍ | 2779/3295 [28:59<05:22,  1.60it/s]
Epoch 0:  84%|████████▍ | 2780/3295 [29:00<05:22,  1.60it/s]
Epoch 0:  84%|████████▍ | 2780/3295 [29:00<05:22,  1.60it/s]
Epoch 0:  84%|████████▍ | 2781/3295 [29:01<05:21,  1.60it/s]
Epoch 0:  84%|████████▍ | 2781/3295 [29:01<05:21,  1.60it/s]
Epoch 0:  84%|████████▍ | 2782/3295 [29:01<05:21,  1.60it/s]
Epoch 0:  84%|████████▍ | 2782/3295 [29:01<05:21,  1.60it/s]
Epoch 0:  84%|████████▍ | 2783/3295 [29:02<05:20,  1.60it/s]
Epoch 0:  84%|████████▍ | 2783/3295 [29:02<05:20,  1.60it/s]
Epoch 0:  84%|████████▍ | 2784/3295 [29:02<05:19,  1.60it/s]
Epoch 0:  84%|████████▍ | 2784/3295 [29:02<05:19,  1.60it/s]
Epoch 0:  85%|████████▍ | 2785/3295 [29:02<05:19,  1.60it/s]
Epoch 0:  85%|████████▍ | 2785/3295 [29:02<05:19,  1.60it/s]
Epoch 0:  85%|████████▍ | 2786/3295 [29:04<05:18,  1.60it/s]
Epoch 0:  85%|████████▍ | 2786/3295 [29:04<05:18,  1.60it/s]
Epoch 0:  85%|████████▍ | 2787/3295 [29:04<05:17,  1.60it/s]
Epoch 0:  85%|████████▍ | 2787/3295 [29:04<05:17,  1.60it/s]
Epoch 0:  85%|████████▍ | 2788/3295 [29:04<05:17,  1.60it/s]
Epoch 0:  85%|████████▍ | 2788/3295 [29:04<05:17,  1.60it/s]
Epoch 0:  85%|████████▍ | 2789/3295 [29:06<05:16,  1.60it/s]
Epoch 0:  85%|████████▍ | 2789/3295 [29:06<05:16,  1.60it/s]
Epoch 0:  85%|████████▍ | 2790/3295 [29:06<05:16,  1.60it/s]
Epoch 0:  85%|████████▍ | 2790/3295 [29:06<05:16,  1.60it/s]
Epoch 0:  85%|████████▍ | 2791/3295 [29:06<05:15,  1.60it/s]
Epoch 0:  85%|████████▍ | 2791/3295 [29:06<05:15,  1.60it/s]
Epoch 0:  85%|████████▍ | 2792/3295 [29:08<05:14,  1.60it/s]
Epoch 0:  85%|████████▍ | 2792/3295 [29:08<05:14,  1.60it/s]
Epoch 0:  85%|████████▍ | 2793/3295 [29:08<05:14,  1.60it/s]
Epoch 0:  85%|████████▍ | 2793/3295 [29:08<05:14,  1.60it/s]
Epoch 0:  85%|████████▍ | 2794/3295 [29:08<05:13,  1.60it/s]
Epoch 0:  85%|████████▍ | 2794/3295 [29:08<05:13,  1.60it/s]
Epoch 0:  85%|████████▍ | 2795/3295 [29:10<05:13,  1.60it/s]
Epoch 0:  85%|████████▍ | 2795/3295 [29:10<05:13,  1.60it/s]
Epoch 0:  85%|████████▍ | 2796/3295 [29:10<05:12,  1.60it/s]
Epoch 0:  85%|████████▍ | 2796/3295 [29:10<05:12,  1.60it/s]
Epoch 0:  85%|████████▍ | 2797/3295 [29:10<05:11,  1.60it/s]
Epoch 0:  85%|████████▍ | 2797/3295 [29:10<05:11,  1.60it/s]
Epoch 0:  85%|████████▍ | 2798/3295 [29:12<05:11,  1.60it/s]
Epoch 0:  85%|████████▍ | 2798/3295 [29:12<05:11,  1.60it/s]
Epoch 0:  85%|████████▍ | 2799/3295 [29:12<05:10,  1.60it/s]
Epoch 0:  85%|████████▍ | 2799/3295 [29:12<05:10,  1.60it/s]
Epoch 0:  85%|████████▍ | 2800/3295 [29:12<05:09,  1.60it/s]
Epoch 0:  85%|████████▍ | 2800/3295 [29:12<05:09,  1.60it/s]
Epoch 0:  85%|████████▌ | 2801/3295 [29:13<05:09,  1.60it/s]
Epoch 0:  85%|████████▌ | 2801/3295 [29:13<05:09,  1.60it/s]
Epoch 0:  85%|████████▌ | 2802/3295 [29:14<05:08,  1.60it/s]
Epoch 0:  85%|████████▌ | 2802/3295 [29:14<05:08,  1.60it/s]
Epoch 0:  85%|████████▌ | 2803/3295 [29:14<05:07,  1.60it/s]
Epoch 0:  85%|████████▌ | 2803/3295 [29:14<05:07,  1.60it/s]
Epoch 0:  85%|████████▌ | 2804/3295 [29:15<05:07,  1.60it/s]
Epoch 0:  85%|████████▌ | 2804/3295 [29:15<05:07,  1.60it/s]
Epoch 0:  85%|████████▌ | 2805/3295 [29:15<05:06,  1.60it/s]
Epoch 0:  85%|████████▌ | 2805/3295 [29:15<05:06,  1.60it/s]
Epoch 0:  85%|████████▌ | 2806/3295 [29:15<05:05,  1.60it/s]
Epoch 0:  85%|████████▌ | 2806/3295 [29:15<05:05,  1.60it/s]
Epoch 0:  85%|████████▌ | 2807/3295 [29:17<05:05,  1.60it/s]
Epoch 0:  85%|████████▌ | 2807/3295 [29:17<05:05,  1.60it/s]
Epoch 0:  85%|████████▌ | 2808/3295 [29:17<05:04,  1.60it/s]
Epoch 0:  85%|████████▌ | 2808/3295 [29:17<05:04,  1.60it/s]
Epoch 0:  85%|████████▌ | 2809/3295 [29:17<05:04,  1.60it/s]
Epoch 0:  85%|████████▌ | 2809/3295 [29:17<05:04,  1.60it/s]
Epoch 0:  85%|████████▌ | 2810/3295 [29:19<05:03,  1.60it/s]
Epoch 0:  85%|████████▌ | 2810/3295 [29:19<05:03,  1.60it/s]
Epoch 0:  85%|████████▌ | 2811/3295 [29:19<05:02,  1.60it/s]
Epoch 0:  85%|████████▌ | 2811/3295 [29:19<05:02,  1.60it/s]
Epoch 0:  85%|████████▌ | 2812/3295 [29:19<05:02,  1.60it/s]
Epoch 0:  85%|████████▌ | 2812/3295 [29:19<05:02,  1.60it/s]
Epoch 0:  85%|████████▌ | 2813/3295 [29:21<05:01,  1.60it/s]
Epoch 0:  85%|████████▌ | 2813/3295 [29:21<05:01,  1.60it/s]
Epoch 0:  85%|████████▌ | 2814/3295 [29:21<05:01,  1.60it/s]
Epoch 0:  85%|████████▌ | 2814/3295 [29:21<05:01,  1.60it/s]
Epoch 0:  85%|████████▌ | 2815/3295 [29:21<05:00,  1.60it/s]
Epoch 0:  85%|████████▌ | 2815/3295 [29:21<05:00,  1.60it/s]
Epoch 0:  85%|████████▌ | 2816/3295 [29:23<04:59,  1.60it/s]
Epoch 0:  85%|████████▌ | 2816/3295 [29:23<04:59,  1.60it/s]
Epoch 0:  85%|████████▌ | 2817/3295 [29:23<04:59,  1.60it/s]
Epoch 0:  85%|████████▌ | 2817/3295 [29:23<04:59,  1.60it/s]
Epoch 0:  86%|████████▌ | 2818/3295 [29:23<04:58,  1.60it/s]
Epoch 0:  86%|████████▌ | 2818/3295 [29:23<04:58,  1.60it/s]
Epoch 0:  86%|████████▌ | 2819/3295 [29:25<04:58,  1.60it/s]
Epoch 0:  86%|████████▌ | 2819/3295 [29:25<04:58,  1.60it/s]
Epoch 0:  86%|████████▌ | 2820/3295 [29:25<04:57,  1.60it/s]
Epoch 0:  86%|████████▌ | 2820/3295 [29:25<04:57,  1.60it/s]
Epoch 0:  86%|████████▌ | 2821/3295 [29:25<04:56,  1.60it/s]
Epoch 0:  86%|████████▌ | 2821/3295 [29:25<04:56,  1.60it/s]
Epoch 0:  86%|████████▌ | 2822/3295 [29:27<04:56,  1.60it/s]
Epoch 0:  86%|████████▌ | 2822/3295 [29:27<04:56,  1.60it/s]
Epoch 0:  86%|████████▌ | 2823/3295 [29:27<04:55,  1.60it/s]
Epoch 0:  86%|████████▌ | 2823/3295 [29:27<04:55,  1.60it/s]
Epoch 0:  86%|████████▌ | 2824/3295 [29:27<04:54,  1.60it/s]
Epoch 0:  86%|████████▌ | 2824/3295 [29:27<04:54,  1.60it/s]
Epoch 0:  86%|████████▌ | 2825/3295 [29:28<04:54,  1.60it/s]
Epoch 0:  86%|████████▌ | 2825/3295 [29:28<04:54,  1.60it/s]
Epoch 0:  86%|████████▌ | 2826/3295 [29:29<04:53,  1.60it/s]
Epoch 0:  86%|████████▌ | 2826/3295 [29:29<04:53,  1.60it/s]
Epoch 0:  86%|████████▌ | 2827/3295 [29:29<04:52,  1.60it/s]
Epoch 0:  86%|████████▌ | 2827/3295 [29:29<04:52,  1.60it/s]
Epoch 0:  86%|████████▌ | 2828/3295 [29:30<04:52,  1.60it/s]
Epoch 0:  86%|████████▌ | 2828/3295 [29:30<04:52,  1.60it/s]
Epoch 0:  86%|████████▌ | 2829/3295 [29:31<04:51,  1.60it/s]
Epoch 0:  86%|████████▌ | 2829/3295 [29:31<04:51,  1.60it/s]
Epoch 0:  86%|████████▌ | 2830/3295 [29:31<04:51,  1.60it/s]
Epoch 0:  86%|████████▌ | 2830/3295 [29:31<04:51,  1.60it/s]
Epoch 0:  86%|████████▌ | 2831/3295 [29:32<04:50,  1.60it/s]
Epoch 0:  86%|████████▌ | 2831/3295 [29:32<04:50,  1.60it/s]
Epoch 0:  86%|████████▌ | 2832/3295 [29:32<04:49,  1.60it/s]
Epoch 0:  86%|████████▌ | 2832/3295 [29:32<04:49,  1.60it/s]
Epoch 0:  86%|████████▌ | 2833/3295 [29:32<04:49,  1.60it/s]
Epoch 0:  86%|████████▌ | 2833/3295 [29:32<04:49,  1.60it/s]
Epoch 0:  86%|████████▌ | 2834/3295 [29:34<04:48,  1.60it/s]
Epoch 0:  86%|████████▌ | 2834/3295 [29:34<04:48,  1.60it/s]
Epoch 0:  86%|████████▌ | 2835/3295 [29:34<04:47,  1.60it/s]
Epoch 0:  86%|████████▌ | 2835/3295 [29:34<04:47,  1.60it/s]
Epoch 0:  86%|████████▌ | 2836/3295 [29:34<04:47,  1.60it/s]
Epoch 0:  86%|████████▌ | 2836/3295 [29:34<04:47,  1.60it/s]
Epoch 0:  86%|████████▌ | 2837/3295 [29:36<04:46,  1.60it/s]
Epoch 0:  86%|████████▌ | 2837/3295 [29:36<04:46,  1.60it/s]
Epoch 0:  86%|████████▌ | 2838/3295 [29:36<04:46,  1.60it/s]
Epoch 0:  86%|████████▌ | 2838/3295 [29:36<04:46,  1.60it/s]
Epoch 0:  86%|████████▌ | 2839/3295 [29:36<04:45,  1.60it/s]
Epoch 0:  86%|████████▌ | 2839/3295 [29:36<04:45,  1.60it/s]
Epoch 0:  86%|████████▌ | 2840/3295 [29:38<04:44,  1.60it/s]
Epoch 0:  86%|████████▌ | 2840/3295 [29:38<04:44,  1.60it/s]
Epoch 0:  86%|████████▌ | 2841/3295 [29:38<04:44,  1.60it/s]
Epoch 0:  86%|████████▌ | 2841/3295 [29:38<04:44,  1.60it/s]
Epoch 0:  86%|████████▋ | 2842/3295 [29:38<04:43,  1.60it/s]
Epoch 0:  86%|████████▋ | 2842/3295 [29:38<04:43,  1.60it/s]
Epoch 0:  86%|████████▋ | 2843/3295 [29:40<04:43,  1.60it/s]
Epoch 0:  86%|████████▋ | 2843/3295 [29:40<04:43,  1.60it/s]
Epoch 0:  86%|████████▋ | 2844/3295 [29:40<04:42,  1.60it/s]
Epoch 0:  86%|████████▋ | 2844/3295 [29:40<04:42,  1.60it/s]
Epoch 0:  86%|████████▋ | 2845/3295 [29:40<04:41,  1.60it/s]
Epoch 0:  86%|████████▋ | 2845/3295 [29:40<04:41,  1.60it/s]
Epoch 0:  86%|████████▋ | 2846/3295 [29:42<04:41,  1.60it/s]
Epoch 0:  86%|████████▋ | 2846/3295 [29:42<04:41,  1.60it/s]
Epoch 0:  86%|████████▋ | 2847/3295 [29:42<04:40,  1.60it/s]
Epoch 0:  86%|████████▋ | 2847/3295 [29:42<04:40,  1.60it/s]
Epoch 0:  86%|████████▋ | 2848/3295 [29:42<04:39,  1.60it/s]
Epoch 0:  86%|████████▋ | 2848/3295 [29:42<04:39,  1.60it/s]
Epoch 0:  86%|████████▋ | 2849/3295 [29:44<04:39,  1.60it/s]
Epoch 0:  86%|████████▋ | 2849/3295 [29:44<04:39,  1.60it/s]
Epoch 0:  86%|████████▋ | 2850/3295 [29:44<04:38,  1.60it/s]
Epoch 0:  86%|████████▋ | 2850/3295 [29:44<04:38,  1.60it/s]
Epoch 0:  87%|████████▋ | 2851/3295 [29:44<04:37,  1.60it/s]
Epoch 0:  87%|████████▋ | 2851/3295 [29:44<04:37,  1.60it/s]
Epoch 0:  87%|████████▋ | 2852/3295 [29:45<04:37,  1.60it/s]
Epoch 0:  87%|████████▋ | 2852/3295 [29:45<04:37,  1.60it/s]
Epoch 0:  87%|████████▋ | 2853/3295 [29:45<04:36,  1.60it/s]
Epoch 0:  87%|████████▋ | 2853/3295 [29:45<04:36,  1.60it/s]
Epoch 0:  87%|████████▋ | 2854/3295 [29:45<04:35,  1.60it/s]
Epoch 0:  87%|████████▋ | 2854/3295 [29:45<04:35,  1.60it/s]
Epoch 0:  87%|████████▋ | 2855/3295 [29:47<04:35,  1.60it/s]
Epoch 0:  87%|████████▋ | 2855/3295 [29:47<04:35,  1.60it/s]
Epoch 0:  87%|████████▋ | 2856/3295 [29:47<04:34,  1.60it/s]
Epoch 0:  87%|████████▋ | 2856/3295 [29:47<04:34,  1.60it/s]
Epoch 0:  87%|████████▋ | 2857/3295 [29:47<04:34,  1.60it/s]
Epoch 0:  87%|████████▋ | 2857/3295 [29:47<04:34,  1.60it/s]
Epoch 0:  87%|████████▋ | 2858/3295 [29:49<04:33,  1.60it/s]
Epoch 0:  87%|████████▋ | 2858/3295 [29:49<04:33,  1.60it/s]
Epoch 0:  87%|████████▋ | 2859/3295 [29:49<04:32,  1.60it/s]
Epoch 0:  87%|████████▋ | 2859/3295 [29:49<04:32,  1.60it/s]
Epoch 0:  87%|████████▋ | 2860/3295 [29:49<04:32,  1.60it/s]
Epoch 0:  87%|████████▋ | 2860/3295 [29:49<04:32,  1.60it/s]
Epoch 0:  87%|████████▋ | 2861/3295 [29:51<04:31,  1.60it/s]
Epoch 0:  87%|████████▋ | 2861/3295 [29:51<04:31,  1.60it/s]
Epoch 0:  87%|████████▋ | 2862/3295 [29:51<04:31,  1.60it/s]
Epoch 0:  87%|████████▋ | 2862/3295 [29:51<04:31,  1.60it/s]
Epoch 0:  87%|████████▋ | 2863/3295 [29:51<04:30,  1.60it/s]
Epoch 0:  87%|████████▋ | 2863/3295 [29:51<04:30,  1.60it/s]
Epoch 0:  87%|████████▋ | 2864/3295 [29:53<04:29,  1.60it/s]
Epoch 0:  87%|████████▋ | 2864/3295 [29:53<04:29,  1.60it/s]
Epoch 0:  87%|████████▋ | 2865/3295 [29:53<04:29,  1.60it/s]
Epoch 0:  87%|████████▋ | 2865/3295 [29:53<04:29,  1.60it/s]
Epoch 0:  87%|████████▋ | 2866/3295 [29:53<04:28,  1.60it/s]
Epoch 0:  87%|████████▋ | 2866/3295 [29:53<04:28,  1.60it/s]
Epoch 0:  87%|████████▋ | 2867/3295 [29:55<04:27,  1.60it/s]
Epoch 0:  87%|████████▋ | 2867/3295 [29:55<04:27,  1.60it/s]
Epoch 0:  87%|████████▋ | 2868/3295 [29:55<04:27,  1.60it/s]
Epoch 0:  87%|████████▋ | 2868/3295 [29:55<04:27,  1.60it/s]
Epoch 0:  87%|████████▋ | 2869/3295 [29:55<04:26,  1.60it/s]
Epoch 0:  87%|████████▋ | 2869/3295 [29:55<04:26,  1.60it/s]
Epoch 0:  87%|████████▋ | 2870/3295 [29:57<04:26,  1.60it/s]
Epoch 0:  87%|████████▋ | 2870/3295 [29:57<04:26,  1.60it/s]
Epoch 0:  87%|████████▋ | 2871/3295 [29:57<04:25,  1.60it/s]
Epoch 0:  87%|████████▋ | 2871/3295 [29:57<04:25,  1.60it/s]
Epoch 0:  87%|████████▋ | 2872/3295 [29:57<04:24,  1.60it/s]
Epoch 0:  87%|████████▋ | 2872/3295 [29:57<04:24,  1.60it/s]
Epoch 0:  87%|████████▋ | 2873/3295 [29:59<04:24,  1.60it/s]
Epoch 0:  87%|████████▋ | 2873/3295 [29:59<04:24,  1.60it/s]
Epoch 0:  87%|████████▋ | 2874/3295 [29:59<04:23,  1.60it/s]
Epoch 0:  87%|████████▋ | 2874/3295 [29:59<04:23,  1.60it/s]
Epoch 0:  87%|████████▋ | 2875/3295 [29:59<04:22,  1.60it/s]
Epoch 0:  87%|████████▋ | 2875/3295 [29:59<04:22,  1.60it/s]
Epoch 0:  87%|████████▋ | 2876/3295 [30:00<04:22,  1.60it/s]
Epoch 0:  87%|████████▋ | 2876/3295 [30:00<04:22,  1.60it/s]
Epoch 0:  87%|████████▋ | 2877/3295 [30:00<04:21,  1.60it/s]
Epoch 0:  87%|████████▋ | 2877/3295 [30:00<04:21,  1.60it/s]
Epoch 0:  87%|████████▋ | 2878/3295 [30:00<04:20,  1.60it/s]
Epoch 0:  87%|████████▋ | 2878/3295 [30:00<04:20,  1.60it/s]
Epoch 0:  87%|████████▋ | 2879/3295 [30:02<04:20,  1.60it/s]
Epoch 0:  87%|████████▋ | 2879/3295 [30:02<04:20,  1.60it/s]
Epoch 0:  87%|████████▋ | 2880/3295 [30:02<04:19,  1.60it/s]
Epoch 0:  87%|████████▋ | 2880/3295 [30:02<04:19,  1.60it/s]
Epoch 0:  87%|████████▋ | 2881/3295 [30:02<04:19,  1.60it/s]
Epoch 0:  87%|████████▋ | 2881/3295 [30:02<04:19,  1.60it/s]
Epoch 0:  87%|████████▋ | 2882/3295 [30:04<04:18,  1.60it/s]
Epoch 0:  87%|████████▋ | 2882/3295 [30:04<04:18,  1.60it/s]
Epoch 0:  87%|████████▋ | 2883/3295 [30:04<04:17,  1.60it/s]
Epoch 0:  87%|████████▋ | 2883/3295 [30:04<04:17,  1.60it/s]
Epoch 0:  88%|████████▊ | 2884/3295 [30:04<04:17,  1.60it/s]
Epoch 0:  88%|████████▊ | 2884/3295 [30:04<04:17,  1.60it/s]
Epoch 0:  88%|████████▊ | 2885/3295 [30:06<04:16,  1.60it/s]
Epoch 0:  88%|████████▊ | 2885/3295 [30:06<04:16,  1.60it/s]
Epoch 0:  88%|████████▊ | 2886/3295 [30:06<04:15,  1.60it/s]
Epoch 0:  88%|████████▊ | 2886/3295 [30:06<04:15,  1.60it/s]
Epoch 0:  88%|████████▊ | 2887/3295 [30:06<04:15,  1.60it/s]
Epoch 0:  88%|████████▊ | 2887/3295 [30:06<04:15,  1.60it/s]
Epoch 0:  88%|████████▊ | 2888/3295 [30:08<04:14,  1.60it/s]
Epoch 0:  88%|████████▊ | 2888/3295 [30:08<04:14,  1.60it/s]
Epoch 0:  88%|████████▊ | 2889/3295 [30:08<04:14,  1.60it/s]
Epoch 0:  88%|████████▊ | 2889/3295 [30:08<04:14,  1.60it/s]
Epoch 0:  88%|████████▊ | 2890/3295 [30:08<04:13,  1.60it/s]
Epoch 0:  88%|████████▊ | 2890/3295 [30:08<04:13,  1.60it/s]
Epoch 0:  88%|████████▊ | 2891/3295 [30:10<04:12,  1.60it/s]
Epoch 0:  88%|████████▊ | 2891/3295 [30:10<04:12,  1.60it/s]
Epoch 0:  88%|████████▊ | 2892/3295 [30:10<04:12,  1.60it/s]
Epoch 0:  88%|████████▊ | 2892/3295 [30:10<04:12,  1.60it/s]
Epoch 0:  88%|████████▊ | 2893/3295 [30:10<04:11,  1.60it/s]
Epoch 0:  88%|████████▊ | 2893/3295 [30:10<04:11,  1.60it/s]
Epoch 0:  88%|████████▊ | 2894/3295 [30:11<04:11,  1.60it/s]
Epoch 0:  88%|████████▊ | 2894/3295 [30:11<04:11,  1.60it/s]
Epoch 0:  88%|████████▊ | 2895/3295 [30:11<04:10,  1.60it/s]
Epoch 0:  88%|████████▊ | 2895/3295 [30:11<04:10,  1.60it/s]
Epoch 0:  88%|████████▊ | 2896/3295 [30:11<04:09,  1.60it/s]
Epoch 0:  88%|████████▊ | 2896/3295 [30:11<04:09,  1.60it/s]
Epoch 0:  88%|████████▊ | 2897/3295 [30:13<04:09,  1.60it/s]
Epoch 0:  88%|████████▊ | 2897/3295 [30:13<04:09,  1.60it/s]
Epoch 0:  88%|████████▊ | 2898/3295 [30:13<04:08,  1.60it/s]
Epoch 0:  88%|████████▊ | 2898/3295 [30:13<04:08,  1.60it/s]
Epoch 0:  88%|████████▊ | 2899/3295 [30:13<04:07,  1.60it/s]
Epoch 0:  88%|████████▊ | 2899/3295 [30:13<04:07,  1.60it/s]
Epoch 0:  88%|████████▊ | 2900/3295 [30:15<04:07,  1.60it/s]
Epoch 0:  88%|████████▊ | 2900/3295 [30:15<04:07,  1.60it/s]
Epoch 0:  88%|████████▊ | 2901/3295 [30:15<04:06,  1.60it/s]
Epoch 0:  88%|████████▊ | 2901/3295 [30:15<04:06,  1.60it/s]
Epoch 0:  88%|████████▊ | 2902/3295 [30:15<04:05,  1.60it/s]
Epoch 0:  88%|████████▊ | 2902/3295 [30:15<04:05,  1.60it/s]
Epoch 0:  88%|████████▊ | 2903/3295 [30:17<04:05,  1.60it/s]
Epoch 0:  88%|████████▊ | 2903/3295 [30:17<04:05,  1.60it/s]
Epoch 0:  88%|████████▊ | 2904/3295 [30:17<04:04,  1.60it/s]
Epoch 0:  88%|████████▊ | 2904/3295 [30:17<04:04,  1.60it/s]
Epoch 0:  88%|████████▊ | 2905/3295 [30:17<04:03,  1.60it/s]
Epoch 0:  88%|████████▊ | 2905/3295 [30:17<04:03,  1.60it/s]
Epoch 0:  88%|████████▊ | 2906/3295 [30:19<04:03,  1.60it/s]
Epoch 0:  88%|████████▊ | 2906/3295 [30:19<04:03,  1.60it/s]
Epoch 0:  88%|████████▊ | 2907/3295 [30:19<04:02,  1.60it/s]
Epoch 0:  88%|████████▊ | 2907/3295 [30:19<04:02,  1.60it/s]
Epoch 0:  88%|████████▊ | 2908/3295 [30:19<04:02,  1.60it/s]
Epoch 0:  88%|████████▊ | 2908/3295 [30:19<04:02,  1.60it/s]
Epoch 0:  88%|████████▊ | 2909/3295 [30:20<04:01,  1.60it/s]
Epoch 0:  88%|████████▊ | 2909/3295 [30:20<04:01,  1.60it/s]
Epoch 0:  88%|████████▊ | 2910/3295 [30:21<04:00,  1.60it/s]
Epoch 0:  88%|████████▊ | 2910/3295 [30:21<04:00,  1.60it/s]
Epoch 0:  88%|████████▊ | 2911/3295 [30:21<04:00,  1.60it/s]
Epoch 0:  88%|████████▊ | 2911/3295 [30:21<04:00,  1.60it/s]
Epoch 0:  88%|████████▊ | 2912/3295 [30:22<03:59,  1.60it/s]
Epoch 0:  88%|████████▊ | 2912/3295 [30:22<03:59,  1.60it/s]
Epoch 0:  88%|████████▊ | 2913/3295 [30:22<03:59,  1.60it/s]
Epoch 0:  88%|████████▊ | 2913/3295 [30:22<03:59,  1.60it/s]
Epoch 0:  88%|████████▊ | 2914/3295 [30:22<03:58,  1.60it/s]
Epoch 0:  88%|████████▊ | 2914/3295 [30:22<03:58,  1.60it/s]
Epoch 0:  88%|████████▊ | 2915/3295 [30:24<03:57,  1.60it/s]
Epoch 0:  88%|████████▊ | 2915/3295 [30:24<03:57,  1.60it/s]
Epoch 0:  88%|████████▊ | 2916/3295 [30:24<03:57,  1.60it/s]
Epoch 0:  88%|████████▊ | 2916/3295 [30:24<03:57,  1.60it/s]
Epoch 0:  89%|████████▊ | 2917/3295 [30:24<03:56,  1.60it/s]
Epoch 0:  89%|████████▊ | 2917/3295 [30:24<03:56,  1.60it/s]
Epoch 0:  89%|████████▊ | 2918/3295 [30:26<03:55,  1.60it/s]
Epoch 0:  89%|████████▊ | 2918/3295 [30:26<03:55,  1.60it/s]
Epoch 0:  89%|████████▊ | 2919/3295 [30:26<03:55,  1.60it/s]
Epoch 0:  89%|████████▊ | 2919/3295 [30:26<03:55,  1.60it/s]
Epoch 0:  89%|████████▊ | 2920/3295 [30:26<03:54,  1.60it/s]
Epoch 0:  89%|████████▊ | 2920/3295 [30:26<03:54,  1.60it/s]
Epoch 0:  89%|████████▊ | 2921/3295 [30:28<03:54,  1.60it/s]
Epoch 0:  89%|████████▊ | 2921/3295 [30:28<03:54,  1.60it/s]
Epoch 0:  89%|████████▊ | 2922/3295 [30:28<03:53,  1.60it/s]
Epoch 0:  89%|████████▊ | 2922/3295 [30:28<03:53,  1.60it/s]
Epoch 0:  89%|████████▊ | 2923/3295 [30:28<03:52,  1.60it/s]
Epoch 0:  89%|████████▊ | 2923/3295 [30:28<03:52,  1.60it/s]
Epoch 0:  89%|████████▊ | 2924/3295 [30:30<03:52,  1.60it/s]
Epoch 0:  89%|████████▊ | 2924/3295 [30:30<03:52,  1.60it/s]
Epoch 0:  89%|████████▉ | 2925/3295 [30:30<03:51,  1.60it/s]
Epoch 0:  89%|████████▉ | 2925/3295 [30:30<03:51,  1.60it/s]
Epoch 0:  89%|████████▉ | 2926/3295 [30:30<03:50,  1.60it/s]
Epoch 0:  89%|████████▉ | 2926/3295 [30:30<03:50,  1.60it/s]
Epoch 0:  89%|████████▉ | 2927/3295 [30:32<03:50,  1.60it/s]
Epoch 0:  89%|████████▉ | 2927/3295 [30:32<03:50,  1.60it/s]
Epoch 0:  89%|████████▉ | 2928/3295 [30:32<03:49,  1.60it/s]
Epoch 0:  89%|████████▉ | 2928/3295 [30:32<03:49,  1.60it/s]
Epoch 0:  89%|████████▉ | 2929/3295 [30:32<03:48,  1.60it/s]
Epoch 0:  89%|████████▉ | 2929/3295 [30:32<03:48,  1.60it/s]
Epoch 0:  89%|████████▉ | 2930/3295 [30:34<03:48,  1.60it/s]
Epoch 0:  89%|████████▉ | 2930/3295 [30:34<03:48,  1.60it/s]
Epoch 0:  89%|████████▉ | 2931/3295 [30:34<03:47,  1.60it/s]
Epoch 0:  89%|████████▉ | 2931/3295 [30:34<03:47,  1.60it/s]
Epoch 0:  89%|████████▉ | 2932/3295 [30:34<03:47,  1.60it/s]
Epoch 0:  89%|████████▉ | 2932/3295 [30:34<03:47,  1.60it/s]
Epoch 0:  89%|████████▉ | 2933/3295 [30:35<03:46,  1.60it/s]
Epoch 0:  89%|████████▉ | 2933/3295 [30:35<03:46,  1.60it/s]
Epoch 0:  89%|████████▉ | 2934/3295 [30:35<03:45,  1.60it/s]
Epoch 0:  89%|████████▉ | 2934/3295 [30:35<03:45,  1.60it/s]
Epoch 0:  89%|████████▉ | 2935/3295 [30:35<03:45,  1.60it/s]
Epoch 0:  89%|████████▉ | 2935/3295 [30:35<03:45,  1.60it/s]
Epoch 0:  89%|████████▉ | 2936/3295 [30:37<03:44,  1.60it/s]
Epoch 0:  89%|████████▉ | 2936/3295 [30:37<03:44,  1.60it/s]
Epoch 0:  89%|████████▉ | 2937/3295 [30:37<03:44,  1.60it/s]
Epoch 0:  89%|████████▉ | 2937/3295 [30:37<03:44,  1.60it/s]
Epoch 0:  89%|████████▉ | 2938/3295 [30:37<03:43,  1.60it/s]
Epoch 0:  89%|████████▉ | 2938/3295 [30:37<03:43,  1.60it/s]
Epoch 0:  89%|████████▉ | 2939/3295 [30:39<03:42,  1.60it/s]
Epoch 0:  89%|████████▉ | 2939/3295 [30:39<03:42,  1.60it/s]
Epoch 0:  89%|████████▉ | 2940/3295 [30:39<03:42,  1.60it/s]
Epoch 0:  89%|████████▉ | 2940/3295 [30:39<03:42,  1.60it/s]
Epoch 0:  89%|████████▉ | 2941/3295 [30:39<03:41,  1.60it/s]
Epoch 0:  89%|████████▉ | 2941/3295 [30:39<03:41,  1.60it/s]
Epoch 0:  89%|████████▉ | 2942/3295 [30:41<03:40,  1.60it/s]
Epoch 0:  89%|████████▉ | 2942/3295 [30:41<03:40,  1.60it/s]
Epoch 0:  89%|████████▉ | 2943/3295 [30:41<03:40,  1.60it/s]
Epoch 0:  89%|████████▉ | 2943/3295 [30:41<03:40,  1.60it/s]
Epoch 0:  89%|████████▉ | 2944/3295 [30:41<03:39,  1.60it/s]
Epoch 0:  89%|████████▉ | 2944/3295 [30:41<03:39,  1.60it/s]
Epoch 0:  89%|████████▉ | 2945/3295 [30:43<03:39,  1.60it/s]
Epoch 0:  89%|████████▉ | 2945/3295 [30:43<03:39,  1.60it/s]
Epoch 0:  89%|████████▉ | 2946/3295 [30:43<03:38,  1.60it/s]
Epoch 0:  89%|████████▉ | 2946/3295 [30:43<03:38,  1.60it/s]
Epoch 0:  89%|████████▉ | 2947/3295 [30:43<03:37,  1.60it/s]
Epoch 0:  89%|████████▉ | 2947/3295 [30:43<03:37,  1.60it/s]
Epoch 0:  89%|████████▉ | 2948/3295 [30:44<03:37,  1.60it/s]
Epoch 0:  89%|████████▉ | 2948/3295 [30:44<03:37,  1.60it/s]
Epoch 0:  89%|████████▉ | 2949/3295 [30:44<03:36,  1.60it/s]
Epoch 0:  89%|████████▉ | 2949/3295 [30:44<03:36,  1.60it/s]
Epoch 0:  90%|████████▉ | 2950/3295 [30:44<03:35,  1.60it/s]
Epoch 0:  90%|████████▉ | 2950/3295 [30:44<03:35,  1.60it/s]
Epoch 0:  90%|████████▉ | 2951/3295 [30:46<03:35,  1.60it/s]
Epoch 0:  90%|████████▉ | 2951/3295 [30:46<03:35,  1.60it/s]
Epoch 0:  90%|████████▉ | 2952/3295 [30:46<03:34,  1.60it/s]
Epoch 0:  90%|████████▉ | 2952/3295 [30:46<03:34,  1.60it/s]
Epoch 0:  90%|████████▉ | 2953/3295 [30:46<03:33,  1.60it/s]
Epoch 0:  90%|████████▉ | 2953/3295 [30:46<03:33,  1.60it/s]
Epoch 0:  90%|████████▉ | 2954/3295 [30:48<03:33,  1.60it/s]
Epoch 0:  90%|████████▉ | 2954/3295 [30:48<03:33,  1.60it/s]
Epoch 0:  90%|████████▉ | 2955/3295 [30:48<03:32,  1.60it/s]
Epoch 0:  90%|████████▉ | 2955/3295 [30:48<03:32,  1.60it/s]
Epoch 0:  90%|████████▉ | 2956/3295 [30:48<03:32,  1.60it/s]
Epoch 0:  90%|████████▉ | 2956/3295 [30:48<03:32,  1.60it/s]
Epoch 0:  90%|████████▉ | 2957/3295 [30:50<03:31,  1.60it/s]
Epoch 0:  90%|████████▉ | 2957/3295 [30:50<03:31,  1.60it/s]
Epoch 0:  90%|████████▉ | 2958/3295 [30:50<03:30,  1.60it/s]
Epoch 0:  90%|████████▉ | 2958/3295 [30:50<03:30,  1.60it/s]
Epoch 0:  90%|████████▉ | 2959/3295 [30:50<03:30,  1.60it/s]
Epoch 0:  90%|████████▉ | 2959/3295 [30:50<03:30,  1.60it/s]
Epoch 0:  90%|████████▉ | 2960/3295 [30:52<03:29,  1.60it/s]
Epoch 0:  90%|████████▉ | 2960/3295 [30:52<03:29,  1.60it/s]
Epoch 0:  90%|████████▉ | 2961/3295 [30:52<03:28,  1.60it/s]
Epoch 0:  90%|████████▉ | 2961/3295 [30:52<03:28,  1.60it/s]
Epoch 0:  90%|████████▉ | 2962/3295 [30:52<03:28,  1.60it/s]
Epoch 0:  90%|████████▉ | 2962/3295 [30:52<03:28,  1.60it/s]
Epoch 0:  90%|████████▉ | 2963/3295 [30:54<03:27,  1.60it/s]
Epoch 0:  90%|████████▉ | 2963/3295 [30:54<03:27,  1.60it/s]
Epoch 0:  90%|████████▉ | 2964/3295 [30:54<03:27,  1.60it/s]
Epoch 0:  90%|████████▉ | 2964/3295 [30:54<03:27,  1.60it/s]
Epoch 0:  90%|████████▉ | 2965/3295 [30:54<03:26,  1.60it/s]
Epoch 0:  90%|████████▉ | 2965/3295 [30:54<03:26,  1.60it/s]
Epoch 0:  90%|█████████ | 2966/3295 [30:56<03:25,  1.60it/s]
Epoch 0:  90%|█████████ | 2966/3295 [30:56<03:25,  1.60it/s]
Epoch 0:  90%|█████████ | 2967/3295 [30:56<03:25,  1.60it/s]
Epoch 0:  90%|█████████ | 2967/3295 [30:56<03:25,  1.60it/s]
Epoch 0:  90%|█████████ | 2968/3295 [30:56<03:24,  1.60it/s]
Epoch 0:  90%|█████████ | 2968/3295 [30:56<03:24,  1.60it/s]
Epoch 0:  90%|█████████ | 2969/3295 [30:57<03:24,  1.60it/s]
Epoch 0:  90%|█████████ | 2969/3295 [30:57<03:24,  1.60it/s]
Epoch 0:  90%|█████████ | 2970/3295 [30:57<03:23,  1.60it/s]
Epoch 0:  90%|█████████ | 2970/3295 [30:57<03:23,  1.60it/s]
Epoch 0:  90%|█████████ | 2971/3295 [30:57<03:22,  1.60it/s]
Epoch 0:  90%|█████████ | 2971/3295 [30:57<03:22,  1.60it/s]
Epoch 0:  90%|█████████ | 2972/3295 [30:59<03:22,  1.60it/s]
Epoch 0:  90%|█████████ | 2972/3295 [30:59<03:22,  1.60it/s]
Epoch 0:  90%|█████████ | 2973/3295 [30:59<03:21,  1.60it/s]
Epoch 0:  90%|█████████ | 2973/3295 [30:59<03:21,  1.60it/s]
Epoch 0:  90%|█████████ | 2974/3295 [30:59<03:20,  1.60it/s]
Epoch 0:  90%|█████████ | 2974/3295 [30:59<03:20,  1.60it/s]
Epoch 0:  90%|█████████ | 2975/3295 [31:01<03:20,  1.60it/s]
Epoch 0:  90%|█████████ | 2975/3295 [31:01<03:20,  1.60it/s]
Epoch 0:  90%|█████████ | 2976/3295 [31:01<03:19,  1.60it/s]
Epoch 0:  90%|█████████ | 2976/3295 [31:01<03:19,  1.60it/s]
Epoch 0:  90%|█████████ | 2977/3295 [31:01<03:18,  1.60it/s]
Epoch 0:  90%|█████████ | 2977/3295 [31:01<03:18,  1.60it/s]
Epoch 0:  90%|█████████ | 2978/3295 [31:03<03:18,  1.60it/s]
Epoch 0:  90%|█████████ | 2978/3295 [31:03<03:18,  1.60it/s]
Epoch 0:  90%|█████████ | 2979/3295 [31:03<03:17,  1.60it/s]
Epoch 0:  90%|█████████ | 2979/3295 [31:03<03:17,  1.60it/s]
Epoch 0:  90%|█████████ | 2980/3295 [31:03<03:17,  1.60it/s]
Epoch 0:  90%|█████████ | 2980/3295 [31:03<03:17,  1.60it/s]
Epoch 0:  90%|█████████ | 2981/3295 [31:05<03:16,  1.60it/s]
Epoch 0:  90%|█████████ | 2981/3295 [31:05<03:16,  1.60it/s]
Epoch 0:  91%|█████████ | 2982/3295 [31:05<03:15,  1.60it/s]
Epoch 0:  91%|█████████ | 2982/3295 [31:05<03:15,  1.60it/s]
Epoch 0:  91%|█████████ | 2983/3295 [31:05<03:15,  1.60it/s]
Epoch 0:  91%|█████████ | 2983/3295 [31:05<03:15,  1.60it/s]
Epoch 0:  91%|█████████ | 2984/3295 [31:07<03:14,  1.60it/s]
Epoch 0:  91%|█████████ | 2984/3295 [31:07<03:14,  1.60it/s]
Epoch 0:  91%|█████████ | 2985/3295 [31:07<03:13,  1.60it/s]
Epoch 0:  91%|█████████ | 2985/3295 [31:07<03:13,  1.60it/s]
Epoch 0:  91%|█████████ | 2986/3295 [31:07<03:13,  1.60it/s]
Epoch 0:  91%|█████████ | 2986/3295 [31:07<03:13,  1.60it/s]
Epoch 0:  91%|█████████ | 2987/3295 [31:09<03:12,  1.60it/s]
Epoch 0:  91%|█████████ | 2987/3295 [31:09<03:12,  1.60it/s]
Epoch 0:  91%|█████████ | 2988/3295 [31:09<03:12,  1.60it/s]
Epoch 0:  91%|█████████ | 2988/3295 [31:09<03:12,  1.60it/s]
Epoch 0:  91%|█████████ | 2989/3295 [31:09<03:11,  1.60it/s]
Epoch 0:  91%|█████████ | 2989/3295 [31:09<03:11,  1.60it/s]
Epoch 0:  91%|█████████ | 2990/3295 [31:11<03:10,  1.60it/s]
Epoch 0:  91%|█████████ | 2990/3295 [31:11<03:10,  1.60it/s]
Epoch 0:  91%|█████████ | 2991/3295 [31:11<03:10,  1.60it/s]
Epoch 0:  91%|█████████ | 2991/3295 [31:11<03:10,  1.60it/s]
Epoch 0:  91%|█████████ | 2992/3295 [31:11<03:09,  1.60it/s]
Epoch 0:  91%|█████████ | 2992/3295 [31:11<03:09,  1.60it/s]
Epoch 0:  91%|█████████ | 2993/3295 [31:12<03:08,  1.60it/s]
Epoch 0:  91%|█████████ | 2993/3295 [31:12<03:08,  1.60it/s]
Epoch 0:  91%|█████████ | 2994/3295 [31:12<03:08,  1.60it/s]
Epoch 0:  91%|█████████ | 2994/3295 [31:12<03:08,  1.60it/s]
Epoch 0:  91%|█████████ | 2995/3295 [31:12<03:07,  1.60it/s]
Epoch 0:  91%|█████████ | 2995/3295 [31:12<03:07,  1.60it/s]
Epoch 0:  91%|█████████ | 2996/3295 [31:14<03:07,  1.60it/s]
Epoch 0:  91%|█████████ | 2996/3295 [31:14<03:07,  1.60it/s]
Epoch 0:  91%|█████████ | 2997/3295 [31:14<03:06,  1.60it/s]
Epoch 0:  91%|█████████ | 2997/3295 [31:14<03:06,  1.60it/s]
Epoch 0:  91%|█████████ | 2998/3295 [31:14<03:05,  1.60it/s]
Epoch 0:  91%|█████████ | 2998/3295 [31:14<03:05,  1.60it/s]
Epoch 0:  91%|█████████ | 2999/3295 [31:16<03:05,  1.60it/s]
Epoch 0:  91%|█████████ | 2999/3295 [31:16<03:05,  1.60it/s]
Epoch 0:  91%|█████████ | 3000/3295 [31:16<03:04,  1.60it/s]
Epoch 0:  91%|█████████ | 3000/3295 [31:16<03:04,  1.60it/s]
Epoch 0:  91%|█████████ | 3001/3295 [31:16<03:03,  1.60it/s]
Epoch 0:  91%|█████████ | 3001/3295 [31:16<03:03,  1.60it/s]
Epoch 0:  91%|█████████ | 3002/3295 [31:18<03:03,  1.60it/s]
Epoch 0:  91%|█████████ | 3002/3295 [31:18<03:03,  1.60it/s]
Epoch 0:  91%|█████████ | 3003/3295 [31:18<03:02,  1.60it/s]
Epoch 0:  91%|█████████ | 3003/3295 [31:18<03:02,  1.60it/s]
Epoch 0:  91%|█████████ | 3004/3295 [31:18<03:01,  1.60it/s]
Epoch 0:  91%|█████████ | 3004/3295 [31:18<03:01,  1.60it/s]
Epoch 0:  91%|█████████ | 3005/3295 [31:20<03:01,  1.60it/s]
Epoch 0:  91%|█████████ | 3005/3295 [31:20<03:01,  1.60it/s]
Epoch 0:  91%|█████████ | 3006/3295 [31:20<03:00,  1.60it/s]
Epoch 0:  91%|█████████ | 3006/3295 [31:20<03:00,  1.60it/s]
Epoch 0:  91%|█████████▏| 3007/3295 [31:20<03:00,  1.60it/s]
Epoch 0:  91%|█████████▏| 3007/3295 [31:20<03:00,  1.60it/s]
Epoch 0:  91%|█████████▏| 3008/3295 [31:21<02:59,  1.60it/s]
Epoch 0:  91%|█████████▏| 3008/3295 [31:21<02:59,  1.60it/s]
Epoch 0:  91%|█████████▏| 3009/3295 [31:21<02:58,  1.60it/s]
Epoch 0:  91%|█████████▏| 3009/3295 [31:21<02:58,  1.60it/s]
Epoch 0:  91%|█████████▏| 3010/3295 [31:21<02:58,  1.60it/s]
Epoch 0:  91%|█████████▏| 3010/3295 [31:21<02:58,  1.60it/s]
Epoch 0:  91%|█████████▏| 3011/3295 [31:23<02:57,  1.60it/s]
Epoch 0:  91%|█████████▏| 3011/3295 [31:23<02:57,  1.60it/s]
Epoch 0:  91%|█████████▏| 3012/3295 [31:23<02:56,  1.60it/s]
Epoch 0:  91%|█████████▏| 3012/3295 [31:23<02:56,  1.60it/s]
Epoch 0:  91%|█████████▏| 3013/3295 [31:23<02:56,  1.60it/s]
Epoch 0:  91%|█████████▏| 3013/3295 [31:23<02:56,  1.60it/s]
Epoch 0:  91%|█████████▏| 3014/3295 [31:25<02:55,  1.60it/s]
Epoch 0:  91%|█████████▏| 3014/3295 [31:25<02:55,  1.60it/s]
Epoch 0:  92%|█████████▏| 3015/3295 [31:25<02:55,  1.60it/s]
Epoch 0:  92%|█████████▏| 3015/3295 [31:25<02:55,  1.60it/s]
Epoch 0:  92%|█████████▏| 3016/3295 [31:25<02:54,  1.60it/s]
Epoch 0:  92%|█████████▏| 3016/3295 [31:25<02:54,  1.60it/s]
Epoch 0:  92%|█████████▏| 3017/3295 [31:27<02:53,  1.60it/s]
Epoch 0:  92%|█████████▏| 3017/3295 [31:27<02:53,  1.60it/s]
Epoch 0:  92%|█████████▏| 3018/3295 [31:27<02:53,  1.60it/s]
Epoch 0:  92%|█████████▏| 3018/3295 [31:27<02:53,  1.60it/s]
Epoch 0:  92%|█████████▏| 3019/3295 [31:27<02:52,  1.60it/s]
Epoch 0:  92%|█████████▏| 3019/3295 [31:27<02:52,  1.60it/s]
Epoch 0:  92%|█████████▏| 3020/3295 [31:29<02:52,  1.60it/s]
Epoch 0:  92%|█████████▏| 3020/3295 [31:29<02:52,  1.60it/s]
Epoch 0:  92%|█████████▏| 3021/3295 [31:29<02:51,  1.60it/s]
Epoch 0:  92%|█████████▏| 3021/3295 [31:29<02:51,  1.60it/s]
Epoch 0:  92%|█████████▏| 3022/3295 [31:29<02:50,  1.60it/s]
Epoch 0:  92%|█████████▏| 3022/3295 [31:29<02:50,  1.60it/s]
Epoch 0:  92%|█████████▏| 3023/3295 [31:31<02:50,  1.60it/s]
Epoch 0:  92%|█████████▏| 3023/3295 [31:31<02:50,  1.60it/s]
Epoch 0:  92%|█████████▏| 3024/3295 [31:31<02:49,  1.60it/s]
Epoch 0:  92%|█████████▏| 3024/3295 [31:31<02:49,  1.60it/s]
Epoch 0:  92%|█████████▏| 3025/3295 [31:31<02:48,  1.60it/s]
Epoch 0:  92%|█████████▏| 3025/3295 [31:31<02:48,  1.60it/s]
Epoch 0:  92%|█████████▏| 3026/3295 [31:32<02:48,  1.60it/s]
Epoch 0:  92%|█████████▏| 3026/3295 [31:32<02:48,  1.60it/s]
Epoch 0:  92%|█████████▏| 3027/3295 [31:32<02:47,  1.60it/s]
Epoch 0:  92%|█████████▏| 3027/3295 [31:32<02:47,  1.60it/s]
Epoch 0:  92%|█████████▏| 3028/3295 [31:33<02:46,  1.60it/s]
Epoch 0:  92%|█████████▏| 3028/3295 [31:33<02:46,  1.60it/s]
Epoch 0:  92%|█████████▏| 3029/3295 [31:34<02:46,  1.60it/s]
Epoch 0:  92%|█████████▏| 3029/3295 [31:34<02:46,  1.60it/s]
Epoch 0:  92%|█████████▏| 3030/3295 [31:35<02:45,  1.60it/s]
Epoch 0:  92%|█████████▏| 3030/3295 [31:35<02:45,  1.60it/s]
Epoch 0:  92%|█████████▏| 3031/3295 [31:35<02:45,  1.60it/s]
Epoch 0:  92%|█████████▏| 3031/3295 [31:35<02:45,  1.60it/s]
Epoch 0:  92%|█████████▏| 3032/3295 [31:36<02:44,  1.60it/s]
Epoch 0:  92%|█████████▏| 3032/3295 [31:36<02:44,  1.60it/s]
Epoch 0:  92%|█████████▏| 3033/3295 [31:36<02:43,  1.60it/s]
Epoch 0:  92%|█████████▏| 3033/3295 [31:36<02:43,  1.60it/s]
Epoch 0:  92%|█████████▏| 3034/3295 [31:36<02:43,  1.60it/s]
Epoch 0:  92%|█████████▏| 3034/3295 [31:36<02:43,  1.60it/s]
Epoch 0:  92%|█████████▏| 3035/3295 [31:38<02:42,  1.60it/s]
Epoch 0:  92%|█████████▏| 3035/3295 [31:38<02:42,  1.60it/s]
Epoch 0:  92%|█████████▏| 3036/3295 [31:38<02:41,  1.60it/s]
Epoch 0:  92%|█████████▏| 3036/3295 [31:38<02:41,  1.60it/s]
Epoch 0:  92%|█████████▏| 3037/3295 [31:38<02:41,  1.60it/s]
Epoch 0:  92%|█████████▏| 3037/3295 [31:38<02:41,  1.60it/s]
Epoch 0:  92%|█████████▏| 3038/3295 [31:40<02:40,  1.60it/s]
Epoch 0:  92%|█████████▏| 3038/3295 [31:40<02:40,  1.60it/s]
Epoch 0:  92%|█████████▏| 3039/3295 [31:40<02:40,  1.60it/s]
Epoch 0:  92%|█████████▏| 3039/3295 [31:40<02:40,  1.60it/s]
Epoch 0:  92%|█████████▏| 3040/3295 [31:40<02:39,  1.60it/s]
Epoch 0:  92%|█████████▏| 3040/3295 [31:40<02:39,  1.60it/s]
Epoch 0:  92%|█████████▏| 3041/3295 [31:42<02:38,  1.60it/s]
Epoch 0:  92%|█████████▏| 3041/3295 [31:42<02:38,  1.60it/s]
Epoch 0:  92%|█████████▏| 3042/3295 [31:42<02:38,  1.60it/s]
Epoch 0:  92%|█████████▏| 3042/3295 [31:42<02:38,  1.60it/s]
Epoch 0:  92%|█████████▏| 3043/3295 [31:42<02:37,  1.60it/s]
Epoch 0:  92%|█████████▏| 3043/3295 [31:42<02:37,  1.60it/s]
Epoch 0:  92%|█████████▏| 3044/3295 [31:44<02:37,  1.60it/s]
Epoch 0:  92%|█████████▏| 3044/3295 [31:44<02:37,  1.60it/s]
Epoch 0:  92%|█████████▏| 3045/3295 [31:44<02:36,  1.60it/s]
Epoch 0:  92%|█████████▏| 3045/3295 [31:44<02:36,  1.60it/s]
Epoch 0:  92%|█████████▏| 3046/3295 [31:44<02:35,  1.60it/s]
Epoch 0:  92%|█████████▏| 3046/3295 [31:44<02:35,  1.60it/s]
Epoch 0:  92%|█████████▏| 3047/3295 [31:46<02:35,  1.60it/s]
Epoch 0:  92%|█████████▏| 3047/3295 [31:46<02:35,  1.60it/s]
Epoch 0:  93%|█████████▎| 3048/3295 [31:46<02:34,  1.60it/s]
Epoch 0:  93%|█████████▎| 3048/3295 [31:46<02:34,  1.60it/s]
Epoch 0:  93%|█████████▎| 3049/3295 [31:46<02:33,  1.60it/s]
Epoch 0:  93%|█████████▎| 3049/3295 [31:46<02:33,  1.60it/s]
Epoch 0:  93%|█████████▎| 3050/3295 [31:47<02:33,  1.60it/s]
Epoch 0:  93%|█████████▎| 3050/3295 [31:47<02:33,  1.60it/s]
Epoch 0:  93%|█████████▎| 3051/3295 [31:47<02:32,  1.60it/s]
Epoch 0:  93%|█████████▎| 3051/3295 [31:47<02:32,  1.60it/s]
Epoch 0:  93%|█████████▎| 3052/3295 [31:47<02:31,  1.60it/s]
Epoch 0:  93%|█████████▎| 3052/3295 [31:47<02:31,  1.60it/s]
Epoch 0:  93%|█████████▎| 3053/3295 [31:49<02:31,  1.60it/s]
Epoch 0:  93%|█████████▎| 3053/3295 [31:49<02:31,  1.60it/s]
Epoch 0:  93%|█████████▎| 3054/3295 [31:49<02:30,  1.60it/s]
Epoch 0:  93%|█████████▎| 3054/3295 [31:49<02:30,  1.60it/s]
Epoch 0:  93%|█████████▎| 3055/3295 [31:49<02:30,  1.60it/s]
Epoch 0:  93%|█████████▎| 3055/3295 [31:49<02:30,  1.60it/s]
Epoch 0:  93%|█████████▎| 3056/3295 [31:51<02:29,  1.60it/s]
Epoch 0:  93%|█████████▎| 3056/3295 [31:51<02:29,  1.60it/s]
Epoch 0:  93%|█████████▎| 3057/3295 [31:51<02:28,  1.60it/s]
Epoch 0:  93%|█████████▎| 3057/3295 [31:51<02:28,  1.60it/s]
Epoch 0:  93%|█████████▎| 3058/3295 [31:51<02:28,  1.60it/s]
Epoch 0:  93%|█████████▎| 3058/3295 [31:51<02:28,  1.60it/s]
Epoch 0:  93%|█████████▎| 3059/3295 [31:53<02:27,  1.60it/s]
Epoch 0:  93%|█████████▎| 3059/3295 [31:53<02:27,  1.60it/s]
Epoch 0:  93%|█████████▎| 3060/3295 [31:53<02:26,  1.60it/s]
Epoch 0:  93%|█████████▎| 3060/3295 [31:53<02:26,  1.60it/s]
Epoch 0:  93%|█████████▎| 3061/3295 [31:53<02:26,  1.60it/s]
Epoch 0:  93%|█████████▎| 3061/3295 [31:53<02:26,  1.60it/s]
Epoch 0:  93%|█████████▎| 3062/3295 [31:55<02:25,  1.60it/s]
Epoch 0:  93%|█████████▎| 3062/3295 [31:55<02:25,  1.60it/s]
Epoch 0:  93%|█████████▎| 3063/3295 [31:55<02:25,  1.60it/s]
Epoch 0:  93%|█████████▎| 3063/3295 [31:55<02:25,  1.60it/s]
Epoch 0:  93%|█████████▎| 3064/3295 [31:55<02:24,  1.60it/s]
Epoch 0:  93%|█████████▎| 3064/3295 [31:55<02:24,  1.60it/s]
Epoch 0:  93%|█████████▎| 3065/3295 [31:57<02:23,  1.60it/s]
Epoch 0:  93%|█████████▎| 3065/3295 [31:57<02:23,  1.60it/s]
Epoch 0:  93%|█████████▎| 3066/3295 [31:57<02:23,  1.60it/s]
Epoch 0:  93%|█████████▎| 3066/3295 [31:57<02:23,  1.60it/s]
Epoch 0:  93%|█████████▎| 3067/3295 [31:57<02:22,  1.60it/s]
Epoch 0:  93%|█████████▎| 3067/3295 [31:57<02:22,  1.60it/s]
Epoch 0:  93%|█████████▎| 3068/3295 [31:59<02:21,  1.60it/s]
Epoch 0:  93%|█████████▎| 3068/3295 [31:59<02:21,  1.60it/s]
Epoch 0:  93%|█████████▎| 3069/3295 [31:59<02:21,  1.60it/s]
Epoch 0:  93%|█████████▎| 3069/3295 [31:59<02:21,  1.60it/s]
Epoch 0:  93%|█████████▎| 3070/3295 [31:59<02:20,  1.60it/s]
Epoch 0:  93%|█████████▎| 3070/3295 [31:59<02:20,  1.60it/s]
Epoch 0:  93%|█████████▎| 3071/3295 [32:01<02:20,  1.60it/s]
Epoch 0:  93%|█████████▎| 3071/3295 [32:01<02:20,  1.60it/s]
Epoch 0:  93%|█████████▎| 3072/3295 [32:01<02:19,  1.60it/s]
Epoch 0:  93%|█████████▎| 3072/3295 [32:01<02:19,  1.60it/s]
Epoch 0:  93%|█████████▎| 3073/3295 [32:01<02:18,  1.60it/s]
Epoch 0:  93%|█████████▎| 3073/3295 [32:01<02:18,  1.60it/s]
Epoch 0:  93%|█████████▎| 3074/3295 [32:02<02:18,  1.60it/s]
Epoch 0:  93%|█████████▎| 3074/3295 [32:02<02:18,  1.60it/s]
Epoch 0:  93%|█████████▎| 3075/3295 [32:02<02:17,  1.60it/s]
Epoch 0:  93%|█████████▎| 3075/3295 [32:02<02:17,  1.60it/s]
Epoch 0:  93%|█████████▎| 3076/3295 [32:02<02:16,  1.60it/s]
Epoch 0:  93%|█████████▎| 3076/3295 [32:02<02:16,  1.60it/s]
Epoch 0:  93%|█████████▎| 3077/3295 [32:04<02:16,  1.60it/s]
Epoch 0:  93%|█████████▎| 3077/3295 [32:04<02:16,  1.60it/s]
Epoch 0:  93%|█████████▎| 3078/3295 [32:04<02:15,  1.60it/s]
Epoch 0:  93%|█████████▎| 3078/3295 [32:04<02:15,  1.60it/s]
Epoch 0:  93%|█████████▎| 3079/3295 [32:04<02:15,  1.60it/s]
Epoch 0:  93%|█████████▎| 3079/3295 [32:04<02:15,  1.60it/s]
Epoch 0:  93%|█████████▎| 3080/3295 [32:06<02:14,  1.60it/s]
Epoch 0:  93%|█████████▎| 3080/3295 [32:06<02:14,  1.60it/s]
Epoch 0:  94%|█████████▎| 3081/3295 [32:06<02:13,  1.60it/s]
Epoch 0:  94%|█████████▎| 3081/3295 [32:06<02:13,  1.60it/s]
Epoch 0:  94%|█████████▎| 3082/3295 [32:06<02:13,  1.60it/s]
Epoch 0:  94%|█████████▎| 3082/3295 [32:06<02:13,  1.60it/s]
Epoch 0:  94%|█████████▎| 3083/3295 [32:08<02:12,  1.60it/s]
Epoch 0:  94%|█████████▎| 3083/3295 [32:08<02:12,  1.60it/s]
Epoch 0:  94%|█████████▎| 3084/3295 [32:08<02:11,  1.60it/s]
Epoch 0:  94%|█████████▎| 3084/3295 [32:08<02:11,  1.60it/s]
Epoch 0:  94%|█████████▎| 3085/3295 [32:08<02:11,  1.60it/s]
Epoch 0:  94%|█████████▎| 3085/3295 [32:08<02:11,  1.60it/s]
Epoch 0:  94%|█████████▎| 3086/3295 [32:10<02:10,  1.60it/s]
Epoch 0:  94%|█████████▎| 3086/3295 [32:10<02:10,  1.60it/s]
Epoch 0:  94%|█████████▎| 3087/3295 [32:10<02:10,  1.60it/s]
Epoch 0:  94%|█████████▎| 3087/3295 [32:10<02:10,  1.60it/s]
Epoch 0:  94%|█████████▎| 3088/3295 [32:10<02:09,  1.60it/s]
Epoch 0:  94%|█████████▎| 3088/3295 [32:10<02:09,  1.60it/s]
Epoch 0:  94%|█████████▎| 3089/3295 [32:12<02:08,  1.60it/s]
Epoch 0:  94%|█████████▎| 3089/3295 [32:12<02:08,  1.60it/s]
Epoch 0:  94%|█████████▍| 3090/3295 [32:12<02:08,  1.60it/s]
Epoch 0:  94%|█████████▍| 3090/3295 [32:12<02:08,  1.60it/s]
Epoch 0:  94%|█████████▍| 3091/3295 [32:12<02:07,  1.60it/s]
Epoch 0:  94%|█████████▍| 3091/3295 [32:12<02:07,  1.60it/s]
Epoch 0:  94%|█████████▍| 3092/3295 [32:13<02:06,  1.60it/s]
Epoch 0:  94%|█████████▍| 3092/3295 [32:13<02:06,  1.60it/s]
Epoch 0:  94%|█████████▍| 3093/3295 [32:13<02:06,  1.60it/s]
Epoch 0:  94%|█████████▍| 3093/3295 [32:13<02:06,  1.60it/s]
Epoch 0:  94%|█████████▍| 3094/3295 [32:13<02:05,  1.60it/s]
Epoch 0:  94%|█████████▍| 3094/3295 [32:13<02:05,  1.60it/s]
Epoch 0:  94%|█████████▍| 3095/3295 [32:15<02:05,  1.60it/s]
Epoch 0:  94%|█████████▍| 3095/3295 [32:15<02:05,  1.60it/s]
Epoch 0:  94%|█████████▍| 3096/3295 [32:15<02:04,  1.60it/s]
Epoch 0:  94%|█████████▍| 3096/3295 [32:15<02:04,  1.60it/s]
Epoch 0:  94%|█████████▍| 3097/3295 [32:15<02:03,  1.60it/s]
Epoch 0:  94%|█████████▍| 3097/3295 [32:15<02:03,  1.60it/s]
Epoch 0:  94%|█████████▍| 3098/3295 [32:17<02:03,  1.60it/s]
Epoch 0:  94%|█████████▍| 3098/3295 [32:17<02:03,  1.60it/s]
Epoch 0:  94%|█████████▍| 3099/3295 [32:17<02:02,  1.60it/s]
Epoch 0:  94%|█████████▍| 3099/3295 [32:17<02:02,  1.60it/s]
Epoch 0:  94%|█████████▍| 3100/3295 [32:17<02:01,  1.60it/s]
Epoch 0:  94%|█████████▍| 3100/3295 [32:17<02:01,  1.60it/s]
Epoch 0:  94%|█████████▍| 3101/3295 [32:19<02:01,  1.60it/s]
Epoch 0:  94%|█████████▍| 3101/3295 [32:19<02:01,  1.60it/s]
Epoch 0:  94%|█████████▍| 3102/3295 [32:19<02:00,  1.60it/s]
Epoch 0:  94%|█████████▍| 3102/3295 [32:19<02:00,  1.60it/s]
Epoch 0:  94%|█████████▍| 3103/3295 [32:19<02:00,  1.60it/s]
Epoch 0:  94%|█████████▍| 3103/3295 [32:19<02:00,  1.60it/s]
Epoch 0:  94%|█████████▍| 3104/3295 [32:21<01:59,  1.60it/s]
Epoch 0:  94%|█████████▍| 3104/3295 [32:21<01:59,  1.60it/s]
Epoch 0:  94%|█████████▍| 3105/3295 [32:21<01:58,  1.60it/s]
Epoch 0:  94%|█████████▍| 3105/3295 [32:21<01:58,  1.60it/s]
Epoch 0:  94%|█████████▍| 3106/3295 [32:21<01:58,  1.60it/s]
Epoch 0:  94%|█████████▍| 3106/3295 [32:21<01:58,  1.60it/s]
Epoch 0:  94%|█████████▍| 3107/3295 [32:23<01:57,  1.60it/s]
Epoch 0:  94%|█████████▍| 3107/3295 [32:23<01:57,  1.60it/s]
Epoch 0:  94%|█████████▍| 3108/3295 [32:23<01:56,  1.60it/s]
Epoch 0:  94%|█████████▍| 3108/3295 [32:23<01:56,  1.60it/s]
Epoch 0:  94%|█████████▍| 3109/3295 [32:23<01:56,  1.60it/s]
Epoch 0:  94%|█████████▍| 3109/3295 [32:23<01:56,  1.60it/s]
Epoch 0:  94%|█████████▍| 3110/3295 [32:25<01:55,  1.60it/s]
Epoch 0:  94%|█████████▍| 3110/3295 [32:25<01:55,  1.60it/s]
Epoch 0:  94%|█████████▍| 3111/3295 [32:25<01:55,  1.60it/s]
Epoch 0:  94%|█████████▍| 3111/3295 [32:25<01:55,  1.60it/s]
Epoch 0:  94%|█████████▍| 3112/3295 [32:25<01:54,  1.60it/s]
Epoch 0:  94%|█████████▍| 3112/3295 [32:25<01:54,  1.60it/s]
Epoch 0:  94%|█████████▍| 3113/3295 [32:27<01:53,  1.60it/s]
Epoch 0:  94%|█████████▍| 3113/3295 [32:27<01:53,  1.60it/s]
Epoch 0:  95%|█████████▍| 3114/3295 [32:27<01:53,  1.60it/s]
Epoch 0:  95%|█████████▍| 3114/3295 [32:27<01:53,  1.60it/s]
Epoch 0:  95%|█████████▍| 3115/3295 [32:27<01:52,  1.60it/s]
Epoch 0:  95%|█████████▍| 3115/3295 [32:27<01:52,  1.60it/s]
Epoch 0:  95%|█████████▍| 3116/3295 [32:28<01:51,  1.60it/s]
Epoch 0:  95%|█████████▍| 3116/3295 [32:28<01:51,  1.60it/s]
Epoch 0:  95%|█████████▍| 3117/3295 [32:28<01:51,  1.60it/s]
Epoch 0:  95%|█████████▍| 3117/3295 [32:28<01:51,  1.60it/s]
Epoch 0:  95%|█████████▍| 3118/3295 [32:28<01:50,  1.60it/s]
Epoch 0:  95%|█████████▍| 3118/3295 [32:28<01:50,  1.60it/s]
Epoch 0:  95%|█████████▍| 3119/3295 [32:30<01:50,  1.60it/s]
Epoch 0:  95%|█████████▍| 3119/3295 [32:30<01:50,  1.60it/s]
Epoch 0:  95%|█████████▍| 3120/3295 [32:30<01:49,  1.60it/s]
Epoch 0:  95%|█████████▍| 3120/3295 [32:30<01:49,  1.60it/s]
Epoch 0:  95%|█████████▍| 3121/3295 [32:30<01:48,  1.60it/s]
Epoch 0:  95%|█████████▍| 3121/3295 [32:30<01:48,  1.60it/s]
Epoch 0:  95%|█████████▍| 3122/3295 [32:32<01:48,  1.60it/s]
Epoch 0:  95%|█████████▍| 3122/3295 [32:32<01:48,  1.60it/s]
Epoch 0:  95%|█████████▍| 3123/3295 [32:32<01:47,  1.60it/s]
Epoch 0:  95%|█████████▍| 3123/3295 [32:32<01:47,  1.60it/s]
Epoch 0:  95%|█████████▍| 3124/3295 [32:32<01:46,  1.60it/s]
Epoch 0:  95%|█████████▍| 3124/3295 [32:32<01:46,  1.60it/s]
Epoch 0:  95%|█████████▍| 3125/3295 [32:34<01:46,  1.60it/s]
Epoch 0:  95%|█████████▍| 3125/3295 [32:34<01:46,  1.60it/s]
Epoch 0:  95%|█████████▍| 3126/3295 [32:34<01:45,  1.60it/s]
Epoch 0:  95%|█████████▍| 3126/3295 [32:34<01:45,  1.60it/s]
Epoch 0:  95%|█████████▍| 3127/3295 [32:34<01:45,  1.60it/s]
Epoch 0:  95%|█████████▍| 3127/3295 [32:34<01:45,  1.60it/s]
Epoch 0:  95%|█████████▍| 3128/3295 [32:36<01:44,  1.60it/s]
Epoch 0:  95%|█████████▍| 3128/3295 [32:36<01:44,  1.60it/s]
Epoch 0:  95%|█████████▍| 3129/3295 [32:36<01:43,  1.60it/s]
Epoch 0:  95%|█████████▍| 3129/3295 [32:36<01:43,  1.60it/s]
Epoch 0:  95%|█████████▍| 3130/3295 [32:36<01:43,  1.60it/s]
Epoch 0:  95%|█████████▍| 3130/3295 [32:36<01:43,  1.60it/s]
Epoch 0:  95%|█████████▌| 3131/3295 [32:38<01:42,  1.60it/s]
Epoch 0:  95%|█████████▌| 3131/3295 [32:38<01:42,  1.60it/s]
Epoch 0:  95%|█████████▌| 3132/3295 [32:38<01:41,  1.60it/s]
Epoch 0:  95%|█████████▌| 3132/3295 [32:38<01:41,  1.60it/s]
Epoch 0:  95%|█████████▌| 3133/3295 [32:38<01:41,  1.60it/s]
Epoch 0:  95%|█████████▌| 3133/3295 [32:38<01:41,  1.60it/s]
Epoch 0:  95%|█████████▌| 3134/3295 [32:39<01:40,  1.60it/s]
Epoch 0:  95%|█████████▌| 3134/3295 [32:39<01:40,  1.60it/s]
Epoch 0:  95%|█████████▌| 3135/3295 [32:39<01:40,  1.60it/s]
Epoch 0:  95%|█████████▌| 3135/3295 [32:39<01:40,  1.60it/s]
Epoch 0:  95%|█████████▌| 3136/3295 [32:39<01:39,  1.60it/s]
Epoch 0:  95%|█████████▌| 3136/3295 [32:39<01:39,  1.60it/s]
Epoch 0:  95%|█████████▌| 3137/3295 [32:41<01:38,  1.60it/s]
Epoch 0:  95%|█████████▌| 3137/3295 [32:41<01:38,  1.60it/s]
Epoch 0:  95%|█████████▌| 3138/3295 [32:41<01:38,  1.60it/s]
Epoch 0:  95%|█████████▌| 3138/3295 [32:41<01:38,  1.60it/s]
Epoch 0:  95%|█████████▌| 3139/3295 [32:41<01:37,  1.60it/s]
Epoch 0:  95%|█████████▌| 3139/3295 [32:41<01:37,  1.60it/s]
Epoch 0:  95%|█████████▌| 3140/3295 [32:43<01:36,  1.60it/s]
Epoch 0:  95%|█████████▌| 3140/3295 [32:43<01:36,  1.60it/s]
Epoch 0:  95%|█████████▌| 3141/3295 [32:43<01:36,  1.60it/s]
Epoch 0:  95%|█████████▌| 3141/3295 [32:43<01:36,  1.60it/s]
Epoch 0:  95%|█████████▌| 3142/3295 [32:43<01:35,  1.60it/s]
Epoch 0:  95%|█████████▌| 3142/3295 [32:43<01:35,  1.60it/s]
Epoch 0:  95%|█████████▌| 3143/3295 [32:45<01:35,  1.60it/s]
Epoch 0:  95%|█████████▌| 3143/3295 [32:45<01:35,  1.60it/s]
Epoch 0:  95%|█████████▌| 3144/3295 [32:45<01:34,  1.60it/s]
Epoch 0:  95%|█████████▌| 3144/3295 [32:45<01:34,  1.60it/s]
Epoch 0:  95%|█████████▌| 3145/3295 [32:45<01:33,  1.60it/s]
Epoch 0:  95%|█████████▌| 3145/3295 [32:45<01:33,  1.60it/s]
Epoch 0:  95%|█████████▌| 3146/3295 [32:47<01:33,  1.60it/s]
Epoch 0:  95%|█████████▌| 3146/3295 [32:47<01:33,  1.60it/s]
Epoch 0:  96%|█████████▌| 3147/3295 [32:47<01:32,  1.60it/s]
Epoch 0:  96%|█████████▌| 3147/3295 [32:47<01:32,  1.60it/s]
Epoch 0:  96%|█████████▌| 3148/3295 [32:47<01:31,  1.60it/s]
Epoch 0:  96%|█████████▌| 3148/3295 [32:47<01:31,  1.60it/s]
Epoch 0:  96%|█████████▌| 3149/3295 [32:49<01:31,  1.60it/s]
Epoch 0:  96%|█████████▌| 3149/3295 [32:49<01:31,  1.60it/s]
Epoch 0:  96%|█████████▌| 3150/3295 [32:49<01:30,  1.60it/s]
Epoch 0:  96%|█████████▌| 3150/3295 [32:49<01:30,  1.60it/s]
Epoch 0:  96%|█████████▌| 3151/3295 [32:49<01:29,  1.60it/s]
Epoch 0:  96%|█████████▌| 3151/3295 [32:49<01:29,  1.60it/s]
Epoch 0:  96%|█████████▌| 3152/3295 [32:50<01:29,  1.60it/s]
Epoch 0:  96%|█████████▌| 3152/3295 [32:50<01:29,  1.60it/s]
Epoch 0:  96%|█████████▌| 3153/3295 [32:50<01:28,  1.60it/s]
Epoch 0:  96%|█████████▌| 3153/3295 [32:50<01:28,  1.60it/s]
Epoch 0:  96%|█████████▌| 3154/3295 [32:50<01:28,  1.60it/s]
Epoch 0:  96%|█████████▌| 3154/3295 [32:50<01:28,  1.60it/s]
Epoch 0:  96%|█████████▌| 3155/3295 [32:52<01:27,  1.60it/s]
Epoch 0:  96%|█████████▌| 3155/3295 [32:52<01:27,  1.60it/s]
Epoch 0:  96%|█████████▌| 3156/3295 [32:52<01:26,  1.60it/s]
Epoch 0:  96%|█████████▌| 3156/3295 [32:52<01:26,  1.60it/s]
Epoch 0:  96%|█████████▌| 3157/3295 [32:52<01:26,  1.60it/s]
Epoch 0:  96%|█████████▌| 3157/3295 [32:52<01:26,  1.60it/s]
Epoch 0:  96%|█████████▌| 3158/3295 [32:54<01:25,  1.60it/s]
Epoch 0:  96%|█████████▌| 3158/3295 [32:54<01:25,  1.60it/s]
Epoch 0:  96%|█████████▌| 3159/3295 [32:54<01:25,  1.60it/s]
Epoch 0:  96%|█████████▌| 3159/3295 [32:54<01:25,  1.60it/s]
Epoch 0:  96%|█████████▌| 3160/3295 [32:54<01:24,  1.60it/s]
Epoch 0:  96%|█████████▌| 3160/3295 [32:54<01:24,  1.60it/s]
Epoch 0:  96%|█████████▌| 3161/3295 [32:56<01:23,  1.60it/s]
Epoch 0:  96%|█████████▌| 3161/3295 [32:56<01:23,  1.60it/s]
Epoch 0:  96%|█████████▌| 3162/3295 [32:56<01:23,  1.60it/s]
Epoch 0:  96%|█████████▌| 3162/3295 [32:56<01:23,  1.60it/s]
Epoch 0:  96%|█████████▌| 3163/3295 [32:56<01:22,  1.60it/s]
Epoch 0:  96%|█████████▌| 3163/3295 [32:56<01:22,  1.60it/s]
Epoch 0:  96%|█████████▌| 3164/3295 [32:58<01:21,  1.60it/s]
Epoch 0:  96%|█████████▌| 3164/3295 [32:58<01:21,  1.60it/s]
Epoch 0:  96%|█████████▌| 3165/3295 [32:58<01:21,  1.60it/s]
Epoch 0:  96%|█████████▌| 3165/3295 [32:58<01:21,  1.60it/s]
Epoch 0:  96%|█████████▌| 3166/3295 [32:58<01:20,  1.60it/s]
Epoch 0:  96%|█████████▌| 3166/3295 [32:58<01:20,  1.60it/s]
Epoch 0:  96%|█████████▌| 3167/3295 [33:00<01:20,  1.60it/s]
Epoch 0:  96%|█████████▌| 3167/3295 [33:00<01:20,  1.60it/s]
Epoch 0:  96%|█████████▌| 3168/3295 [33:00<01:19,  1.60it/s]
Epoch 0:  96%|█████████▌| 3168/3295 [33:00<01:19,  1.60it/s]
Epoch 0:  96%|█████████▌| 3169/3295 [33:00<01:18,  1.60it/s]
Epoch 0:  96%|█████████▌| 3169/3295 [33:00<01:18,  1.60it/s]
Epoch 0:  96%|█████████▌| 3170/3295 [33:02<01:18,  1.60it/s]
Epoch 0:  96%|█████████▌| 3170/3295 [33:02<01:18,  1.60it/s]
Epoch 0:  96%|█████████▌| 3171/3295 [33:02<01:17,  1.60it/s]
Epoch 0:  96%|█████████▌| 3171/3295 [33:02<01:17,  1.60it/s]
Epoch 0:  96%|█████████▋| 3172/3295 [33:02<01:16,  1.60it/s]
Epoch 0:  96%|█████████▋| 3172/3295 [33:02<01:16,  1.60it/s]
Epoch 0:  96%|█████████▋| 3173/3295 [33:03<01:16,  1.60it/s]
Epoch 0:  96%|█████████▋| 3173/3295 [33:03<01:16,  1.60it/s]
Epoch 0:  96%|█████████▋| 3174/3295 [33:03<01:15,  1.60it/s]
Epoch 0:  96%|█████████▋| 3174/3295 [33:03<01:15,  1.60it/s]
Epoch 0:  96%|█████████▋| 3175/3295 [33:04<01:14,  1.60it/s]
Epoch 0:  96%|█████████▋| 3175/3295 [33:04<01:14,  1.60it/s]
Epoch 0:  96%|█████████▋| 3176/3295 [33:05<01:14,  1.60it/s]
Epoch 0:  96%|█████████▋| 3176/3295 [33:05<01:14,  1.60it/s]
Epoch 0:  96%|█████████▋| 3177/3295 [33:05<01:13,  1.60it/s]
Epoch 0:  96%|█████████▋| 3177/3295 [33:05<01:13,  1.60it/s]
Epoch 0:  96%|█████████▋| 3178/3295 [33:05<01:13,  1.60it/s]
Epoch 0:  96%|█████████▋| 3178/3295 [33:05<01:13,  1.60it/s]
Epoch 0:  96%|█████████▋| 3179/3295 [33:07<01:12,  1.60it/s]
Epoch 0:  96%|█████████▋| 3179/3295 [33:07<01:12,  1.60it/s]
Epoch 0:  97%|█████████▋| 3180/3295 [33:07<01:11,  1.60it/s]
Epoch 0:  97%|█████████▋| 3180/3295 [33:07<01:11,  1.60it/s]
Epoch 0:  97%|█████████▋| 3181/3295 [33:07<01:11,  1.60it/s]
Epoch 0:  97%|█████████▋| 3181/3295 [33:07<01:11,  1.60it/s]
Epoch 0:  97%|█████████▋| 3182/3295 [33:09<01:10,  1.60it/s]
Epoch 0:  97%|█████████▋| 3182/3295 [33:09<01:10,  1.60it/s]
Epoch 0:  97%|█████████▋| 3183/3295 [33:09<01:10,  1.60it/s]
Epoch 0:  97%|█████████▋| 3183/3295 [33:09<01:10,  1.60it/s]
Epoch 0:  97%|█████████▋| 3184/3295 [33:09<01:09,  1.60it/s]
Epoch 0:  97%|█████████▋| 3184/3295 [33:09<01:09,  1.60it/s]
Epoch 0:  97%|█████████▋| 3185/3295 [33:11<01:08,  1.60it/s]
Epoch 0:  97%|█████████▋| 3185/3295 [33:11<01:08,  1.60it/s]
Epoch 0:  97%|█████████▋| 3186/3295 [33:11<01:08,  1.60it/s]
Epoch 0:  97%|█████████▋| 3186/3295 [33:11<01:08,  1.60it/s]
Epoch 0:  97%|█████████▋| 3187/3295 [33:11<01:07,  1.60it/s]
Epoch 0:  97%|█████████▋| 3187/3295 [33:11<01:07,  1.60it/s]
Epoch 0:  97%|█████████▋| 3188/3295 [33:13<01:06,  1.60it/s]
Epoch 0:  97%|█████████▋| 3188/3295 [33:13<01:06,  1.60it/s]
Epoch 0:  97%|█████████▋| 3189/3295 [33:13<01:06,  1.60it/s]
Epoch 0:  97%|█████████▋| 3189/3295 [33:13<01:06,  1.60it/s]
Epoch 0:  97%|█████████▋| 3190/3295 [33:13<01:05,  1.60it/s]
Epoch 0:  97%|█████████▋| 3190/3295 [33:13<01:05,  1.60it/s]
Epoch 0:  97%|█████████▋| 3191/3295 [33:15<01:05,  1.60it/s]
Epoch 0:  97%|█████████▋| 3191/3295 [33:15<01:05,  1.60it/s]
Epoch 0:  97%|█████████▋| 3192/3295 [33:15<01:04,  1.60it/s]
Epoch 0:  97%|█████████▋| 3192/3295 [33:15<01:04,  1.60it/s]
Epoch 0:  97%|█████████▋| 3193/3295 [33:15<01:03,  1.60it/s]
Epoch 0:  97%|█████████▋| 3193/3295 [33:15<01:03,  1.60it/s]
Epoch 0:  97%|█████████▋| 3194/3295 [33:16<01:03,  1.60it/s]
Epoch 0:  97%|█████████▋| 3194/3295 [33:16<01:03,  1.60it/s]
Epoch 0:  97%|█████████▋| 3195/3295 [33:16<01:02,  1.60it/s]
Epoch 0:  97%|█████████▋| 3195/3295 [33:16<01:02,  1.60it/s]
Epoch 0:  97%|█████████▋| 3196/3295 [33:16<01:01,  1.60it/s]
Epoch 0:  97%|█████████▋| 3196/3295 [33:16<01:01,  1.60it/s]
Epoch 0:  97%|█████████▋| 3197/3295 [33:18<01:01,  1.60it/s]
Epoch 0:  97%|█████████▋| 3197/3295 [33:18<01:01,  1.60it/s]
Epoch 0:  97%|█████████▋| 3198/3295 [33:18<01:00,  1.60it/s]
Epoch 0:  97%|█████████▋| 3198/3295 [33:18<01:00,  1.60it/s]
Epoch 0:  97%|█████████▋| 3199/3295 [33:18<00:59,  1.60it/s]
Epoch 0:  97%|█████████▋| 3199/3295 [33:18<00:59,  1.60it/s]
Epoch 0:  97%|█████████▋| 3200/3295 [33:20<00:59,  1.60it/s]
Epoch 0:  97%|█████████▋| 3200/3295 [33:20<00:59,  1.60it/s]
Epoch 0:  97%|█████████▋| 3201/3295 [33:20<00:58,  1.60it/s]
Epoch 0:  97%|█████████▋| 3201/3295 [33:20<00:58,  1.60it/s]
Epoch 0:  97%|█████████▋| 3202/3295 [33:20<00:58,  1.60it/s]
Epoch 0:  97%|█████████▋| 3202/3295 [33:20<00:58,  1.60it/s]
Epoch 0:  97%|█████████▋| 3203/3295 [33:22<00:57,  1.60it/s]
Epoch 0:  97%|█████████▋| 3203/3295 [33:22<00:57,  1.60it/s]
Epoch 0:  97%|█████████▋| 3204/3295 [33:22<00:56,  1.60it/s]
Epoch 0:  97%|█████████▋| 3204/3295 [33:22<00:56,  1.60it/s]
Epoch 0:  97%|█████████▋| 3205/3295 [33:22<00:56,  1.60it/s]
Epoch 0:  97%|█████████▋| 3205/3295 [33:22<00:56,  1.60it/s]
Epoch 0:  97%|█████████▋| 3206/3295 [33:24<00:55,  1.60it/s]
Epoch 0:  97%|█████████▋| 3206/3295 [33:24<00:55,  1.60it/s]
Epoch 0:  97%|█████████▋| 3207/3295 [33:24<00:54,  1.60it/s]
Epoch 0:  97%|█████████▋| 3207/3295 [33:24<00:54,  1.60it/s]
Epoch 0:  97%|█████████▋| 3208/3295 [33:24<00:54,  1.60it/s]
Epoch 0:  97%|█████████▋| 3208/3295 [33:24<00:54,  1.60it/s]
Epoch 0:  97%|█████████▋| 3209/3295 [33:25<00:53,  1.60it/s]
Epoch 0:  97%|█████████▋| 3209/3295 [33:25<00:53,  1.60it/s]
Epoch 0:  97%|█████████▋| 3210/3295 [33:26<00:53,  1.60it/s]
Epoch 0:  97%|█████████▋| 3210/3295 [33:26<00:53,  1.60it/s]
Epoch 0:  97%|█████████▋| 3211/3295 [33:26<00:52,  1.60it/s]
Epoch 0:  97%|█████████▋| 3211/3295 [33:26<00:52,  1.60it/s]
Epoch 0:  97%|█████████▋| 3212/3295 [33:27<00:51,  1.60it/s]
Epoch 0:  97%|█████████▋| 3212/3295 [33:27<00:51,  1.60it/s]
Epoch 0:  98%|█████████▊| 3213/3295 [33:27<00:51,  1.60it/s]
Epoch 0:  98%|█████████▊| 3213/3295 [33:27<00:51,  1.60it/s]
Epoch 0:  98%|█████████▊| 3214/3295 [33:27<00:50,  1.60it/s]
Epoch 0:  98%|█████████▊| 3214/3295 [33:27<00:50,  1.60it/s]
Epoch 0:  98%|█████████▊| 3215/3295 [33:29<00:50,  1.60it/s]
Epoch 0:  98%|█████████▊| 3215/3295 [33:29<00:50,  1.60it/s]
Epoch 0:  98%|█████████▊| 3216/3295 [33:29<00:49,  1.60it/s]
Epoch 0:  98%|█████████▊| 3216/3295 [33:29<00:49,  1.60it/s]
Epoch 0:  98%|█████████▊| 3217/3295 [33:29<00:48,  1.60it/s]
Epoch 0:  98%|█████████▊| 3217/3295 [33:29<00:48,  1.60it/s]
Epoch 0:  98%|█████████▊| 3218/3295 [33:31<00:48,  1.60it/s]
Epoch 0:  98%|█████████▊| 3218/3295 [33:31<00:48,  1.60it/s]
Epoch 0:  98%|█████████▊| 3219/3295 [33:31<00:47,  1.60it/s]
Epoch 0:  98%|█████████▊| 3219/3295 [33:31<00:47,  1.60it/s]
Epoch 0:  98%|█████████▊| 3220/3295 [33:31<00:46,  1.60it/s]
Epoch 0:  98%|█████████▊| 3220/3295 [33:31<00:46,  1.60it/s]
Epoch 0:  98%|█████████▊| 3221/3295 [33:33<00:46,  1.60it/s]
Epoch 0:  98%|█████████▊| 3221/3295 [33:33<00:46,  1.60it/s]
Epoch 0:  98%|█████████▊| 3222/3295 [33:33<00:45,  1.60it/s]
Epoch 0:  98%|█████████▊| 3222/3295 [33:33<00:45,  1.60it/s]
Epoch 0:  98%|█████████▊| 3223/3295 [33:33<00:44,  1.60it/s]
Epoch 0:  98%|█████████▊| 3223/3295 [33:33<00:44,  1.60it/s]
Epoch 0:  98%|█████████▊| 3224/3295 [33:35<00:44,  1.60it/s]
Epoch 0:  98%|█████████▊| 3224/3295 [33:35<00:44,  1.60it/s]
Epoch 0:  98%|█████████▊| 3225/3295 [33:35<00:43,  1.60it/s]
Epoch 0:  98%|█████████▊| 3225/3295 [33:35<00:43,  1.60it/s]
Epoch 0:  98%|█████████▊| 3226/3295 [33:35<00:43,  1.60it/s]
Epoch 0:  98%|█████████▊| 3226/3295 [33:35<00:43,  1.60it/s]
Epoch 0:  98%|█████████▊| 3227/3295 [33:36<00:42,  1.60it/s]
Epoch 0:  98%|█████████▊| 3227/3295 [33:36<00:42,  1.60it/s]
Epoch 0:  98%|█████████▊| 3228/3295 [33:37<00:41,  1.60it/s]
Epoch 0:  98%|█████████▊| 3228/3295 [33:37<00:41,  1.60it/s]
Epoch 0:  98%|█████████▊| 3229/3295 [33:37<00:41,  1.60it/s]
Epoch 0:  98%|█████████▊| 3229/3295 [33:37<00:41,  1.60it/s]
Epoch 0:  98%|█████████▊| 3230/3295 [33:38<00:40,  1.60it/s]
Epoch 0:  98%|█████████▊| 3230/3295 [33:38<00:40,  1.60it/s]
Epoch 0:  98%|█████████▊| 3231/3295 [33:38<00:39,  1.60it/s]
Epoch 0:  98%|█████████▊| 3231/3295 [33:38<00:39,  1.60it/s]
Epoch 0:  98%|█████████▊| 3232/3295 [33:38<00:39,  1.60it/s]
Epoch 0:  98%|█████████▊| 3232/3295 [33:38<00:39,  1.60it/s]
Epoch 0:  98%|█████████▊| 3233/3295 [33:40<00:38,  1.60it/s]
Epoch 0:  98%|█████████▊| 3233/3295 [33:40<00:38,  1.60it/s]
Epoch 0:  98%|█████████▊| 3234/3295 [33:40<00:38,  1.60it/s]
Epoch 0:  98%|█████████▊| 3234/3295 [33:40<00:38,  1.60it/s]
Epoch 0:  98%|█████████▊| 3235/3295 [33:40<00:37,  1.60it/s]
Epoch 0:  98%|█████████▊| 3235/3295 [33:40<00:37,  1.60it/s]
Epoch 0:  98%|█████████▊| 3236/3295 [33:42<00:36,  1.60it/s]
Epoch 0:  98%|█████████▊| 3236/3295 [33:42<00:36,  1.60it/s]
Epoch 0:  98%|█████████▊| 3237/3295 [33:42<00:36,  1.60it/s]
Epoch 0:  98%|█████████▊| 3237/3295 [33:42<00:36,  1.60it/s]
Epoch 0:  98%|█████████▊| 3238/3295 [33:42<00:35,  1.60it/s]
Epoch 0:  98%|█████████▊| 3238/3295 [33:42<00:35,  1.60it/s]
Epoch 0:  98%|█████████▊| 3239/3295 [33:44<00:34,  1.60it/s]
Epoch 0:  98%|█████████▊| 3239/3295 [33:44<00:34,  1.60it/s]
Epoch 0:  98%|█████████▊| 3240/3295 [33:44<00:34,  1.60it/s]
Epoch 0:  98%|█████████▊| 3240/3295 [33:44<00:34,  1.60it/s]
Epoch 0:  98%|█████████▊| 3241/3295 [33:44<00:33,  1.60it/s]
Epoch 0:  98%|█████████▊| 3241/3295 [33:44<00:33,  1.60it/s]
Epoch 0:  98%|█████████▊| 3242/3295 [33:46<00:33,  1.60it/s]
Epoch 0:  98%|█████████▊| 3242/3295 [33:46<00:33,  1.60it/s]
Epoch 0:  98%|█████████▊| 3243/3295 [33:46<00:32,  1.60it/s]
Epoch 0:  98%|█████████▊| 3243/3295 [33:46<00:32,  1.60it/s]
Epoch 0:  98%|█████████▊| 3244/3295 [33:46<00:31,  1.60it/s]
Epoch 0:  98%|█████████▊| 3244/3295 [33:46<00:31,  1.60it/s]
Epoch 0:  98%|█████████▊| 3245/3295 [33:48<00:31,  1.60it/s]
Epoch 0:  98%|█████████▊| 3245/3295 [33:48<00:31,  1.60it/s]
Epoch 0:  99%|█████████▊| 3246/3295 [33:48<00:30,  1.60it/s]
Epoch 0:  99%|█████████▊| 3246/3295 [33:48<00:30,  1.60it/s]
Epoch 0:  99%|█████████▊| 3247/3295 [33:48<00:29,  1.60it/s]
Epoch 0:  99%|█████████▊| 3247/3295 [33:48<00:29,  1.60it/s]
Epoch 0:  99%|█████████▊| 3248/3295 [33:49<00:29,  1.60it/s]
Epoch 0:  99%|█████████▊| 3248/3295 [33:49<00:29,  1.60it/s]
Epoch 0:  99%|█████████▊| 3249/3295 [33:49<00:28,  1.60it/s]
Epoch 0:  99%|█████████▊| 3249/3295 [33:49<00:28,  1.60it/s]
Epoch 0:  99%|█████████▊| 3250/3295 [33:49<00:28,  1.60it/s]
Epoch 0:  99%|█████████▊| 3250/3295 [33:49<00:28,  1.60it/s]
Epoch 0:  99%|█████████▊| 3251/3295 [33:51<00:27,  1.60it/s]
Epoch 0:  99%|█████████▊| 3251/3295 [33:51<00:27,  1.60it/s]
Epoch 0:  99%|█████████▊| 3252/3295 [33:51<00:26,  1.60it/s]
Epoch 0:  99%|█████████▊| 3252/3295 [33:51<00:26,  1.60it/s]
Epoch 0:  99%|█████████▊| 3253/3295 [33:51<00:26,  1.60it/s]
Epoch 0:  99%|█████████▊| 3253/3295 [33:51<00:26,  1.60it/s]
Epoch 0:  99%|█████████▉| 3254/3295 [33:53<00:25,  1.60it/s]
Epoch 0:  99%|█████████▉| 3254/3295 [33:53<00:25,  1.60it/s]
Epoch 0:  99%|█████████▉| 3255/3295 [33:53<00:24,  1.60it/s]
Epoch 0:  99%|█████████▉| 3255/3295 [33:53<00:24,  1.60it/s]
Epoch 0:  99%|█████████▉| 3256/3295 [33:53<00:24,  1.60it/s]
Epoch 0:  99%|█████████▉| 3256/3295 [33:53<00:24,  1.60it/s]
Epoch 0:  99%|█████████▉| 3257/3295 [33:55<00:23,  1.60it/s]
Epoch 0:  99%|█████████▉| 3257/3295 [33:55<00:23,  1.60it/s]
Epoch 0:  99%|█████████▉| 3258/3295 [33:55<00:23,  1.60it/s]
Epoch 0:  99%|█████████▉| 3258/3295 [33:55<00:23,  1.60it/s]
Epoch 0:  99%|█████████▉| 3259/3295 [33:55<00:22,  1.60it/s]
Epoch 0:  99%|█████████▉| 3259/3295 [33:55<00:22,  1.60it/s]
Epoch 0:  99%|█████████▉| 3260/3295 [33:57<00:21,  1.60it/s]
Epoch 0:  99%|█████████▉| 3260/3295 [33:57<00:21,  1.60it/s]
Epoch 0:  99%|█████████▉| 3261/3295 [33:57<00:21,  1.60it/s]
Epoch 0:  99%|█████████▉| 3261/3295 [33:57<00:21,  1.60it/s]
Epoch 0:  99%|█████████▉| 3262/3295 [33:57<00:20,  1.60it/s]
Epoch 0:  99%|█████████▉| 3262/3295 [33:57<00:20,  1.60it/s]
Epoch 0:  99%|█████████▉| 3263/3295 [33:59<00:19,  1.60it/s]
Epoch 0:  99%|█████████▉| 3263/3295 [33:59<00:19,  1.60it/s]
Epoch 0:  99%|█████████▉| 3264/3295 [33:59<00:19,  1.60it/s]
Epoch 0:  99%|█████████▉| 3264/3295 [33:59<00:19,  1.60it/s]
Epoch 0:  99%|█████████▉| 3265/3295 [33:59<00:18,  1.60it/s]
Epoch 0:  99%|█████████▉| 3265/3295 [33:59<00:18,  1.60it/s]
Epoch 0:  99%|█████████▉| 3266/3295 [34:01<00:18,  1.60it/s]
Epoch 0:  99%|█████████▉| 3266/3295 [34:01<00:18,  1.60it/s]
Epoch 0:  99%|█████████▉| 3267/3295 [34:01<00:17,  1.60it/s]
Epoch 0:  99%|█████████▉| 3267/3295 [34:01<00:17,  1.60it/s]
Epoch 0:  99%|█████████▉| 3268/3295 [34:01<00:16,  1.60it/s]
Epoch 0:  99%|█████████▉| 3268/3295 [34:01<00:16,  1.60it/s]
Epoch 0:  99%|█████████▉| 3269/3295 [34:03<00:16,  1.60it/s]
Epoch 0:  99%|█████████▉| 3269/3295 [34:03<00:16,  1.60it/s]
Epoch 0:  99%|█████████▉| 3270/3295 [34:03<00:15,  1.60it/s]
Epoch 0:  99%|█████████▉| 3270/3295 [34:03<00:15,  1.60it/s]
Epoch 0:  99%|█████████▉| 3271/3295 [34:03<00:14,  1.60it/s]
Epoch 0:  99%|█████████▉| 3271/3295 [34:03<00:14,  1.60it/s]
Epoch 0:  99%|█████████▉| 3272/3295 [34:05<00:14,  1.60it/s]
Epoch 0:  99%|█████████▉| 3272/3295 [34:05<00:14,  1.60it/s]
Epoch 0:  99%|█████████▉| 3273/3295 [34:05<00:13,  1.60it/s]
Epoch 0:  99%|█████████▉| 3273/3295 [34:05<00:13,  1.60it/s]
Epoch 0:  99%|█████████▉| 3274/3295 [34:05<00:13,  1.60it/s]
Epoch 0:  99%|█████████▉| 3274/3295 [34:05<00:13,  1.60it/s]
Epoch 0:  99%|█████████▉| 3275/3295 [34:06<00:12,  1.60it/s]
Epoch 0:  99%|█████████▉| 3275/3295 [34:06<00:12,  1.60it/s]
Epoch 0:  99%|█████████▉| 3276/3295 [34:06<00:11,  1.60it/s]
Epoch 0:  99%|█████████▉| 3276/3295 [34:06<00:11,  1.60it/s]
Epoch 0:  99%|█████████▉| 3277/3295 [34:06<00:11,  1.60it/s]
Epoch 0:  99%|█████████▉| 3277/3295 [34:06<00:11,  1.60it/s]
Epoch 0:  99%|█████████▉| 3278/3295 [34:08<00:10,  1.60it/s]
Epoch 0:  99%|█████████▉| 3278/3295 [34:08<00:10,  1.60it/s]
Epoch 0: 100%|█████████▉| 3279/3295 [34:08<00:09,  1.60it/s]
Epoch 0: 100%|█████████▉| 3279/3295 [34:08<00:09,  1.60it/s]
Epoch 0: 100%|█████████▉| 3280/3295 [34:08<00:09,  1.60it/s]
Epoch 0: 100%|█████████▉| 3280/3295 [34:08<00:09,  1.60it/s]
Epoch 0: 100%|█████████▉| 3281/3295 [34:10<00:08,  1.60it/s]
Epoch 0: 100%|█████████▉| 3281/3295 [34:10<00:08,  1.60it/s]
Epoch 0: 100%|█████████▉| 3282/3295 [34:10<00:08,  1.60it/s]
Epoch 0: 100%|█████████▉| 3282/3295 [34:10<00:08,  1.60it/s]
Epoch 0: 100%|█████████▉| 3283/3295 [34:10<00:07,  1.60it/s]
Epoch 0: 100%|█████████▉| 3283/3295 [34:10<00:07,  1.60it/s]
Epoch 0: 100%|█████████▉| 3284/3295 [34:12<00:06,  1.60it/s]
Epoch 0: 100%|█████████▉| 3284/3295 [34:12<00:06,  1.60it/s]
Epoch 0: 100%|█████████▉| 3285/3295 [34:12<00:06,  1.60it/s]
Epoch 0: 100%|█████████▉| 3285/3295 [34:12<00:06,  1.60it/s]
Epoch 0: 100%|█████████▉| 3286/3295 [34:12<00:05,  1.60it/s]
Epoch 0: 100%|█████████▉| 3286/3295 [34:12<00:05,  1.60it/s]
Epoch 0: 100%|█████████▉| 3287/3295 [34:14<00:05,  1.60it/s]
Epoch 0: 100%|█████████▉| 3287/3295 [34:14<00:05,  1.60it/s]
Epoch 0: 100%|█████████▉| 3288/3295 [34:14<00:04,  1.60it/s]
Epoch 0: 100%|█████████▉| 3288/3295 [34:14<00:04,  1.60it/s]
Epoch 0: 100%|█████████▉| 3289/3295 [34:14<00:03,  1.60it/s]
Epoch 0: 100%|█████████▉| 3289/3295 [34:14<00:03,  1.60it/s]
Epoch 0: 100%|█████████▉| 3290/3295 [34:16<00:03,  1.60it/s]
Epoch 0: 100%|█████████▉| 3290/3295 [34:16<00:03,  1.60it/s]
Epoch 0: 100%|█████████▉| 3291/3295 [34:16<00:02,  1.60it/s]
Epoch 0: 100%|█████████▉| 3291/3295 [34:16<00:02,  1.60it/s]
Epoch 0: 100%|█████████▉| 3292/3295 [34:16<00:01,  1.60it/s]
Epoch 0: 100%|█████████▉| 3292/3295 [34:16<00:01,  1.60it/s]
Epoch 0: 100%|█████████▉| 3293/3295 [34:18<00:01,  1.60it/s]
Epoch 0: 100%|█████████▉| 3293/3295 [34:18<00:01,  1.60it/s]
Epoch 0: 100%|█████████▉| 3294/3295 [34:18<00:00,  1.60it/s]
Epoch 0: 100%|█████████▉| 3294/3295 [34:18<00:00,  1.60it/s]
Epoch 0: 100%|██████████| 3295/3295 [34:19<00:00,  1.60it/s]
Epoch 0: 100%|██████████| 3295/3295 [34:19<00:00,  1.60it/s]

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 0/602 [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 1/602 [00:00<00:10, 60.07it/s][A

Validation DataLoader 0:   0%|          | 2/602 [00:00<00:10, 59.08it/s][A

Validation DataLoader 0:   0%|          | 3/602 [00:00<00:10, 58.97it/s][A

Validation DataLoader 0:   1%|          | 4/602 [00:00<00:09, 61.36it/s][A

Validation DataLoader 0:   1%|          | 5/602 [00:00<00:09, 63.75it/s][A

Validation DataLoader 0:   1%|          | 6/602 [00:00<00:09, 64.96it/s][A

Validation DataLoader 0:   1%|          | 7/602 [00:00<00:37, 15.72it/s][A

Validation DataLoader 0:   1%|▏         | 8/602 [00:00<00:39, 14.91it/s][A

Validation DataLoader 0:   1%|▏         | 9/602 [00:00<00:36, 16.32it/s][A

Validation DataLoader 0:   2%|▏         | 10/602 [00:00<00:49, 11.95it/s][A

Validation DataLoader 0:   2%|▏         | 11/602 [00:00<00:52, 11.16it/s][A

Validation DataLoader 0:   2%|▏         | 12/602 [00:00<00:49, 12.00it/s][A

Validation DataLoader 0:   2%|▏         | 13/602 [00:01<00:56, 10.48it/s][A

Validation DataLoader 0:   2%|▏         | 14/602 [00:01<00:59,  9.93it/s][A

Validation DataLoader 0:   2%|▏         | 15/602 [00:01<00:55, 10.53it/s][A

Validation DataLoader 0:   3%|▎         | 16/602 [00:01<01:00,  9.67it/s][A

Validation DataLoader 0:   3%|▎         | 17/602 [00:01<01:05,  8.96it/s][A

Validation DataLoader 0:   3%|▎         | 18/602 [00:01<01:01,  9.42it/s][A

Validation DataLoader 0:   3%|▎         | 19/602 [00:02<01:02,  9.28it/s][A

Validation DataLoader 0:   3%|▎         | 20/602 [00:02<01:09,  8.32it/s][A

Validation DataLoader 0:   3%|▎         | 21/602 [00:02<01:06,  8.68it/s][A

Validation DataLoader 0:   4%|▎         | 22/602 [00:02<01:06,  8.68it/s][A

Validation DataLoader 0:   4%|▍         | 23/602 [00:02<01:12,  8.04it/s][A

Validation DataLoader 0:   4%|▍         | 24/602 [00:02<01:09,  8.35it/s][A

Validation DataLoader 0:   4%|▍         | 25/602 [00:02<01:09,  8.34it/s][A

Validation DataLoader 0:   4%|▍         | 26/602 [00:03<01:13,  7.85it/s][A

Validation DataLoader 0:   4%|▍         | 27/602 [00:03<01:10,  8.12it/s][A

Validation DataLoader 0:   5%|▍         | 28/602 [00:03<01:11,  8.05it/s][A

Validation DataLoader 0:   5%|▍         | 29/602 [00:03<01:14,  7.74it/s][A

Validation DataLoader 0:   5%|▍         | 30/602 [00:03<01:11,  7.98it/s][A

Validation DataLoader 0:   5%|▌         | 31/602 [00:03<01:12,  7.88it/s][A

Validation DataLoader 0:   5%|▌         | 32/602 [00:04<01:13,  7.71it/s][A

Validation DataLoader 0:   5%|▌         | 33/602 [00:04<01:11,  7.93it/s][A

Validation DataLoader 0:   6%|▌         | 34/602 [00:04<01:13,  7.70it/s][A

Validation DataLoader 0:   6%|▌         | 35/602 [00:04<01:14,  7.66it/s][A

Validation DataLoader 0:   6%|▌         | 36/602 [00:04<01:12,  7.85it/s][A

Validation DataLoader 0:   6%|▌         | 37/602 [00:04<01:14,  7.59it/s][A

Validation DataLoader 0:   6%|▋         | 38/602 [00:05<01:14,  7.56it/s][A

Validation DataLoader 0:   6%|▋         | 39/602 [00:05<01:12,  7.74it/s][A

Validation DataLoader 0:   7%|▋         | 40/602 [00:05<01:14,  7.52it/s][A

Validation DataLoader 0:   7%|▋         | 41/602 [00:05<01:15,  7.43it/s][A

Validation DataLoader 0:   7%|▋         | 42/602 [00:05<01:13,  7.59it/s][A

Validation DataLoader 0:   7%|▋         | 43/602 [00:05<01:14,  7.47it/s][A

Validation DataLoader 0:   7%|▋         | 44/602 [00:05<01:15,  7.37it/s][A

Validation DataLoader 0:   7%|▋         | 45/602 [00:05<01:14,  7.52it/s][A

Validation DataLoader 0:   8%|▊         | 46/602 [00:06<01:14,  7.42it/s][A

Validation DataLoader 0:   8%|▊         | 47/602 [00:06<01:15,  7.39it/s][A

Validation DataLoader 0:   8%|▊         | 48/602 [00:06<01:13,  7.53it/s][A

Validation DataLoader 0:   8%|▊         | 49/602 [00:06<01:15,  7.29it/s][A

Validation DataLoader 0:   8%|▊         | 50/602 [00:06<01:15,  7.32it/s][A

Validation DataLoader 0:   8%|▊         | 51/602 [00:06<01:13,  7.45it/s][A

Validation DataLoader 0:   9%|▊         | 52/602 [00:07<01:15,  7.25it/s][A

Validation DataLoader 0:   9%|▉         | 53/602 [00:07<01:15,  7.27it/s][A

Validation DataLoader 0:   9%|▉         | 54/602 [00:07<01:14,  7.40it/s][A

Validation DataLoader 0:   9%|▉         | 55/602 [00:07<01:15,  7.21it/s][A

Validation DataLoader 0:   9%|▉         | 56/602 [00:07<01:15,  7.27it/s][A

Validation DataLoader 0:   9%|▉         | 57/602 [00:07<01:13,  7.38it/s][A

Validation DataLoader 0:  10%|▉         | 58/602 [00:08<01:15,  7.24it/s][A

Validation DataLoader 0:  10%|▉         | 59/602 [00:08<01:15,  7.23it/s][A

Validation DataLoader 0:  10%|▉         | 60/602 [00:08<01:13,  7.34it/s][A

Validation DataLoader 0:  10%|█         | 61/602 [00:08<01:15,  7.21it/s][A

Validation DataLoader 0:  10%|█         | 62/602 [00:08<01:15,  7.16it/s][A

Validation DataLoader 0:  10%|█         | 63/602 [00:08<01:14,  7.27it/s][A

Validation DataLoader 0:  11%|█         | 64/602 [00:08<01:15,  7.16it/s][A

Validation DataLoader 0:  11%|█         | 65/602 [00:09<01:15,  7.12it/s][A

Validation DataLoader 0:  11%|█         | 66/602 [00:09<01:14,  7.22it/s][A

Validation DataLoader 0:  11%|█         | 67/602 [00:09<01:14,  7.15it/s][A

Validation DataLoader 0:  11%|█▏        | 68/602 [00:09<01:16,  7.01it/s][A

Validation DataLoader 0:  11%|█▏        | 69/602 [00:09<01:15,  7.10it/s][A

Validation DataLoader 0:  12%|█▏        | 70/602 [00:09<01:14,  7.15it/s][A

Validation DataLoader 0:  12%|█▏        | 71/602 [00:10<01:15,  7.02it/s][A

Validation DataLoader 0:  12%|█▏        | 72/602 [00:10<01:14,  7.11it/s][A

Validation DataLoader 0:  12%|█▏        | 73/602 [00:10<01:14,  7.13it/s][A

Validation DataLoader 0:  12%|█▏        | 74/602 [00:10<01:14,  7.05it/s][A

Validation DataLoader 0:  12%|█▏        | 75/602 [00:10<01:13,  7.13it/s][A

Validation DataLoader 0:  13%|█▎        | 76/602 [00:10<01:14,  7.06it/s][A

Validation DataLoader 0:  13%|█▎        | 77/602 [00:10<01:14,  7.05it/s][A

Validation DataLoader 0:  13%|█▎        | 78/602 [00:10<01:13,  7.13it/s][A

Validation DataLoader 0:  13%|█▎        | 79/602 [00:11<01:14,  7.03it/s][A

Validation DataLoader 0:  13%|█▎        | 80/602 [00:11<01:13,  7.07it/s][A

Validation DataLoader 0:  13%|█▎        | 81/602 [00:11<01:12,  7.15it/s][A

Validation DataLoader 0:  14%|█▎        | 82/602 [00:11<01:14,  6.99it/s][A

Validation DataLoader 0:  14%|█▍        | 83/602 [00:11<01:13,  7.06it/s][A

Validation DataLoader 0:  14%|█▍        | 84/602 [00:11<01:12,  7.14it/s][A

Validation DataLoader 0:  14%|█▍        | 85/602 [00:12<01:13,  7.00it/s][A

Validation DataLoader 0:  14%|█▍        | 86/602 [00:12<01:13,  7.02it/s][A

Validation DataLoader 0:  14%|█▍        | 87/602 [00:12<01:12,  7.09it/s][A

Validation DataLoader 0:  15%|█▍        | 88/602 [00:12<01:13,  7.01it/s][A

Validation DataLoader 0:  15%|█▍        | 89/602 [00:12<01:12,  7.03it/s][A

Validation DataLoader 0:  15%|█▍        | 90/602 [00:12<01:12,  7.10it/s][A

Validation DataLoader 0:  15%|█▌        | 91/602 [00:12<01:12,  7.03it/s][A

Validation DataLoader 0:  15%|█▌        | 92/602 [00:13<01:13,  6.98it/s][A

Validation DataLoader 0:  15%|█▌        | 93/602 [00:13<01:12,  7.05it/s][A

Validation DataLoader 0:  16%|█▌        | 94/602 [00:13<01:12,  7.04it/s][A

Validation DataLoader 0:  16%|█▌        | 95/602 [00:13<01:12,  6.98it/s][A

Validation DataLoader 0:  16%|█▌        | 96/602 [00:13<01:11,  7.05it/s][A

Validation DataLoader 0:  16%|█▌        | 97/602 [00:13<01:11,  7.03it/s][A

Validation DataLoader 0:  16%|█▋        | 98/602 [00:14<01:12,  6.97it/s][A

Validation DataLoader 0:  16%|█▋        | 99/602 [00:14<01:11,  7.03it/s][A

Validation DataLoader 0:  17%|█▋        | 100/602 [00:14<01:11,  7.01it/s][A

Validation DataLoader 0:  17%|█▋        | 101/602 [00:14<01:11,  6.97it/s][A

Validation DataLoader 0:  17%|█▋        | 102/602 [00:14<01:11,  7.03it/s][A

Validation DataLoader 0:  17%|█▋        | 103/602 [00:14<01:11,  7.00it/s][A

Validation DataLoader 0:  17%|█▋        | 104/602 [00:14<01:11,  6.95it/s][A

Validation DataLoader 0:  17%|█▋        | 105/602 [00:14<01:10,  7.01it/s][A

Validation DataLoader 0:  18%|█▊        | 106/602 [00:15<01:11,  6.95it/s][A

Validation DataLoader 0:  18%|█▊        | 107/602 [00:15<01:11,  6.95it/s][A

Validation DataLoader 0:  18%|█▊        | 108/602 [00:15<01:10,  7.01it/s][A

Validation DataLoader 0:  18%|█▊        | 109/602 [00:15<01:11,  6.94it/s][A

Validation DataLoader 0:  18%|█▊        | 110/602 [00:15<01:10,  6.93it/s][A

Validation DataLoader 0:  18%|█▊        | 111/602 [00:15<01:10,  6.99it/s][A

Validation DataLoader 0:  19%|█▊        | 112/602 [00:16<01:10,  6.93it/s][A

Validation DataLoader 0:  19%|█▉        | 113/602 [00:16<01:10,  6.91it/s][A

Validation DataLoader 0:  19%|█▉        | 114/602 [00:16<01:10,  6.96it/s][A

Validation DataLoader 0:  19%|█▉        | 115/602 [00:16<01:10,  6.92it/s][A

Validation DataLoader 0:  19%|█▉        | 116/602 [00:16<01:10,  6.89it/s][A

Validation DataLoader 0:  19%|█▉        | 117/602 [00:16<01:09,  6.94it/s][A

Validation DataLoader 0:  20%|█▉        | 118/602 [00:17<01:09,  6.93it/s][A

Validation DataLoader 0:  20%|█▉        | 119/602 [00:17<01:10,  6.90it/s][A

Validation DataLoader 0:  20%|█▉        | 120/602 [00:17<01:09,  6.95it/s][A

Validation DataLoader 0:  20%|██        | 121/602 [00:17<01:09,  6.93it/s][A

Validation DataLoader 0:  20%|██        | 122/602 [00:17<01:09,  6.88it/s][A

Validation DataLoader 0:  20%|██        | 123/602 [00:17<01:09,  6.93it/s][A

Validation DataLoader 0:  21%|██        | 124/602 [00:17<01:09,  6.90it/s][A

Validation DataLoader 0:  21%|██        | 125/602 [00:18<01:09,  6.86it/s][A

Validation DataLoader 0:  21%|██        | 126/602 [00:18<01:08,  6.91it/s][A

Validation DataLoader 0:  21%|██        | 127/602 [00:18<01:08,  6.90it/s][A

Validation DataLoader 0:  21%|██▏       | 128/602 [00:18<01:08,  6.88it/s][A

Validation DataLoader 0:  21%|██▏       | 129/602 [00:18<01:08,  6.93it/s][A

Validation DataLoader 0:  22%|██▏       | 130/602 [00:18<01:08,  6.90it/s][A

Validation DataLoader 0:  22%|██▏       | 131/602 [00:19<01:08,  6.89it/s][A

Validation DataLoader 0:  22%|██▏       | 132/602 [00:19<01:07,  6.94it/s][A

Validation DataLoader 0:  22%|██▏       | 133/602 [00:19<01:07,  6.90it/s][A

Validation DataLoader 0:  22%|██▏       | 134/602 [00:19<01:07,  6.89it/s][A

Validation DataLoader 0:  22%|██▏       | 135/602 [00:19<01:07,  6.93it/s][A

Validation DataLoader 0:  23%|██▎       | 136/602 [00:19<01:07,  6.88it/s][A

Validation DataLoader 0:  23%|██▎       | 137/602 [00:19<01:07,  6.89it/s][A

Validation DataLoader 0:  23%|██▎       | 138/602 [00:19<01:06,  6.93it/s][A

Validation DataLoader 0:  23%|██▎       | 139/602 [00:20<01:07,  6.86it/s][A

Validation DataLoader 0:  23%|██▎       | 140/602 [00:20<01:07,  6.88it/s][A

Validation DataLoader 0:  23%|██▎       | 141/602 [00:20<01:06,  6.92it/s][A

Validation DataLoader 0:  24%|██▎       | 142/602 [00:20<01:06,  6.89it/s][A

Validation DataLoader 0:  24%|██▍       | 143/602 [00:20<01:06,  6.88it/s][A

Validation DataLoader 0:  24%|██▍       | 144/602 [00:20<01:06,  6.92it/s][A

Validation DataLoader 0:  24%|██▍       | 145/602 [00:21<01:06,  6.88it/s][A

Validation DataLoader 0:  24%|██▍       | 146/602 [00:21<01:06,  6.86it/s][A

Validation DataLoader 0:  24%|██▍       | 147/602 [00:21<01:05,  6.91it/s][A

Validation DataLoader 0:  25%|██▍       | 148/602 [00:21<01:06,  6.87it/s][A

Validation DataLoader 0:  25%|██▍       | 149/602 [00:21<01:06,  6.86it/s][A

Validation DataLoader 0:  25%|██▍       | 150/602 [00:21<01:05,  6.90it/s][A

Validation DataLoader 0:  25%|██▌       | 151/602 [00:21<01:05,  6.87it/s][A

Validation DataLoader 0:  25%|██▌       | 152/602 [00:22<01:05,  6.85it/s][A

Validation DataLoader 0:  25%|██▌       | 153/602 [00:22<01:05,  6.89it/s][A

Validation DataLoader 0:  26%|██▌       | 154/602 [00:22<01:05,  6.88it/s][A

Validation DataLoader 0:  26%|██▌       | 155/602 [00:22<01:05,  6.85it/s][A

Validation DataLoader 0:  26%|██▌       | 156/602 [00:22<01:04,  6.89it/s][A

Validation DataLoader 0:  26%|██▌       | 157/602 [00:22<01:04,  6.87it/s][A

Validation DataLoader 0:  26%|██▌       | 158/602 [00:23<01:04,  6.85it/s][A

Validation DataLoader 0:  26%|██▋       | 159/602 [00:23<01:04,  6.88it/s][A

Validation DataLoader 0:  27%|██▋       | 160/602 [00:23<01:04,  6.87it/s][A

Validation DataLoader 0:  27%|██▋       | 161/602 [00:23<01:04,  6.85it/s][A

Validation DataLoader 0:  27%|██▋       | 162/602 [00:23<01:03,  6.89it/s][A

Validation DataLoader 0:  27%|██▋       | 163/602 [00:23<01:03,  6.87it/s][A

Validation DataLoader 0:  27%|██▋       | 164/602 [00:23<01:03,  6.85it/s][A

Validation DataLoader 0:  27%|██▋       | 165/602 [00:23<01:03,  6.89it/s][A

Validation DataLoader 0:  28%|██▊       | 166/602 [00:24<01:03,  6.86it/s][A

Validation DataLoader 0:  28%|██▊       | 167/602 [00:24<01:03,  6.84it/s][A

Validation DataLoader 0:  28%|██▊       | 168/602 [00:24<01:03,  6.88it/s][A

Validation DataLoader 0:  28%|██▊       | 169/602 [00:24<01:03,  6.85it/s][A

Validation DataLoader 0:  28%|██▊       | 170/602 [00:24<01:03,  6.84it/s][A

Validation DataLoader 0:  28%|██▊       | 171/602 [00:24<01:02,  6.88it/s][A

Validation DataLoader 0:  29%|██▊       | 172/602 [00:25<01:02,  6.85it/s][A

Validation DataLoader 0:  29%|██▊       | 173/602 [00:25<01:02,  6.84it/s][A

Validation DataLoader 0:  29%|██▉       | 174/602 [00:25<01:02,  6.88it/s][A

Validation DataLoader 0:  29%|██▉       | 175/602 [00:25<01:02,  6.84it/s][A

Validation DataLoader 0:  29%|██▉       | 176/602 [00:25<01:02,  6.84it/s][A

Validation DataLoader 0:  29%|██▉       | 177/602 [00:25<01:01,  6.87it/s][A

Validation DataLoader 0:  30%|██▉       | 178/602 [00:26<01:02,  6.84it/s][A

Validation DataLoader 0:  30%|██▉       | 179/602 [00:26<01:01,  6.83it/s][A

Validation DataLoader 0:  30%|██▉       | 180/602 [00:26<01:01,  6.86it/s][A

Validation DataLoader 0:  30%|███       | 181/602 [00:26<01:01,  6.82it/s][A

Validation DataLoader 0:  30%|███       | 182/602 [00:26<01:01,  6.85it/s][A

Validation DataLoader 0:  30%|███       | 183/602 [00:26<01:00,  6.88it/s][A

Validation DataLoader 0:  31%|███       | 184/602 [00:26<01:01,  6.82it/s][A

Validation DataLoader 0:  31%|███       | 185/602 [00:26<01:00,  6.86it/s][A

Validation DataLoader 0:  31%|███       | 186/602 [00:27<01:00,  6.89it/s][A

Validation DataLoader 0:  31%|███       | 187/602 [00:27<01:00,  6.81it/s][A

Validation DataLoader 0:  31%|███       | 188/602 [00:27<01:00,  6.84it/s][A

Validation DataLoader 0:  31%|███▏      | 189/602 [00:27<01:00,  6.88it/s][A

Validation DataLoader 0:  32%|███▏      | 190/602 [00:27<01:00,  6.82it/s][A

Validation DataLoader 0:  32%|███▏      | 191/602 [00:27<01:00,  6.85it/s][A

Validation DataLoader 0:  32%|███▏      | 192/602 [00:27<00:59,  6.88it/s][A

Validation DataLoader 0:  32%|███▏      | 193/602 [00:28<01:00,  6.81it/s][A

Validation DataLoader 0:  32%|███▏      | 194/602 [00:28<00:59,  6.84it/s][A

Validation DataLoader 0:  32%|███▏      | 195/602 [00:28<00:59,  6.88it/s][A

Validation DataLoader 0:  33%|███▎      | 196/602 [00:28<00:59,  6.81it/s][A

Validation DataLoader 0:  33%|███▎      | 197/602 [00:28<00:59,  6.84it/s][A

Validation DataLoader 0:  33%|███▎      | 198/602 [00:28<00:58,  6.87it/s][A

Validation DataLoader 0:  33%|███▎      | 199/602 [00:29<00:59,  6.80it/s][A

Validation DataLoader 0:  33%|███▎      | 200/602 [00:29<00:58,  6.84it/s][A

Validation DataLoader 0:  33%|███▎      | 201/602 [00:29<00:58,  6.87it/s][A

Validation DataLoader 0:  34%|███▎      | 202/602 [00:29<00:58,  6.79it/s][A

Validation DataLoader 0:  34%|███▎      | 203/602 [00:29<00:58,  6.82it/s][A

Validation DataLoader 0:  34%|███▍      | 204/602 [00:29<00:58,  6.85it/s][A

Validation DataLoader 0:  34%|███▍      | 205/602 [00:30<00:58,  6.80it/s][A

Validation DataLoader 0:  34%|███▍      | 206/602 [00:30<00:58,  6.82it/s][A

Validation DataLoader 0:  34%|███▍      | 207/602 [00:30<00:57,  6.85it/s][A

Validation DataLoader 0:  35%|███▍      | 208/602 [00:30<00:58,  6.78it/s][A

Validation DataLoader 0:  35%|███▍      | 209/602 [00:30<00:57,  6.81it/s][A

Validation DataLoader 0:  35%|███▍      | 210/602 [00:30<00:57,  6.84it/s][A

Validation DataLoader 0:  35%|███▌      | 211/602 [00:31<00:57,  6.78it/s][A

Validation DataLoader 0:  35%|███▌      | 212/602 [00:31<00:57,  6.80it/s][A

Validation DataLoader 0:  35%|███▌      | 213/602 [00:31<00:56,  6.83it/s][A

Validation DataLoader 0:  36%|███▌      | 214/602 [00:31<00:57,  6.76it/s][A

Validation DataLoader 0:  36%|███▌      | 215/602 [00:31<00:56,  6.79it/s][A

Validation DataLoader 0:  36%|███▌      | 216/602 [00:31<00:56,  6.82it/s][A

Validation DataLoader 0:  36%|███▌      | 217/602 [00:32<00:56,  6.76it/s][A

Validation DataLoader 0:  36%|███▌      | 218/602 [00:32<00:56,  6.79it/s][A

Validation DataLoader 0:  36%|███▋      | 219/602 [00:32<00:56,  6.81it/s][A

Validation DataLoader 0:  37%|███▋      | 220/602 [00:32<00:56,  6.75it/s][A

Validation DataLoader 0:  37%|███▋      | 221/602 [00:32<00:56,  6.78it/s][A

Validation DataLoader 0:  37%|███▋      | 222/602 [00:32<00:55,  6.81it/s][A

Validation DataLoader 0:  37%|███▋      | 223/602 [00:33<00:56,  6.75it/s][A

Validation DataLoader 0:  37%|███▋      | 224/602 [00:33<00:55,  6.77it/s][A

Validation DataLoader 0:  37%|███▋      | 225/602 [00:33<00:55,  6.80it/s][A

Validation DataLoader 0:  38%|███▊      | 226/602 [00:33<00:55,  6.75it/s][A

Validation DataLoader 0:  38%|███▊      | 227/602 [00:33<00:55,  6.77it/s][A

Validation DataLoader 0:  38%|███▊      | 228/602 [00:33<00:54,  6.80it/s][A

Validation DataLoader 0:  38%|███▊      | 229/602 [00:33<00:55,  6.75it/s][A

Validation DataLoader 0:  38%|███▊      | 230/602 [00:33<00:54,  6.78it/s][A

Validation DataLoader 0:  38%|███▊      | 231/602 [00:33<00:54,  6.81it/s][A

Validation DataLoader 0:  39%|███▊      | 232/602 [00:34<00:54,  6.75it/s][A

Validation DataLoader 0:  39%|███▊      | 233/602 [00:34<00:54,  6.77it/s][A

Validation DataLoader 0:  39%|███▉      | 234/602 [00:34<00:54,  6.80it/s][A

Validation DataLoader 0:  39%|███▉      | 235/602 [00:34<00:54,  6.74it/s][A

Validation DataLoader 0:  39%|███▉      | 236/602 [00:34<00:54,  6.76it/s][A

Validation DataLoader 0:  39%|███▉      | 237/602 [00:34<00:53,  6.79it/s][A

Validation DataLoader 0:  40%|███▉      | 238/602 [00:35<00:54,  6.74it/s][A

Validation DataLoader 0:  40%|███▉      | 239/602 [00:35<00:53,  6.76it/s][A

Validation DataLoader 0:  40%|███▉      | 240/602 [00:35<00:53,  6.79it/s][A

Validation DataLoader 0:  40%|████      | 241/602 [00:35<00:53,  6.75it/s][A

Validation DataLoader 0:  40%|████      | 242/602 [00:35<00:53,  6.78it/s][A

Validation DataLoader 0:  40%|████      | 243/602 [00:35<00:52,  6.80it/s][A

Validation DataLoader 0:  41%|████      | 244/602 [00:36<00:53,  6.74it/s][A

Validation DataLoader 0:  41%|████      | 245/602 [00:36<00:52,  6.77it/s][A

Validation DataLoader 0:  41%|████      | 246/602 [00:36<00:52,  6.79it/s][A

Validation DataLoader 0:  41%|████      | 247/602 [00:36<00:52,  6.73it/s][A

Validation DataLoader 0:  41%|████      | 248/602 [00:36<00:52,  6.75it/s][A

Validation DataLoader 0:  41%|████▏     | 249/602 [00:36<00:52,  6.78it/s][A

Validation DataLoader 0:  42%|████▏     | 250/602 [00:37<00:52,  6.71it/s][A

Validation DataLoader 0:  42%|████▏     | 251/602 [00:37<00:52,  6.74it/s][A

Validation DataLoader 0:  42%|████▏     | 252/602 [00:37<00:51,  6.76it/s][A

Validation DataLoader 0:  42%|████▏     | 253/602 [00:37<00:51,  6.72it/s][A

Validation DataLoader 0:  42%|████▏     | 254/602 [00:37<00:51,  6.74it/s][A

Validation DataLoader 0:  42%|████▏     | 255/602 [00:37<00:51,  6.76it/s][A

Validation DataLoader 0:  43%|████▎     | 256/602 [00:38<00:51,  6.72it/s][A

Validation DataLoader 0:  43%|████▎     | 257/602 [00:38<00:51,  6.75it/s][A

Validation DataLoader 0:  43%|████▎     | 258/602 [00:38<00:50,  6.77it/s][A

Validation DataLoader 0:  43%|████▎     | 259/602 [00:38<00:51,  6.72it/s][A

Validation DataLoader 0:  43%|████▎     | 260/602 [00:38<00:50,  6.75it/s][A

Validation DataLoader 0:  43%|████▎     | 261/602 [00:38<00:50,  6.77it/s][A

Validation DataLoader 0:  44%|████▎     | 262/602 [00:38<00:50,  6.73it/s][A

Validation DataLoader 0:  44%|████▎     | 263/602 [00:38<00:50,  6.75it/s][A

Validation DataLoader 0:  44%|████▍     | 264/602 [00:38<00:49,  6.77it/s][A

Validation DataLoader 0:  44%|████▍     | 265/602 [00:39<00:49,  6.74it/s][A

Validation DataLoader 0:  44%|████▍     | 266/602 [00:39<00:49,  6.76it/s][A

Validation DataLoader 0:  44%|████▍     | 267/602 [00:39<00:49,  6.78it/s][A

Validation DataLoader 0:  45%|████▍     | 268/602 [00:39<00:49,  6.75it/s][A

Validation DataLoader 0:  45%|████▍     | 269/602 [00:39<00:49,  6.77it/s][A

Validation DataLoader 0:  45%|████▍     | 270/602 [00:39<00:48,  6.79it/s][A

Validation DataLoader 0:  45%|████▌     | 271/602 [00:40<00:49,  6.74it/s][A

Validation DataLoader 0:  45%|████▌     | 272/602 [00:40<00:48,  6.76it/s][A

Validation DataLoader 0:  45%|████▌     | 273/602 [00:40<00:48,  6.78it/s][A

Validation DataLoader 0:  46%|████▌     | 274/602 [00:40<00:48,  6.75it/s][A

Validation DataLoader 0:  46%|████▌     | 275/602 [00:40<00:48,  6.77it/s][A

Validation DataLoader 0:  46%|████▌     | 276/602 [00:40<00:48,  6.79it/s][A

Validation DataLoader 0:  46%|████▌     | 277/602 [00:41<00:48,  6.75it/s][A

Validation DataLoader 0:  46%|████▌     | 278/602 [00:41<00:47,  6.77it/s][A

Validation DataLoader 0:  46%|████▋     | 279/602 [00:41<00:47,  6.79it/s][A

Validation DataLoader 0:  47%|████▋     | 280/602 [00:41<00:47,  6.75it/s][A

Validation DataLoader 0:  47%|████▋     | 281/602 [00:41<00:47,  6.77it/s][A

Validation DataLoader 0:  47%|████▋     | 282/602 [00:41<00:47,  6.79it/s][A

Validation DataLoader 0:  47%|████▋     | 283/602 [00:41<00:47,  6.75it/s][A

Validation DataLoader 0:  47%|████▋     | 284/602 [00:41<00:46,  6.77it/s][A

Validation DataLoader 0:  47%|████▋     | 285/602 [00:41<00:46,  6.79it/s][A

Validation DataLoader 0:  48%|████▊     | 286/602 [00:42<00:46,  6.76it/s][A

Validation DataLoader 0:  48%|████▊     | 287/602 [00:42<00:46,  6.78it/s][A

Validation DataLoader 0:  48%|████▊     | 288/602 [00:42<00:46,  6.80it/s][A

Validation DataLoader 0:  48%|████▊     | 289/602 [00:42<00:46,  6.75it/s][A

Validation DataLoader 0:  48%|████▊     | 290/602 [00:42<00:46,  6.77it/s][A

Validation DataLoader 0:  48%|████▊     | 291/602 [00:42<00:45,  6.79it/s][A

Validation DataLoader 0:  49%|████▊     | 292/602 [00:43<00:45,  6.75it/s][A

Validation DataLoader 0:  49%|████▊     | 293/602 [00:43<00:45,  6.77it/s][A

Validation DataLoader 0:  49%|████▉     | 294/602 [00:43<00:45,  6.79it/s][A

Validation DataLoader 0:  49%|████▉     | 295/602 [00:43<00:45,  6.75it/s][A

Validation DataLoader 0:  49%|████▉     | 296/602 [00:43<00:45,  6.77it/s][A

Validation DataLoader 0:  49%|████▉     | 297/602 [00:43<00:44,  6.79it/s][A

Validation DataLoader 0:  50%|████▉     | 298/602 [00:44<00:45,  6.75it/s][A

Validation DataLoader 0:  50%|████▉     | 299/602 [00:44<00:44,  6.77it/s][A

Validation DataLoader 0:  50%|████▉     | 300/602 [00:44<00:44,  6.79it/s][A

Validation DataLoader 0:  50%|█████     | 301/602 [00:44<00:44,  6.75it/s][A

Validation DataLoader 0:  50%|█████     | 302/602 [00:44<00:44,  6.77it/s][A

Validation DataLoader 0:  50%|█████     | 303/602 [00:44<00:44,  6.79it/s][A

Validation DataLoader 0:  50%|█████     | 304/602 [00:45<00:44,  6.75it/s][A

Validation DataLoader 0:  51%|█████     | 305/602 [00:45<00:43,  6.77it/s][A

Validation DataLoader 0:  51%|█████     | 306/602 [00:45<00:43,  6.79it/s][A

Validation DataLoader 0:  51%|█████     | 307/602 [00:45<00:43,  6.74it/s][A

Validation DataLoader 0:  51%|█████     | 308/602 [00:45<00:43,  6.76it/s][A

Validation DataLoader 0:  51%|█████▏    | 309/602 [00:45<00:43,  6.78it/s][A

Validation DataLoader 0:  51%|█████▏    | 310/602 [00:46<00:43,  6.72it/s][A

Validation DataLoader 0:  52%|█████▏    | 311/602 [00:46<00:43,  6.74it/s][A

Validation DataLoader 0:  52%|█████▏    | 312/602 [00:46<00:42,  6.76it/s][A

Validation DataLoader 0:  52%|█████▏    | 313/602 [00:46<00:43,  6.72it/s][A

Validation DataLoader 0:  52%|█████▏    | 314/602 [00:46<00:42,  6.74it/s][A

Validation DataLoader 0:  52%|█████▏    | 315/602 [00:46<00:42,  6.75it/s][A

Validation DataLoader 0:  52%|█████▏    | 316/602 [00:46<00:42,  6.73it/s][A

Validation DataLoader 0:  53%|█████▎    | 317/602 [00:46<00:42,  6.75it/s][A

Validation DataLoader 0:  53%|█████▎    | 318/602 [00:46<00:41,  6.77it/s][A

Validation DataLoader 0:  53%|█████▎    | 319/602 [00:47<00:42,  6.73it/s][A

Validation DataLoader 0:  53%|█████▎    | 320/602 [00:47<00:41,  6.75it/s][A

Validation DataLoader 0:  53%|█████▎    | 321/602 [00:47<00:41,  6.77it/s][A

Validation DataLoader 0:  53%|█████▎    | 322/602 [00:47<00:41,  6.72it/s][A

Validation DataLoader 0:  54%|█████▎    | 323/602 [00:47<00:41,  6.74it/s][A

Validation DataLoader 0:  54%|█████▍    | 324/602 [00:47<00:41,  6.76it/s][A

Validation DataLoader 0:  54%|█████▍    | 325/602 [00:48<00:41,  6.72it/s][A

Validation DataLoader 0:  54%|█████▍    | 326/602 [00:48<00:40,  6.74it/s][A

Validation DataLoader 0:  54%|█████▍    | 327/602 [00:48<00:40,  6.76it/s][A

Validation DataLoader 0:  54%|█████▍    | 328/602 [00:48<00:40,  6.71it/s][A

Validation DataLoader 0:  55%|█████▍    | 329/602 [00:48<00:40,  6.73it/s][A

Validation DataLoader 0:  55%|█████▍    | 330/602 [00:48<00:40,  6.75it/s][A

Validation DataLoader 0:  55%|█████▍    | 331/602 [00:49<00:40,  6.71it/s][A

Validation DataLoader 0:  55%|█████▌    | 332/602 [00:49<00:40,  6.73it/s][A

Validation DataLoader 0:  55%|█████▌    | 333/602 [00:49<00:39,  6.75it/s][A

Validation DataLoader 0:  55%|█████▌    | 334/602 [00:49<00:39,  6.71it/s][A

Validation DataLoader 0:  56%|█████▌    | 335/602 [00:49<00:39,  6.73it/s][A

Validation DataLoader 0:  56%|█████▌    | 336/602 [00:49<00:39,  6.75it/s][A

Validation DataLoader 0:  56%|█████▌    | 337/602 [00:50<00:39,  6.73it/s][A

Validation DataLoader 0:  56%|█████▌    | 338/602 [00:50<00:39,  6.75it/s][A

Validation DataLoader 0:  56%|█████▋    | 339/602 [00:50<00:38,  6.76it/s][A

Validation DataLoader 0:  56%|█████▋    | 340/602 [00:50<00:38,  6.74it/s][A

Validation DataLoader 0:  57%|█████▋    | 341/602 [00:50<00:38,  6.76it/s][A

Validation DataLoader 0:  57%|█████▋    | 342/602 [00:50<00:38,  6.78it/s][A

Validation DataLoader 0:  57%|█████▋    | 343/602 [00:50<00:38,  6.73it/s][A

Validation DataLoader 0:  57%|█████▋    | 344/602 [00:50<00:38,  6.75it/s][A

Validation DataLoader 0:  57%|█████▋    | 345/602 [00:50<00:37,  6.77it/s][A

Validation DataLoader 0:  57%|█████▋    | 346/602 [00:51<00:38,  6.73it/s][A

Validation DataLoader 0:  58%|█████▊    | 347/602 [00:51<00:37,  6.75it/s][A

Validation DataLoader 0:  58%|█████▊    | 348/602 [00:51<00:37,  6.77it/s][A

Validation DataLoader 0:  58%|█████▊    | 349/602 [00:51<00:37,  6.74it/s][A

Validation DataLoader 0:  58%|█████▊    | 350/602 [00:51<00:37,  6.76it/s][A

Validation DataLoader 0:  58%|█████▊    | 351/602 [00:51<00:37,  6.77it/s][A

Validation DataLoader 0:  58%|█████▊    | 352/602 [00:52<00:37,  6.74it/s][A

Validation DataLoader 0:  59%|█████▊    | 353/602 [00:52<00:36,  6.76it/s][A

Validation DataLoader 0:  59%|█████▉    | 354/602 [00:52<00:36,  6.76it/s][A

Validation DataLoader 0:  59%|█████▉    | 355/602 [00:52<00:36,  6.75it/s][A

Validation DataLoader 0:  59%|█████▉    | 356/602 [00:52<00:36,  6.76it/s][A

Validation DataLoader 0:  59%|█████▉    | 357/602 [00:52<00:36,  6.76it/s][A

Validation DataLoader 0:  59%|█████▉    | 358/602 [00:53<00:36,  6.74it/s][A

Validation DataLoader 0:  60%|█████▉    | 359/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|█████▉    | 360/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|█████▉    | 361/602 [00:53<00:35,  6.74it/s][A

Validation DataLoader 0:  60%|██████    | 362/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|██████    | 363/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|██████    | 364/602 [00:54<00:35,  6.74it/s][A

Validation DataLoader 0:  61%|██████    | 365/602 [00:54<00:35,  6.75it/s][A

Validation DataLoader 0:  61%|██████    | 366/602 [00:54<00:34,  6.76it/s][A

Validation DataLoader 0:  61%|██████    | 367/602 [00:54<00:34,  6.73it/s][A

Validation DataLoader 0:  61%|██████    | 368/602 [00:54<00:34,  6.75it/s][A

Validation DataLoader 0:  61%|██████▏   | 369/602 [00:54<00:34,  6.76it/s][A

Validation DataLoader 0:  61%|██████▏   | 370/602 [00:55<00:34,  6.72it/s][A

Validation DataLoader 0:  62%|██████▏   | 371/602 [00:55<00:34,  6.74it/s][A

Validation DataLoader 0:  62%|██████▏   | 372/602 [00:55<00:34,  6.76it/s][A

Validation DataLoader 0:  62%|██████▏   | 373/602 [00:55<00:34,  6.73it/s][A

Validation DataLoader 0:  62%|██████▏   | 374/602 [00:55<00:33,  6.74it/s][A

Validation DataLoader 0:  62%|██████▏   | 375/602 [00:55<00:33,  6.76it/s][A

Validation DataLoader 0:  62%|██████▏   | 376/602 [00:55<00:33,  6.72it/s][A

Validation DataLoader 0:  63%|██████▎   | 377/602 [00:55<00:33,  6.74it/s][A

Validation DataLoader 0:  63%|██████▎   | 378/602 [00:55<00:33,  6.76it/s][A

Validation DataLoader 0:  63%|██████▎   | 379/602 [00:56<00:33,  6.71it/s][A

Validation DataLoader 0:  63%|██████▎   | 380/602 [00:56<00:32,  6.73it/s][A

Validation DataLoader 0:  63%|██████▎   | 381/602 [00:56<00:32,  6.74it/s][A

Validation DataLoader 0:  63%|██████▎   | 382/602 [00:56<00:32,  6.71it/s][A

Validation DataLoader 0:  64%|██████▎   | 383/602 [00:56<00:32,  6.72it/s][A

Validation DataLoader 0:  64%|██████▍   | 384/602 [00:56<00:32,  6.74it/s][A

Validation DataLoader 0:  64%|██████▍   | 385/602 [00:57<00:32,  6.70it/s][A

Validation DataLoader 0:  64%|██████▍   | 386/602 [00:57<00:32,  6.71it/s][A

Validation DataLoader 0:  64%|██████▍   | 387/602 [00:57<00:31,  6.73it/s][A

Validation DataLoader 0:  64%|██████▍   | 388/602 [00:57<00:31,  6.70it/s][A

Validation DataLoader 0:  65%|██████▍   | 389/602 [00:57<00:31,  6.72it/s][A

Validation DataLoader 0:  65%|██████▍   | 390/602 [00:57<00:31,  6.73it/s][A

Validation DataLoader 0:  65%|██████▍   | 391/602 [00:58<00:31,  6.71it/s][A

Validation DataLoader 0:  65%|██████▌   | 392/602 [00:58<00:31,  6.72it/s][A

Validation DataLoader 0:  65%|██████▌   | 393/602 [00:58<00:31,  6.74it/s][A

Validation DataLoader 0:  65%|██████▌   | 394/602 [00:58<00:31,  6.70it/s][A

Validation DataLoader 0:  66%|██████▌   | 395/602 [00:58<00:30,  6.72it/s][A

Validation DataLoader 0:  66%|██████▌   | 396/602 [00:58<00:30,  6.73it/s][A

Validation DataLoader 0:  66%|██████▌   | 397/602 [00:59<00:30,  6.70it/s][A

Validation DataLoader 0:  66%|██████▌   | 398/602 [00:59<00:30,  6.72it/s][A

Validation DataLoader 0:  66%|██████▋   | 399/602 [00:59<00:30,  6.73it/s][A

Validation DataLoader 0:  66%|██████▋   | 400/602 [00:59<00:30,  6.70it/s][A

Validation DataLoader 0:  67%|██████▋   | 401/602 [00:59<00:29,  6.72it/s][A

Validation DataLoader 0:  67%|██████▋   | 402/602 [00:59<00:29,  6.73it/s][A

Validation DataLoader 0:  67%|██████▋   | 403/602 [01:00<00:29,  6.71it/s][A

Validation DataLoader 0:  67%|██████▋   | 404/602 [01:00<00:29,  6.72it/s][A

Validation DataLoader 0:  67%|██████▋   | 405/602 [01:00<00:29,  6.74it/s][A

Validation DataLoader 0:  67%|██████▋   | 406/602 [01:00<00:29,  6.71it/s][A

Validation DataLoader 0:  68%|██████▊   | 407/602 [01:00<00:29,  6.72it/s][A

Validation DataLoader 0:  68%|██████▊   | 408/602 [01:00<00:28,  6.74it/s][A

Validation DataLoader 0:  68%|██████▊   | 409/602 [01:00<00:28,  6.71it/s][A

Validation DataLoader 0:  68%|██████▊   | 410/602 [01:00<00:28,  6.72it/s][A

Validation DataLoader 0:  68%|██████▊   | 411/602 [01:00<00:28,  6.74it/s][A

Validation DataLoader 0:  68%|██████▊   | 412/602 [01:01<00:28,  6.71it/s][A

Validation DataLoader 0:  69%|██████▊   | 413/602 [01:01<00:28,  6.73it/s][A

Validation DataLoader 0:  69%|██████▉   | 414/602 [01:01<00:27,  6.74it/s][A

Validation DataLoader 0:  69%|██████▉   | 415/602 [01:01<00:27,  6.71it/s][A

Validation DataLoader 0:  69%|██████▉   | 416/602 [01:01<00:27,  6.73it/s][A

Validation DataLoader 0:  69%|██████▉   | 417/602 [01:01<00:27,  6.74it/s][A

Validation DataLoader 0:  69%|██████▉   | 418/602 [01:02<00:27,  6.71it/s][A

Validation DataLoader 0:  70%|██████▉   | 419/602 [01:02<00:27,  6.72it/s][A

Validation DataLoader 0:  70%|██████▉   | 420/602 [01:02<00:27,  6.74it/s][A

Validation DataLoader 0:  70%|██████▉   | 421/602 [01:02<00:27,  6.70it/s][A

Validation DataLoader 0:  70%|███████   | 422/602 [01:02<00:26,  6.72it/s][A

Validation DataLoader 0:  70%|███████   | 423/602 [01:02<00:26,  6.73it/s][A

Validation DataLoader 0:  70%|███████   | 424/602 [01:03<00:26,  6.71it/s][A

Validation DataLoader 0:  71%|███████   | 425/602 [01:03<00:26,  6.72it/s][A

Validation DataLoader 0:  71%|███████   | 426/602 [01:03<00:26,  6.74it/s][A

Validation DataLoader 0:  71%|███████   | 427/602 [01:03<00:26,  6.70it/s][A

Validation DataLoader 0:  71%|███████   | 428/602 [01:03<00:25,  6.71it/s][A

Validation DataLoader 0:  71%|███████▏  | 429/602 [01:03<00:25,  6.72it/s][A

Validation DataLoader 0:  71%|███████▏  | 430/602 [01:04<00:25,  6.70it/s][A

Validation DataLoader 0:  72%|███████▏  | 431/602 [01:04<00:25,  6.71it/s][A

Validation DataLoader 0:  72%|███████▏  | 432/602 [01:04<00:25,  6.72it/s][A

Validation DataLoader 0:  72%|███████▏  | 433/602 [01:04<00:25,  6.70it/s][A

Validation DataLoader 0:  72%|███████▏  | 434/602 [01:04<00:25,  6.71it/s][A

Validation DataLoader 0:  72%|███████▏  | 435/602 [01:04<00:24,  6.73it/s][A

Validation DataLoader 0:  72%|███████▏  | 436/602 [01:05<00:24,  6.70it/s][A

Validation DataLoader 0:  73%|███████▎  | 437/602 [01:05<00:24,  6.72it/s][A

Validation DataLoader 0:  73%|███████▎  | 438/602 [01:05<00:24,  6.73it/s][A

Validation DataLoader 0:  73%|███████▎  | 439/602 [01:05<00:24,  6.70it/s][A

Validation DataLoader 0:  73%|███████▎  | 440/602 [01:05<00:24,  6.72it/s][A

Validation DataLoader 0:  73%|███████▎  | 441/602 [01:05<00:23,  6.73it/s][A

Validation DataLoader 0:  73%|███████▎  | 442/602 [01:06<00:23,  6.69it/s][A

Validation DataLoader 0:  74%|███████▎  | 443/602 [01:06<00:23,  6.71it/s][A

Validation DataLoader 0:  74%|███████▍  | 444/602 [01:06<00:23,  6.72it/s][A

Validation DataLoader 0:  74%|███████▍  | 445/602 [01:06<00:23,  6.69it/s][A

Validation DataLoader 0:  74%|███████▍  | 446/602 [01:06<00:23,  6.71it/s][A

Validation DataLoader 0:  74%|███████▍  | 447/602 [01:06<00:23,  6.72it/s][A

Validation DataLoader 0:  74%|███████▍  | 448/602 [01:06<00:23,  6.69it/s][A

Validation DataLoader 0:  75%|███████▍  | 449/602 [01:06<00:22,  6.70it/s][A

Validation DataLoader 0:  75%|███████▍  | 450/602 [01:07<00:22,  6.72it/s][A

Validation DataLoader 0:  75%|███████▍  | 451/602 [01:07<00:22,  6.69it/s][A

Validation DataLoader 0:  75%|███████▌  | 452/602 [01:07<00:22,  6.70it/s][A

Validation DataLoader 0:  75%|███████▌  | 453/602 [01:07<00:22,  6.72it/s][A

Validation DataLoader 0:  75%|███████▌  | 454/602 [01:07<00:22,  6.70it/s][A

Validation DataLoader 0:  76%|███████▌  | 455/602 [01:07<00:21,  6.71it/s][A

Validation DataLoader 0:  76%|███████▌  | 456/602 [01:07<00:21,  6.73it/s][A

Validation DataLoader 0:  76%|███████▌  | 457/602 [01:08<00:21,  6.70it/s][A

Validation DataLoader 0:  76%|███████▌  | 458/602 [01:08<00:21,  6.71it/s][A

Validation DataLoader 0:  76%|███████▌  | 459/602 [01:08<00:21,  6.72it/s][A

Validation DataLoader 0:  76%|███████▋  | 460/602 [01:08<00:21,  6.69it/s][A

Validation DataLoader 0:  77%|███████▋  | 461/602 [01:08<00:21,  6.70it/s][A

Validation DataLoader 0:  77%|███████▋  | 462/602 [01:08<00:20,  6.71it/s][A

Validation DataLoader 0:  77%|███████▋  | 463/602 [01:09<00:20,  6.68it/s][A

Validation DataLoader 0:  77%|███████▋  | 464/602 [01:09<00:20,  6.69it/s][A

Validation DataLoader 0:  77%|███████▋  | 465/602 [01:09<00:20,  6.71it/s][A

Validation DataLoader 0:  77%|███████▋  | 466/602 [01:09<00:20,  6.67it/s][A

Validation DataLoader 0:  78%|███████▊  | 467/602 [01:09<00:20,  6.68it/s][A

Validation DataLoader 0:  78%|███████▊  | 468/602 [01:09<00:20,  6.69it/s][A

Validation DataLoader 0:  78%|███████▊  | 469/602 [01:10<00:19,  6.66it/s][A

Validation DataLoader 0:  78%|███████▊  | 470/602 [01:10<00:19,  6.68it/s][A

Validation DataLoader 0:  78%|███████▊  | 471/602 [01:10<00:19,  6.69it/s][A

Validation DataLoader 0:  78%|███████▊  | 472/602 [01:10<00:19,  6.66it/s][A

Validation DataLoader 0:  79%|███████▊  | 473/602 [01:10<00:19,  6.67it/s][A

Validation DataLoader 0:  79%|███████▊  | 474/602 [01:10<00:19,  6.69it/s][A

Validation DataLoader 0:  79%|███████▉  | 475/602 [01:11<00:19,  6.66it/s][A

Validation DataLoader 0:  79%|███████▉  | 476/602 [01:11<00:18,  6.67it/s][A

Validation DataLoader 0:  79%|███████▉  | 477/602 [01:11<00:18,  6.69it/s][A

Validation DataLoader 0:  79%|███████▉  | 478/602 [01:11<00:18,  6.66it/s][A

Validation DataLoader 0:  80%|███████▉  | 479/602 [01:11<00:18,  6.67it/s][A

Validation DataLoader 0:  80%|███████▉  | 480/602 [01:11<00:18,  6.69it/s][A

Validation DataLoader 0:  80%|███████▉  | 481/602 [01:12<00:18,  6.67it/s][A

Validation DataLoader 0:  80%|████████  | 482/602 [01:12<00:17,  6.68it/s][A

Validation DataLoader 0:  80%|████████  | 483/602 [01:12<00:17,  6.70it/s][A

Validation DataLoader 0:  80%|████████  | 484/602 [01:12<00:17,  6.67it/s][A

Validation DataLoader 0:  81%|████████  | 485/602 [01:12<00:17,  6.68it/s][A

Validation DataLoader 0:  81%|████████  | 486/602 [01:12<00:17,  6.70it/s][A

Validation DataLoader 0:  81%|████████  | 487/602 [01:12<00:17,  6.67it/s][A

Validation DataLoader 0:  81%|████████  | 488/602 [01:13<00:17,  6.68it/s][A

Validation DataLoader 0:  81%|████████  | 489/602 [01:13<00:16,  6.70it/s][A

Validation DataLoader 0:  81%|████████▏ | 490/602 [01:13<00:16,  6.67it/s][A

Validation DataLoader 0:  82%|████████▏ | 491/602 [01:13<00:16,  6.69it/s][A

Validation DataLoader 0:  82%|████████▏ | 492/602 [01:13<00:16,  6.70it/s][A

Validation DataLoader 0:  82%|████████▏ | 493/602 [01:13<00:16,  6.67it/s][A

Validation DataLoader 0:  82%|████████▏ | 494/602 [01:13<00:16,  6.69it/s][A

Validation DataLoader 0:  82%|████████▏ | 495/602 [01:13<00:15,  6.70it/s][A

Validation DataLoader 0:  82%|████████▏ | 496/602 [01:14<00:15,  6.68it/s][A

Validation DataLoader 0:  83%|████████▎ | 497/602 [01:14<00:15,  6.69it/s][A

Validation DataLoader 0:  83%|████████▎ | 498/602 [01:14<00:15,  6.70it/s][A

Validation DataLoader 0:  83%|████████▎ | 499/602 [01:14<00:15,  6.67it/s][A

Validation DataLoader 0:  83%|████████▎ | 500/602 [01:14<00:15,  6.69it/s][A

Validation DataLoader 0:  83%|████████▎ | 501/602 [01:14<00:15,  6.70it/s][A

Validation DataLoader 0:  83%|████████▎ | 502/602 [01:15<00:14,  6.67it/s][A

Validation DataLoader 0:  84%|████████▎ | 503/602 [01:15<00:14,  6.69it/s][A

Validation DataLoader 0:  84%|████████▎ | 504/602 [01:15<00:14,  6.70it/s][A

Validation DataLoader 0:  84%|████████▍ | 505/602 [01:15<00:14,  6.68it/s][A

Validation DataLoader 0:  84%|████████▍ | 506/602 [01:15<00:14,  6.69it/s][A

Validation DataLoader 0:  84%|████████▍ | 507/602 [01:15<00:14,  6.70it/s][A

Validation DataLoader 0:  84%|████████▍ | 508/602 [01:16<00:14,  6.68it/s][A

Validation DataLoader 0:  85%|████████▍ | 509/602 [01:16<00:13,  6.69it/s][A

Validation DataLoader 0:  85%|████████▍ | 510/602 [01:16<00:13,  6.70it/s][A

Validation DataLoader 0:  85%|████████▍ | 511/602 [01:16<00:13,  6.68it/s][A

Validation DataLoader 0:  85%|████████▌ | 512/602 [01:16<00:13,  6.69it/s][A

Validation DataLoader 0:  85%|████████▌ | 513/602 [01:16<00:13,  6.70it/s][A

Validation DataLoader 0:  85%|████████▌ | 514/602 [01:16<00:13,  6.68it/s][A

Validation DataLoader 0:  86%|████████▌ | 515/602 [01:16<00:13,  6.69it/s][A

Validation DataLoader 0:  86%|████████▌ | 516/602 [01:17<00:12,  6.70it/s][A

Validation DataLoader 0:  86%|████████▌ | 517/602 [01:17<00:12,  6.67it/s][A

Validation DataLoader 0:  86%|████████▌ | 518/602 [01:17<00:12,  6.69it/s][A

Validation DataLoader 0:  86%|████████▌ | 519/602 [01:17<00:12,  6.70it/s][A

Validation DataLoader 0:  86%|████████▋ | 520/602 [01:17<00:12,  6.67it/s][A

Validation DataLoader 0:  87%|████████▋ | 521/602 [01:17<00:12,  6.68it/s][A

Validation DataLoader 0:  87%|████████▋ | 522/602 [01:17<00:11,  6.69it/s][A

Validation DataLoader 0:  87%|████████▋ | 523/602 [01:18<00:11,  6.67it/s][A

Validation DataLoader 0:  87%|████████▋ | 524/602 [01:18<00:11,  6.68it/s][A

Validation DataLoader 0:  87%|████████▋ | 525/602 [01:18<00:11,  6.69it/s][A

Validation DataLoader 0:  87%|████████▋ | 526/602 [01:18<00:11,  6.67it/s][A

Validation DataLoader 0:  88%|████████▊ | 527/602 [01:18<00:11,  6.69it/s][A

Validation DataLoader 0:  88%|████████▊ | 528/602 [01:18<00:11,  6.70it/s][A

Validation DataLoader 0:  88%|████████▊ | 529/602 [01:19<00:10,  6.68it/s][A

Validation DataLoader 0:  88%|████████▊ | 530/602 [01:19<00:10,  6.69it/s][A

Validation DataLoader 0:  88%|████████▊ | 531/602 [01:19<00:10,  6.70it/s][A

Validation DataLoader 0:  88%|████████▊ | 532/602 [01:19<00:10,  6.68it/s][A

Validation DataLoader 0:  89%|████████▊ | 533/602 [01:19<00:10,  6.69it/s][A

Validation DataLoader 0:  89%|████████▊ | 534/602 [01:19<00:10,  6.71it/s][A

Validation DataLoader 0:  89%|████████▉ | 535/602 [01:20<00:10,  6.68it/s][A

Validation DataLoader 0:  89%|████████▉ | 536/602 [01:20<00:09,  6.69it/s][A

Validation DataLoader 0:  89%|████████▉ | 537/602 [01:20<00:09,  6.70it/s][A

Validation DataLoader 0:  89%|████████▉ | 538/602 [01:20<00:09,  6.68it/s][A

Validation DataLoader 0:  90%|████████▉ | 539/602 [01:20<00:09,  6.69it/s][A

Validation DataLoader 0:  90%|████████▉ | 540/602 [01:20<00:09,  6.70it/s][A

Validation DataLoader 0:  90%|████████▉ | 541/602 [01:21<00:09,  6.68it/s][A

Validation DataLoader 0:  90%|█████████ | 542/602 [01:21<00:08,  6.69it/s][A

Validation DataLoader 0:  90%|█████████ | 543/602 [01:21<00:08,  6.70it/s][A

Validation DataLoader 0:  90%|█████████ | 544/602 [01:21<00:08,  6.68it/s][A

Validation DataLoader 0:  91%|█████████ | 545/602 [01:21<00:08,  6.69it/s][A

Validation DataLoader 0:  91%|█████████ | 546/602 [01:21<00:08,  6.70it/s][A

Validation DataLoader 0:  91%|█████████ | 547/602 [01:21<00:08,  6.69it/s][A

Validation DataLoader 0:  91%|█████████ | 548/602 [01:21<00:08,  6.70it/s][A

Validation DataLoader 0:  91%|█████████ | 549/602 [01:21<00:07,  6.71it/s][A

Validation DataLoader 0:  91%|█████████▏| 550/602 [01:22<00:07,  6.69it/s][A

Validation DataLoader 0:  92%|█████████▏| 551/602 [01:22<00:07,  6.70it/s][A

Validation DataLoader 0:  92%|█████████▏| 552/602 [01:22<00:07,  6.71it/s][A

Validation DataLoader 0:  92%|█████████▏| 553/602 [01:22<00:07,  6.69it/s][A

Validation DataLoader 0:  92%|█████████▏| 554/602 [01:22<00:07,  6.70it/s][A

Validation DataLoader 0:  92%|█████████▏| 555/602 [01:22<00:06,  6.71it/s][A

Validation DataLoader 0:  92%|█████████▏| 556/602 [01:23<00:06,  6.69it/s][A

Validation DataLoader 0:  93%|█████████▎| 557/602 [01:23<00:06,  6.70it/s][A

Validation DataLoader 0:  93%|█████████▎| 558/602 [01:23<00:06,  6.71it/s][A

Validation DataLoader 0:  93%|█████████▎| 559/602 [01:23<00:06,  6.69it/s][A

Validation DataLoader 0:  93%|█████████▎| 560/602 [01:23<00:06,  6.70it/s][A

Validation DataLoader 0:  93%|█████████▎| 561/602 [01:23<00:06,  6.71it/s][A

Validation DataLoader 0:  93%|█████████▎| 562/602 [01:23<00:05,  6.69it/s][A

Validation DataLoader 0:  94%|█████████▎| 563/602 [01:23<00:05,  6.70it/s][A

Validation DataLoader 0:  94%|█████████▎| 564/602 [01:23<00:05,  6.72it/s][A

Validation DataLoader 0:  94%|█████████▍| 565/602 [01:24<00:05,  6.69it/s][A

Validation DataLoader 0:  94%|█████████▍| 566/602 [01:24<00:05,  6.70it/s][A

Validation DataLoader 0:  94%|█████████▍| 567/602 [01:24<00:05,  6.71it/s][A

Validation DataLoader 0:  94%|█████████▍| 568/602 [01:24<00:05,  6.69it/s][A

Validation DataLoader 0:  95%|█████████▍| 569/602 [01:24<00:04,  6.70it/s][A

Validation DataLoader 0:  95%|█████████▍| 570/602 [01:24<00:04,  6.71it/s][A

Validation DataLoader 0:  95%|█████████▍| 571/602 [01:25<00:04,  6.69it/s][A

Validation DataLoader 0:  95%|█████████▌| 572/602 [01:25<00:04,  6.70it/s][A

Validation DataLoader 0:  95%|█████████▌| 573/602 [01:25<00:04,  6.71it/s][A

Validation DataLoader 0:  95%|█████████▌| 574/602 [01:25<00:04,  6.69it/s][A

Validation DataLoader 0:  96%|█████████▌| 575/602 [01:25<00:04,  6.70it/s][A

Validation DataLoader 0:  96%|█████████▌| 576/602 [01:25<00:03,  6.71it/s][A

Validation DataLoader 0:  96%|█████████▌| 577/602 [01:26<00:03,  6.69it/s][A

Validation DataLoader 0:  96%|█████████▌| 578/602 [01:26<00:03,  6.70it/s][A

Validation DataLoader 0:  96%|█████████▌| 579/602 [01:26<00:03,  6.71it/s][A

Validation DataLoader 0:  96%|█████████▋| 580/602 [01:26<00:03,  6.69it/s][A

Validation DataLoader 0:  97%|█████████▋| 581/602 [01:26<00:03,  6.70it/s][A

Validation DataLoader 0:  97%|█████████▋| 582/602 [01:26<00:02,  6.71it/s][A

Validation DataLoader 0:  97%|█████████▋| 583/602 [01:27<00:02,  6.69it/s][A

Validation DataLoader 0:  97%|█████████▋| 584/602 [01:27<00:02,  6.70it/s][A

Validation DataLoader 0:  97%|█████████▋| 585/602 [01:27<00:02,  6.71it/s][A

Validation DataLoader 0:  97%|█████████▋| 586/602 [01:27<00:02,  6.69it/s][A

Validation DataLoader 0:  98%|█████████▊| 587/602 [01:27<00:02,  6.70it/s][A

Validation DataLoader 0:  98%|█████████▊| 588/602 [01:27<00:02,  6.71it/s][A

Validation DataLoader 0:  98%|█████████▊| 589/602 [01:28<00:01,  6.69it/s][A

Validation DataLoader 0:  98%|█████████▊| 590/602 [01:28<00:01,  6.70it/s][A

Validation DataLoader 0:  98%|█████████▊| 591/602 [01:28<00:01,  6.71it/s][A

Validation DataLoader 0:  98%|█████████▊| 592/602 [01:28<00:01,  6.69it/s][A

Validation DataLoader 0:  99%|█████████▊| 593/602 [01:28<00:01,  6.70it/s][A

Validation DataLoader 0:  99%|█████████▊| 594/602 [01:28<00:01,  6.71it/s][A

Validation DataLoader 0:  99%|█████████▉| 595/602 [01:28<00:01,  6.69it/s][A

Validation DataLoader 0:  99%|█████████▉| 596/602 [01:28<00:00,  6.70it/s][A

Validation DataLoader 0:  99%|█████████▉| 597/602 [01:28<00:00,  6.71it/s][A

Validation DataLoader 0:  99%|█████████▉| 598/602 [01:29<00:00,  6.70it/s][A

Validation DataLoader 0: 100%|█████████▉| 599/602 [01:29<00:00,  6.71it/s][A

Validation DataLoader 0: 100%|█████████▉| 600/602 [01:29<00:00,  6.72it/s][A

Validation DataLoader 0: 100%|█████████▉| 601/602 [01:29<00:00,  6.70it/s][A

Validation DataLoader 0: 100%|██████████| 602/602 [01:29<00:00,  6.70it/s][A

                                                                          [A
Epoch 0: 100%|██████████| 3295/3295 [36:08<00:00,  1.52it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772]
Epoch 0: 100%|██████████| 3295/3295 [36:08<00:00,  1.52it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]Epoch 0, global step 3295: 'R@1 (val)' reached 0.36299 (best 0.36299), saving model to '/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=00-R@1 (val)=0.363.ckpt' as top 1

Epoch 0:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]           
Epoch 1:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 1/3295 [00:02<1:53:45,  0.48it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 1/3295 [00:02<1:53:47,  0.48it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 2/3295 [00:02<57:29,  0.95it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]  
Epoch 1:   0%|          | 2/3295 [00:02<57:30,  0.95it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 3/3295 [00:02<38:46,  1.41it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 3/3295 [00:02<38:46,  1.41it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 4/3295 [00:03<54:34,  1.01it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 4/3295 [00:03<54:34,  1.01it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 5/3295 [00:04<43:58,  1.25it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 5/3295 [00:04<43:58,  1.25it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 6/3295 [00:04<36:54,  1.49it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 6/3295 [00:04<36:54,  1.49it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 7/3295 [00:05<44:49,  1.22it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 7/3295 [00:05<44:49,  1.22it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 8/3295 [00:05<39:21,  1.39it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 8/3295 [00:05<39:21,  1.39it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 9/3295 [00:05<35:32,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 9/3295 [00:05<35:33,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 10/3295 [00:07<41:20,  1.32it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 10/3295 [00:07<41:20,  1.32it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 11/3295 [00:07<37:42,  1.45it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 11/3295 [00:07<37:42,  1.45it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 12/3295 [00:07<35:32,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 12/3295 [00:07<35:32,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 13/3295 [00:09<39:18,  1.39it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 13/3295 [00:09<39:18,  1.39it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 14/3295 [00:09<36:35,  1.49it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 14/3295 [00:09<36:36,  1.49it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 15/3295 [00:09<35:16,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 15/3295 [00:09<35:16,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 16/3295 [00:11<38:16,  1.43it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   0%|          | 16/3295 [00:11<38:16,  1.43it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 17/3295 [00:11<36:06,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 17/3295 [00:11<36:06,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 18/3295 [00:11<34:59,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 18/3295 [00:11<34:59,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 19/3295 [00:13<37:51,  1.44it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 19/3295 [00:13<37:51,  1.44it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 20/3295 [00:13<36:02,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 20/3295 [00:13<36:02,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 21/3295 [00:13<35:11,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 21/3295 [00:13<35:11,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 22/3295 [00:15<37:34,  1.45it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 22/3295 [00:15<37:34,  1.45it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 23/3295 [00:15<36:00,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 23/3295 [00:15<36:00,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 24/3295 [00:15<34:59,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 24/3295 [00:15<34:59,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 25/3295 [00:17<37:27,  1.46it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 25/3295 [00:17<37:27,  1.46it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 26/3295 [00:17<36:04,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 26/3295 [00:17<36:04,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 27/3295 [00:17<34:55,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 27/3295 [00:17<34:55,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 28/3295 [00:18<36:54,  1.48it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 28/3295 [00:18<36:54,  1.48it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 29/3295 [00:19<35:40,  1.53it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 29/3295 [00:19<35:40,  1.53it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 30/3295 [00:19<34:32,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 30/3295 [00:19<34:32,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 31/3295 [00:20<36:20,  1.50it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 31/3295 [00:20<36:20,  1.50it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 32/3295 [00:20<35:14,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 32/3295 [00:20<35:14,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 33/3295 [00:20<34:13,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 33/3295 [00:20<34:13,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 34/3295 [00:22<36:01,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 34/3295 [00:22<36:01,  1.51it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 35/3295 [00:22<35:01,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 35/3295 [00:22<35:01,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 36/3295 [00:22<34:04,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 36/3295 [00:22<34:04,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 37/3295 [00:24<35:37,  1.52it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 37/3295 [00:24<35:37,  1.52it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 38/3295 [00:24<34:43,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 38/3295 [00:24<34:43,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 39/3295 [00:24<33:51,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 39/3295 [00:24<33:51,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 40/3295 [00:26<35:25,  1.53it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 40/3295 [00:26<35:25,  1.53it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 41/3295 [00:26<34:35,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|          | 41/3295 [00:26<34:35,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 42/3295 [00:26<33:47,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 42/3295 [00:26<33:47,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 43/3295 [00:27<35:15,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 43/3295 [00:27<35:15,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 44/3295 [00:28<34:29,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 44/3295 [00:28<34:29,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 45/3295 [00:28<33:45,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 45/3295 [00:28<33:45,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 46/3295 [00:29<35:12,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 46/3295 [00:29<35:12,  1.54it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 47/3295 [00:29<34:29,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 47/3295 [00:29<34:29,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 48/3295 [00:30<33:51,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 48/3295 [00:30<33:51,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 49/3295 [00:31<34:59,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   1%|▏         | 49/3295 [00:31<34:59,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 50/3295 [00:31<34:23,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 50/3295 [00:31<34:23,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 51/3295 [00:31<33:49,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 51/3295 [00:31<33:49,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 52/3295 [00:33<34:55,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 52/3295 [00:33<34:55,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 53/3295 [00:33<34:17,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 53/3295 [00:33<34:17,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 54/3295 [00:33<33:47,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 54/3295 [00:33<33:47,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 55/3295 [00:35<34:43,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 55/3295 [00:35<34:43,  1.55it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 56/3295 [00:35<34:10,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 56/3295 [00:35<34:10,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 57/3295 [00:35<33:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 57/3295 [00:35<33:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 58/3295 [00:37<34:34,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 58/3295 [00:37<34:34,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 59/3295 [00:37<34:00,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 59/3295 [00:37<34:00,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 60/3295 [00:37<33:37,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 60/3295 [00:37<33:37,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 61/3295 [00:39<34:31,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 61/3295 [00:39<34:31,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 62/3295 [00:39<34:01,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 62/3295 [00:39<34:01,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 63/3295 [00:39<33:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 63/3295 [00:39<33:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 64/3295 [00:41<34:29,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 64/3295 [00:41<34:29,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 65/3295 [00:41<33:58,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 65/3295 [00:41<33:58,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 66/3295 [00:41<33:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 66/3295 [00:41<33:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 67/3295 [00:42<34:25,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 67/3295 [00:42<34:25,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 68/3295 [00:42<33:58,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 68/3295 [00:42<33:58,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 69/3295 [00:42<33:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 69/3295 [00:42<33:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 70/3295 [00:44<34:30,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 70/3295 [00:44<34:30,  1.56it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 71/3295 [00:44<34:02,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 71/3295 [00:44<34:02,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 72/3295 [00:44<33:34,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 72/3295 [00:45<33:34,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 73/3295 [00:46<34:18,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 73/3295 [00:46<34:18,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 74/3295 [00:46<33:55,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 74/3295 [00:46<33:55,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 75/3295 [00:46<33:29,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 75/3295 [00:46<33:29,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 76/3295 [00:48<34:09,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 76/3295 [00:48<34:09,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 77/3295 [00:48<33:48,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 77/3295 [00:48<33:48,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 78/3295 [00:48<33:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 78/3295 [00:48<33:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 79/3295 [00:50<34:07,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 79/3295 [00:50<34:07,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 80/3295 [00:50<33:46,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 80/3295 [00:50<33:46,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 81/3295 [00:50<33:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 81/3295 [00:50<33:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 82/3295 [00:52<34:00,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   2%|▏         | 82/3295 [00:52<34:00,  1.57it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 83/3295 [00:52<33:38,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 83/3295 [00:52<33:38,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 84/3295 [00:52<33:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 84/3295 [00:52<33:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 85/3295 [00:53<33:55,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 85/3295 [00:53<33:55,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 86/3295 [00:54<33:36,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 86/3295 [00:54<33:36,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 87/3295 [00:54<33:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 87/3295 [00:54<33:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 88/3295 [00:55<33:51,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 88/3295 [00:55<33:51,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 89/3295 [00:56<33:37,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 89/3295 [00:56<33:37,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 90/3295 [00:56<33:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 90/3295 [00:56<33:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 91/3295 [00:57<33:50,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 91/3295 [00:57<33:50,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 92/3295 [00:57<33:36,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 92/3295 [00:57<33:36,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 93/3295 [00:57<33:15,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 93/3295 [00:57<33:15,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 94/3295 [00:59<33:45,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 94/3295 [00:59<33:45,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 95/3295 [00:59<33:36,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 95/3295 [00:59<33:36,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 96/3295 [00:59<33:16,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 96/3295 [00:59<33:16,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 97/3295 [01:01<33:41,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 97/3295 [01:01<33:41,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 98/3295 [01:01<33:32,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 98/3295 [01:01<33:32,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 99/3295 [01:01<33:12,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 99/3295 [01:01<33:12,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 100/3295 [01:03<33:39,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 100/3295 [01:03<33:39,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 101/3295 [01:03<33:30,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 101/3295 [01:03<33:30,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 102/3295 [01:03<33:10,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 102/3295 [01:03<33:10,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 103/3295 [01:05<33:36,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 103/3295 [01:05<33:36,  1.58it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 104/3295 [01:05<33:25,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 104/3295 [01:05<33:25,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 105/3295 [01:05<33:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 105/3295 [01:05<33:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 106/3295 [01:06<33:27,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 106/3295 [01:06<33:27,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 107/3295 [01:07<33:25,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 107/3295 [01:07<33:25,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 108/3295 [01:07<33:07,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 108/3295 [01:07<33:07,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 109/3295 [01:08<33:25,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 109/3295 [01:08<33:25,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 110/3295 [01:09<33:23,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 110/3295 [01:09<33:23,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 111/3295 [01:09<33:05,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 111/3295 [01:09<33:05,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 112/3295 [01:10<33:20,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 112/3295 [01:10<33:20,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 113/3295 [01:11<33:20,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 113/3295 [01:11<33:20,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 114/3295 [01:11<33:03,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 114/3295 [01:11<33:03,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 115/3295 [01:12<33:19,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   3%|▎         | 115/3295 [01:12<33:19,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 116/3295 [01:12<33:14,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 116/3295 [01:12<33:14,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 117/3295 [01:12<32:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 117/3295 [01:12<32:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 118/3295 [01:14<33:21,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 118/3295 [01:14<33:21,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 119/3295 [01:14<33:11,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 119/3295 [01:14<33:11,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 120/3295 [01:14<32:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 120/3295 [01:14<32:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 121/3295 [01:16<33:18,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 121/3295 [01:16<33:18,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 122/3295 [01:16<33:13,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 122/3295 [01:16<33:13,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 123/3295 [01:16<32:57,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▎         | 123/3295 [01:16<32:57,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 124/3295 [01:18<33:16,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 124/3295 [01:18<33:16,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 125/3295 [01:18<33:11,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 125/3295 [01:18<33:11,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 126/3295 [01:18<32:55,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 126/3295 [01:18<32:55,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 127/3295 [01:19<33:13,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 127/3295 [01:19<33:13,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 128/3295 [01:20<33:09,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 128/3295 [01:20<33:09,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 129/3295 [01:20<32:53,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 129/3295 [01:20<32:53,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 130/3295 [01:21<33:08,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 130/3295 [01:21<33:08,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 131/3295 [01:22<33:08,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 131/3295 [01:22<33:08,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 132/3295 [01:22<32:53,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 132/3295 [01:22<32:53,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 133/3295 [01:23<33:01,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 133/3295 [01:23<33:01,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 134/3295 [01:24<33:05,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 134/3295 [01:24<33:05,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 135/3295 [01:24<32:50,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 135/3295 [01:24<32:50,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 136/3295 [01:25<32:57,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 136/3295 [01:25<32:57,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 137/3295 [01:26<33:04,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 137/3295 [01:26<33:04,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 138/3295 [01:26<32:49,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 138/3295 [01:26<32:49,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 139/3295 [01:27<33:01,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 139/3295 [01:27<33:01,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 140/3295 [01:27<33:01,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 140/3295 [01:27<33:01,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 141/3295 [01:27<32:47,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 141/3295 [01:27<32:47,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 142/3295 [01:29<33:02,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 142/3295 [01:29<33:02,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 143/3295 [01:29<32:57,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 143/3295 [01:29<32:57,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 144/3295 [01:29<32:43,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 144/3295 [01:29<32:43,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 145/3295 [01:31<33:01,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 145/3295 [01:31<33:01,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 146/3295 [01:31<32:54,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 146/3295 [01:31<32:54,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 147/3295 [01:31<32:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 147/3295 [01:31<32:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 148/3295 [01:33<32:58,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   4%|▍         | 148/3295 [01:33<32:58,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 149/3295 [01:33<32:49,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 149/3295 [01:33<32:49,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 150/3295 [01:33<32:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 150/3295 [01:33<32:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 151/3295 [01:34<32:55,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 151/3295 [01:34<32:55,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 152/3295 [01:35<32:48,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 152/3295 [01:35<32:48,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 153/3295 [01:35<32:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 153/3295 [01:35<32:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 154/3295 [01:36<32:51,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 154/3295 [01:36<32:51,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 155/3295 [01:37<32:48,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 155/3295 [01:37<32:48,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 156/3295 [01:37<32:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 156/3295 [01:37<32:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 157/3295 [01:38<32:47,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 157/3295 [01:38<32:47,  1.59it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 158/3295 [01:39<32:45,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 158/3295 [01:39<32:45,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 159/3295 [01:39<32:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 159/3295 [01:39<32:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 160/3295 [01:40<32:43,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 160/3295 [01:40<32:43,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 161/3295 [01:40<32:41,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 161/3295 [01:40<32:41,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 162/3295 [01:40<32:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 162/3295 [01:40<32:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 163/3295 [01:41<32:39,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 163/3295 [01:41<32:39,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 164/3295 [01:42<32:34,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▍         | 164/3295 [01:42<32:34,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 165/3295 [01:42<32:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 165/3295 [01:42<32:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 166/3295 [01:43<32:36,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 166/3295 [01:43<32:36,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 167/3295 [01:44<32:30,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 167/3295 [01:44<32:30,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 168/3295 [01:44<32:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 168/3295 [01:44<32:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 169/3295 [01:45<32:35,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 169/3295 [01:45<32:35,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 170/3295 [01:45<32:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 170/3295 [01:45<32:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 171/3295 [01:45<32:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 171/3295 [01:45<32:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 172/3295 [01:47<32:32,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 172/3295 [01:47<32:32,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 173/3295 [01:47<32:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 173/3295 [01:47<32:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 174/3295 [01:47<32:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 174/3295 [01:47<32:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 175/3295 [01:49<32:29,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 175/3295 [01:49<32:29,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 176/3295 [01:49<32:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 176/3295 [01:49<32:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 177/3295 [01:49<32:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 177/3295 [01:49<32:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 178/3295 [01:51<32:28,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 178/3295 [01:51<32:28,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 179/3295 [01:51<32:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 179/3295 [01:51<32:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 180/3295 [01:51<32:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 180/3295 [01:51<32:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 181/3295 [01:53<32:25,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   5%|▌         | 181/3295 [01:53<32:25,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 182/3295 [01:53<32:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 182/3295 [01:53<32:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 183/3295 [01:53<32:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 183/3295 [01:53<32:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 184/3295 [01:55<32:24,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 184/3295 [01:55<32:24,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 185/3295 [01:55<32:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 185/3295 [01:55<32:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 186/3295 [01:55<32:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 186/3295 [01:55<32:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 187/3295 [01:56<32:20,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 187/3295 [01:56<32:20,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 188/3295 [01:56<32:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 188/3295 [01:56<32:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 189/3295 [01:56<32:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 189/3295 [01:56<32:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 190/3295 [01:58<32:17,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 190/3295 [01:58<32:17,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 191/3295 [01:58<32:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 191/3295 [01:58<32:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 192/3295 [01:58<32:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 192/3295 [01:58<32:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 193/3295 [02:00<32:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 193/3295 [02:00<32:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 194/3295 [02:00<32:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 194/3295 [02:00<32:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 195/3295 [02:00<31:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 195/3295 [02:00<31:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 196/3295 [02:02<32:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 196/3295 [02:02<32:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 197/3295 [02:02<32:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 197/3295 [02:02<32:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 198/3295 [02:02<31:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 198/3295 [02:02<31:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 199/3295 [02:03<32:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 199/3295 [02:03<32:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 200/3295 [02:04<32:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 200/3295 [02:04<32:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 201/3295 [02:04<31:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 201/3295 [02:04<31:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 202/3295 [02:05<32:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 202/3295 [02:05<32:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 203/3295 [02:06<31:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 203/3295 [02:06<31:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 204/3295 [02:06<31:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 204/3295 [02:06<31:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 205/3295 [02:07<32:05,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▌         | 205/3295 [02:07<32:05,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 206/3295 [02:07<31:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 206/3295 [02:07<31:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 207/3295 [02:08<31:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 207/3295 [02:08<31:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 208/3295 [02:09<32:03,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 208/3295 [02:09<32:03,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 209/3295 [02:09<31:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 209/3295 [02:09<31:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 210/3295 [02:09<31:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 210/3295 [02:09<31:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 211/3295 [02:11<32:01,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 211/3295 [02:11<32:01,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 212/3295 [02:11<31:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 212/3295 [02:11<31:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 213/3295 [02:11<31:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 213/3295 [02:11<31:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 214/3295 [02:13<32:00,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   6%|▋         | 214/3295 [02:13<32:00,  1.60it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 215/3295 [02:13<31:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 215/3295 [02:13<31:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 216/3295 [02:13<31:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 216/3295 [02:13<31:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 217/3295 [02:15<31:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 217/3295 [02:15<31:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 218/3295 [02:15<31:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 218/3295 [02:15<31:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 219/3295 [02:15<31:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 219/3295 [02:15<31:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 220/3295 [02:17<31:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 220/3295 [02:17<31:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 221/3295 [02:17<31:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 221/3295 [02:17<31:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 222/3295 [02:17<31:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 222/3295 [02:17<31:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 223/3295 [02:18<31:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 223/3295 [02:18<31:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 224/3295 [02:18<31:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 224/3295 [02:18<31:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 225/3295 [02:19<31:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 225/3295 [02:19<31:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 226/3295 [02:20<31:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 226/3295 [02:20<31:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 227/3295 [02:20<31:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 227/3295 [02:20<31:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 228/3295 [02:21<31:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 228/3295 [02:21<31:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 229/3295 [02:22<31:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 229/3295 [02:22<31:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 230/3295 [02:22<31:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 230/3295 [02:22<31:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 231/3295 [02:23<31:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 231/3295 [02:23<31:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 232/3295 [02:24<31:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 232/3295 [02:24<31:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 233/3295 [02:24<31:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 233/3295 [02:24<31:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 234/3295 [02:25<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 234/3295 [02:25<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 235/3295 [02:26<31:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 235/3295 [02:26<31:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 236/3295 [02:26<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 236/3295 [02:26<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 237/3295 [02:27<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 237/3295 [02:27<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 238/3295 [02:28<31:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 238/3295 [02:28<31:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 239/3295 [02:28<31:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 239/3295 [02:28<31:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 240/3295 [02:28<31:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 240/3295 [02:28<31:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 241/3295 [02:30<31:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 241/3295 [02:30<31:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 242/3295 [02:30<31:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 242/3295 [02:30<31:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 243/3295 [02:30<31:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 243/3295 [02:30<31:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 244/3295 [02:31<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 244/3295 [02:31<31:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 245/3295 [02:32<31:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 245/3295 [02:32<31:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 246/3295 [02:32<31:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 246/3295 [02:32<31:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 247/3295 [02:33<31:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   7%|▋         | 247/3295 [02:33<31:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 248/3295 [02:34<31:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 248/3295 [02:34<31:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 249/3295 [02:34<31:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 249/3295 [02:34<31:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 250/3295 [02:35<31:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 250/3295 [02:35<31:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 251/3295 [02:36<31:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 251/3295 [02:36<31:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 252/3295 [02:36<31:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 252/3295 [02:36<31:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 253/3295 [02:37<31:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 253/3295 [02:37<31:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 254/3295 [02:37<31:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 254/3295 [02:37<31:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 255/3295 [02:38<31:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 255/3295 [02:38<31:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 256/3295 [02:39<31:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 256/3295 [02:39<31:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 257/3295 [02:39<31:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 257/3295 [02:39<31:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 258/3295 [02:39<31:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 258/3295 [02:39<31:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 259/3295 [02:40<31:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 259/3295 [02:40<31:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 260/3295 [02:41<31:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 260/3295 [02:41<31:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 261/3295 [02:41<31:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 261/3295 [02:41<31:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 262/3295 [02:42<31:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 262/3295 [02:42<31:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 263/3295 [02:43<31:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 263/3295 [02:43<31:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 264/3295 [02:43<31:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 264/3295 [02:43<31:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 265/3295 [02:44<31:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 265/3295 [02:44<31:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 266/3295 [02:45<31:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 266/3295 [02:45<31:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 267/3295 [02:45<31:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 267/3295 [02:45<31:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 268/3295 [02:46<31:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 268/3295 [02:46<31:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 269/3295 [02:47<31:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 269/3295 [02:47<31:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 270/3295 [02:47<31:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 270/3295 [02:47<31:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 271/3295 [02:48<31:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 271/3295 [02:48<31:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 272/3295 [02:49<31:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 272/3295 [02:49<31:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 273/3295 [02:49<31:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 273/3295 [02:49<31:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 274/3295 [02:50<31:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 274/3295 [02:50<31:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 275/3295 [02:51<31:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 275/3295 [02:51<31:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 276/3295 [02:51<31:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 276/3295 [02:51<31:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 277/3295 [02:51<31:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 277/3295 [02:51<31:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 278/3295 [02:53<31:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 278/3295 [02:53<31:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 279/3295 [02:53<31:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 279/3295 [02:53<31:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 280/3295 [02:53<31:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   8%|▊         | 280/3295 [02:53<31:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 281/3295 [02:55<31:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 281/3295 [02:55<31:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 282/3295 [02:55<31:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 282/3295 [02:55<31:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 283/3295 [02:55<31:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 283/3295 [02:55<31:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 284/3295 [02:56<31:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 284/3295 [02:56<31:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 285/3295 [02:56<31:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 285/3295 [02:56<31:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 286/3295 [02:57<31:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 286/3295 [02:57<31:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 287/3295 [02:58<31:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 287/3295 [02:58<31:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 288/3295 [02:58<31:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▊         | 288/3295 [02:58<31:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 289/3295 [02:59<31:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 289/3295 [02:59<31:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 290/3295 [03:00<31:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 290/3295 [03:00<31:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 291/3295 [03:00<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 291/3295 [03:00<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 292/3295 [03:01<31:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 292/3295 [03:01<31:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 293/3295 [03:02<31:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 293/3295 [03:02<31:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 294/3295 [03:02<30:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 294/3295 [03:02<30:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 295/3295 [03:03<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 295/3295 [03:03<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 296/3295 [03:03<31:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 296/3295 [03:03<31:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 297/3295 [03:04<30:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 297/3295 [03:04<30:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 298/3295 [03:05<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 298/3295 [03:05<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 299/3295 [03:05<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 299/3295 [03:05<31:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 300/3295 [03:05<30:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 300/3295 [03:05<30:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 301/3295 [03:06<30:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 301/3295 [03:06<30:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 302/3295 [03:07<31:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 302/3295 [03:07<31:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 303/3295 [03:07<30:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 303/3295 [03:07<30:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 304/3295 [03:09<30:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 304/3295 [03:09<30:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 305/3295 [03:09<30:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 305/3295 [03:09<30:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 306/3295 [03:09<30:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 306/3295 [03:09<30:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 307/3295 [03:10<30:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 307/3295 [03:10<30:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 308/3295 [03:11<30:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 308/3295 [03:11<30:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 309/3295 [03:11<30:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 309/3295 [03:11<30:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 310/3295 [03:12<30:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 310/3295 [03:12<30:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 311/3295 [03:13<30:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 311/3295 [03:13<30:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 312/3295 [03:13<30:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 312/3295 [03:13<30:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 313/3295 [03:14<30:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:   9%|▉         | 313/3295 [03:14<30:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 314/3295 [03:15<30:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 314/3295 [03:15<30:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 315/3295 [03:15<30:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 315/3295 [03:15<30:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 316/3295 [03:16<30:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 316/3295 [03:16<30:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 317/3295 [03:16<30:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 317/3295 [03:16<30:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 318/3295 [03:17<30:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 318/3295 [03:17<30:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 319/3295 [03:18<30:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 319/3295 [03:18<30:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 320/3295 [03:18<30:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 320/3295 [03:18<30:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 321/3295 [03:18<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 321/3295 [03:18<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 322/3295 [03:19<30:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 322/3295 [03:19<30:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 323/3295 [03:20<30:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 323/3295 [03:20<30:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 324/3295 [03:20<30:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 324/3295 [03:20<30:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 325/3295 [03:21<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 325/3295 [03:21<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 326/3295 [03:22<30:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 326/3295 [03:22<30:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 327/3295 [03:22<30:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 327/3295 [03:22<30:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 328/3295 [03:23<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 328/3295 [03:23<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 329/3295 [03:24<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|▉         | 329/3295 [03:24<30:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 330/3295 [03:24<30:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 330/3295 [03:24<30:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 331/3295 [03:25<30:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 331/3295 [03:25<30:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 332/3295 [03:26<30:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 332/3295 [03:26<30:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 333/3295 [03:26<30:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 333/3295 [03:26<30:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 334/3295 [03:27<30:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 334/3295 [03:27<30:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 335/3295 [03:28<30:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 335/3295 [03:28<30:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 336/3295 [03:28<30:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 336/3295 [03:28<30:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 337/3295 [03:29<30:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 337/3295 [03:29<30:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 338/3295 [03:29<30:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 338/3295 [03:29<30:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 339/3295 [03:30<30:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 339/3295 [03:30<30:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 340/3295 [03:30<30:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 340/3295 [03:30<30:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 341/3295 [03:31<30:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 341/3295 [03:31<30:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 342/3295 [03:32<30:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 342/3295 [03:32<30:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 343/3295 [03:32<30:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 343/3295 [03:32<30:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 344/3295 [03:33<30:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 344/3295 [03:33<30:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 345/3295 [03:33<30:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  10%|█         | 345/3295 [03:33<30:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 346/3295 [03:34<30:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 346/3295 [03:34<30:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 347/3295 [03:35<30:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 347/3295 [03:35<30:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 348/3295 [03:35<30:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 348/3295 [03:35<30:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 349/3295 [03:36<30:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 349/3295 [03:36<30:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 350/3295 [03:37<30:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 350/3295 [03:37<30:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 351/3295 [03:37<30:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 351/3295 [03:37<30:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 352/3295 [03:38<30:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 352/3295 [03:38<30:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 353/3295 [03:38<30:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 353/3295 [03:38<30:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 354/3295 [03:39<30:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 354/3295 [03:39<30:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 355/3295 [03:40<30:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 355/3295 [03:40<30:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 356/3295 [03:40<30:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 356/3295 [03:40<30:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 357/3295 [03:41<30:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 357/3295 [03:41<30:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 358/3295 [03:42<30:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 358/3295 [03:42<30:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 359/3295 [03:42<30:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 359/3295 [03:42<30:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 360/3295 [03:43<30:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 360/3295 [03:43<30:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 361/3295 [03:44<30:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 361/3295 [03:44<30:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 362/3295 [03:44<30:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 362/3295 [03:44<30:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 363/3295 [03:45<30:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 363/3295 [03:45<30:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 364/3295 [03:45<30:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 364/3295 [03:45<30:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 365/3295 [03:46<30:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 365/3295 [03:46<30:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 366/3295 [03:47<30:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 366/3295 [03:47<30:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 367/3295 [03:47<30:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 367/3295 [03:47<30:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 368/3295 [03:48<30:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 368/3295 [03:48<30:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 369/3295 [03:49<30:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 369/3295 [03:49<30:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 370/3295 [03:49<30:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█         | 370/3295 [03:49<30:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 371/3295 [03:50<30:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 371/3295 [03:50<30:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 372/3295 [03:50<30:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 372/3295 [03:50<30:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 373/3295 [03:51<30:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 373/3295 [03:51<30:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 374/3295 [03:52<30:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 374/3295 [03:52<30:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 375/3295 [03:52<30:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 375/3295 [03:52<30:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 376/3295 [03:53<30:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 376/3295 [03:53<30:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 377/3295 [03:53<30:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 377/3295 [03:53<30:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 378/3295 [03:54<30:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  11%|█▏        | 378/3295 [03:54<30:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 379/3295 [03:55<30:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 379/3295 [03:55<30:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 380/3295 [03:55<30:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 380/3295 [03:55<30:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 381/3295 [03:56<30:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 381/3295 [03:56<30:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 382/3295 [03:56<30:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 382/3295 [03:56<30:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 383/3295 [03:57<30:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 383/3295 [03:57<30:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 384/3295 [03:58<30:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 384/3295 [03:58<30:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 385/3295 [03:58<30:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 385/3295 [03:58<30:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 386/3295 [03:59<30:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 386/3295 [03:59<30:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 387/3295 [04:00<30:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 387/3295 [04:00<30:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 388/3295 [04:00<30:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 388/3295 [04:00<30:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 389/3295 [04:01<30:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 389/3295 [04:01<30:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 390/3295 [04:01<30:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 390/3295 [04:01<30:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 391/3295 [04:02<30:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 391/3295 [04:02<30:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 392/3295 [04:03<30:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 392/3295 [04:03<30:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 393/3295 [04:03<29:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 393/3295 [04:03<29:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 394/3295 [04:04<29:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 394/3295 [04:04<29:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 395/3295 [04:05<29:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 395/3295 [04:05<29:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 396/3295 [04:05<29:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 396/3295 [04:05<29:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 397/3295 [04:06<29:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 397/3295 [04:06<29:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 398/3295 [04:06<29:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 398/3295 [04:06<29:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 399/3295 [04:07<29:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 399/3295 [04:07<29:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 400/3295 [04:08<29:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 400/3295 [04:08<29:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 401/3295 [04:08<29:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 401/3295 [04:08<29:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 402/3295 [04:09<29:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 402/3295 [04:09<29:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 403/3295 [04:10<29:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 403/3295 [04:10<29:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 404/3295 [04:10<29:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 404/3295 [04:10<29:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 405/3295 [04:10<29:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 405/3295 [04:10<29:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 406/3295 [04:11<29:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 406/3295 [04:11<29:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 407/3295 [04:12<29:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 407/3295 [04:12<29:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 408/3295 [04:12<29:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 408/3295 [04:12<29:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 409/3295 [04:13<29:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 409/3295 [04:13<29:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 410/3295 [04:14<29:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 410/3295 [04:14<29:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 411/3295 [04:14<29:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  12%|█▏        | 411/3295 [04:14<29:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 412/3295 [04:15<29:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 412/3295 [04:15<29:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 413/3295 [04:16<29:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 413/3295 [04:16<29:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 414/3295 [04:16<29:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 414/3295 [04:16<29:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 415/3295 [04:17<29:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 415/3295 [04:17<29:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 416/3295 [04:17<29:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 416/3295 [04:17<29:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 417/3295 [04:18<29:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 417/3295 [04:18<29:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 418/3295 [04:19<29:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 418/3295 [04:19<29:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 419/3295 [04:19<29:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 419/3295 [04:19<29:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 420/3295 [04:20<29:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 420/3295 [04:20<29:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 421/3295 [04:20<29:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 421/3295 [04:20<29:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 422/3295 [04:21<29:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 422/3295 [04:21<29:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 423/3295 [04:22<29:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 423/3295 [04:22<29:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 424/3295 [04:22<29:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 424/3295 [04:22<29:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 425/3295 [04:23<29:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 425/3295 [04:23<29:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 426/3295 [04:24<29:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 426/3295 [04:24<29:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 427/3295 [04:24<29:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 427/3295 [04:24<29:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 428/3295 [04:25<29:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 428/3295 [04:25<29:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 429/3295 [04:26<29:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 429/3295 [04:26<29:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 430/3295 [04:26<29:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 430/3295 [04:26<29:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 431/3295 [04:27<29:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 431/3295 [04:27<29:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 432/3295 [04:27<29:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 432/3295 [04:27<29:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 433/3295 [04:28<29:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 433/3295 [04:28<29:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 434/3295 [04:29<29:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 434/3295 [04:29<29:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 435/3295 [04:29<29:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 435/3295 [04:29<29:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 436/3295 [04:29<29:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 436/3295 [04:29<29:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 437/3295 [04:31<29:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 437/3295 [04:31<29:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 438/3295 [04:31<29:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 438/3295 [04:31<29:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 439/3295 [04:31<29:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 439/3295 [04:31<29:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 440/3295 [04:32<29:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 440/3295 [04:32<29:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 441/3295 [04:33<29:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 441/3295 [04:33<29:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 442/3295 [04:33<29:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 442/3295 [04:33<29:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 443/3295 [04:34<29:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 443/3295 [04:34<29:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 444/3295 [04:35<29:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  13%|█▎        | 444/3295 [04:35<29:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 445/3295 [04:35<29:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 445/3295 [04:35<29:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 446/3295 [04:36<29:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 446/3295 [04:36<29:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 447/3295 [04:37<29:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 447/3295 [04:37<29:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 448/3295 [04:37<29:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 448/3295 [04:37<29:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 449/3295 [04:38<29:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 449/3295 [04:38<29:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 450/3295 [04:38<29:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 450/3295 [04:38<29:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 451/3295 [04:38<29:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 451/3295 [04:38<29:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 452/3295 [04:40<29:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 452/3295 [04:40<29:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 453/3295 [04:40<29:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▎        | 453/3295 [04:40<29:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 454/3295 [04:40<29:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 454/3295 [04:40<29:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 455/3295 [04:42<29:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 455/3295 [04:42<29:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 456/3295 [04:42<29:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 456/3295 [04:42<29:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 457/3295 [04:42<29:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 457/3295 [04:42<29:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 458/3295 [04:43<29:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 458/3295 [04:43<29:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 459/3295 [04:44<29:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 459/3295 [04:44<29:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 460/3295 [04:44<29:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 460/3295 [04:44<29:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 461/3295 [04:45<29:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 461/3295 [04:45<29:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 462/3295 [04:46<29:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 462/3295 [04:46<29:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 463/3295 [04:46<29:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 463/3295 [04:46<29:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 464/3295 [04:47<29:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 464/3295 [04:47<29:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 465/3295 [04:47<29:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 465/3295 [04:47<29:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 466/3295 [04:48<29:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 466/3295 [04:48<29:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 467/3295 [04:49<29:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 467/3295 [04:49<29:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 468/3295 [04:49<29:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 468/3295 [04:49<29:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 469/3295 [04:49<29:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 469/3295 [04:49<29:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 470/3295 [04:51<29:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 470/3295 [04:51<29:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 471/3295 [04:51<29:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 471/3295 [04:51<29:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 472/3295 [04:51<29:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 472/3295 [04:51<29:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 473/3295 [04:53<29:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 473/3295 [04:53<29:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 474/3295 [04:53<29:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 474/3295 [04:53<29:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 475/3295 [04:53<29:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 475/3295 [04:53<29:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 476/3295 [04:55<29:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 476/3295 [04:55<29:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 477/3295 [04:55<29:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  14%|█▍        | 477/3295 [04:55<29:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 478/3295 [04:55<29:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 478/3295 [04:55<29:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 479/3295 [04:56<29:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 479/3295 [04:56<29:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 480/3295 [04:57<29:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 480/3295 [04:57<29:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 481/3295 [04:57<29:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 481/3295 [04:57<29:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 482/3295 [04:58<29:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 482/3295 [04:58<29:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 483/3295 [04:59<29:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 483/3295 [04:59<29:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 484/3295 [04:59<28:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 484/3295 [04:59<28:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 485/3295 [05:00<29:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 485/3295 [05:00<29:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 486/3295 [05:00<28:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 486/3295 [05:00<28:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 487/3295 [05:01<28:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 487/3295 [05:01<28:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 488/3295 [05:02<29:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 488/3295 [05:02<29:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 489/3295 [05:03<28:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 489/3295 [05:03<28:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 490/3295 [05:03<28:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 490/3295 [05:03<28:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 491/3295 [05:04<28:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 491/3295 [05:04<28:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 492/3295 [05:04<28:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 492/3295 [05:04<28:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 493/3295 [05:04<28:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 493/3295 [05:04<28:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 494/3295 [05:06<28:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▍        | 494/3295 [05:06<28:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 495/3295 [05:06<28:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 495/3295 [05:06<28:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 496/3295 [05:06<28:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 496/3295 [05:06<28:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 497/3295 [05:08<28:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 497/3295 [05:08<28:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 498/3295 [05:08<28:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 498/3295 [05:08<28:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 499/3295 [05:08<28:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 499/3295 [05:08<28:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 500/3295 [05:10<28:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 500/3295 [05:10<28:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 501/3295 [05:10<28:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 501/3295 [05:10<28:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 502/3295 [05:10<28:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 502/3295 [05:10<28:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 503/3295 [05:11<28:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 503/3295 [05:11<28:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 504/3295 [05:12<28:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 504/3295 [05:12<28:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 505/3295 [05:12<28:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 505/3295 [05:12<28:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 506/3295 [05:13<28:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 506/3295 [05:13<28:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 507/3295 [05:13<28:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 507/3295 [05:13<28:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 508/3295 [05:14<28:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 508/3295 [05:14<28:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 509/3295 [05:15<28:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 509/3295 [05:15<28:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 510/3295 [05:15<28:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  15%|█▌        | 510/3295 [05:15<28:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 511/3295 [05:16<28:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 511/3295 [05:16<28:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 512/3295 [05:17<28:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 512/3295 [05:17<28:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 513/3295 [05:17<28:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 513/3295 [05:17<28:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 514/3295 [05:18<28:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 514/3295 [05:18<28:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 515/3295 [05:19<28:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 515/3295 [05:19<28:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 516/3295 [05:19<28:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 516/3295 [05:19<28:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 517/3295 [05:20<28:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 517/3295 [05:20<28:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 518/3295 [05:21<28:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 518/3295 [05:21<28:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 519/3295 [05:21<28:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 519/3295 [05:21<28:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 520/3295 [05:21<28:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 520/3295 [05:21<28:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 521/3295 [05:23<28:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 521/3295 [05:23<28:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 522/3295 [05:23<28:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 522/3295 [05:23<28:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 523/3295 [05:23<28:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 523/3295 [05:23<28:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 524/3295 [05:24<28:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 524/3295 [05:24<28:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 525/3295 [05:25<28:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 525/3295 [05:25<28:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 526/3295 [05:25<28:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 526/3295 [05:25<28:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 527/3295 [05:26<28:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 527/3295 [05:26<28:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 528/3295 [05:26<28:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 528/3295 [05:26<28:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 529/3295 [05:27<28:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 529/3295 [05:27<28:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 530/3295 [05:28<28:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 530/3295 [05:28<28:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 531/3295 [05:28<28:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 531/3295 [05:28<28:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 532/3295 [05:29<28:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 532/3295 [05:29<28:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 533/3295 [05:30<28:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 533/3295 [05:30<28:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 534/3295 [05:30<28:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 534/3295 [05:30<28:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 535/3295 [05:31<28:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▌        | 535/3295 [05:31<28:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 536/3295 [05:31<28:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 536/3295 [05:31<28:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 537/3295 [05:32<28:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 537/3295 [05:32<28:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 538/3295 [05:33<28:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 538/3295 [05:33<28:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 539/3295 [05:34<28:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 539/3295 [05:34<28:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 540/3295 [05:34<28:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 540/3295 [05:34<28:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 541/3295 [05:34<28:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 541/3295 [05:34<28:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 542/3295 [05:35<28:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 542/3295 [05:35<28:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 543/3295 [05:36<28:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  16%|█▋        | 543/3295 [05:36<28:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 544/3295 [05:36<28:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 544/3295 [05:36<28:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 545/3295 [05:37<28:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 545/3295 [05:37<28:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 546/3295 [05:37<28:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 546/3295 [05:37<28:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 547/3295 [05:38<28:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 547/3295 [05:38<28:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 548/3295 [05:39<28:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 548/3295 [05:39<28:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 549/3295 [05:39<28:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 549/3295 [05:39<28:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 550/3295 [05:40<28:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 550/3295 [05:40<28:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 551/3295 [05:41<28:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 551/3295 [05:41<28:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 552/3295 [05:41<28:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 552/3295 [05:41<28:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 553/3295 [05:42<28:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 553/3295 [05:42<28:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 554/3295 [05:43<28:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 554/3295 [05:43<28:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 555/3295 [05:43<28:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 555/3295 [05:43<28:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 556/3295 [05:44<28:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 556/3295 [05:44<28:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 557/3295 [05:45<28:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 557/3295 [05:45<28:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 558/3295 [05:45<28:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 558/3295 [05:45<28:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 559/3295 [05:45<28:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 559/3295 [05:45<28:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 560/3295 [05:47<28:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 560/3295 [05:47<28:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 561/3295 [05:47<28:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 561/3295 [05:47<28:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 562/3295 [05:47<28:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 562/3295 [05:47<28:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 563/3295 [05:48<28:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 563/3295 [05:48<28:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 564/3295 [05:49<28:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 564/3295 [05:49<28:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 565/3295 [05:49<28:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 565/3295 [05:49<28:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 566/3295 [05:50<28:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 566/3295 [05:50<28:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 567/3295 [05:51<28:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 567/3295 [05:51<28:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 568/3295 [05:51<28:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 568/3295 [05:51<28:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 569/3295 [05:52<28:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 569/3295 [05:52<28:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 570/3295 [05:52<28:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 570/3295 [05:52<28:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 571/3295 [05:53<28:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 571/3295 [05:53<28:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 572/3295 [05:54<28:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 572/3295 [05:54<28:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 573/3295 [05:54<28:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 573/3295 [05:54<28:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 574/3295 [05:54<28:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 574/3295 [05:54<28:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 575/3295 [05:56<28:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 575/3295 [05:56<28:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 576/3295 [05:56<28:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  17%|█▋        | 576/3295 [05:56<28:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 577/3295 [05:56<28:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 577/3295 [05:56<28:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 578/3295 [05:58<28:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 578/3295 [05:58<28:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 579/3295 [05:58<28:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 579/3295 [05:58<28:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 580/3295 [05:58<27:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 580/3295 [05:58<27:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 581/3295 [06:00<28:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 581/3295 [06:00<28:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 582/3295 [06:00<27:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 582/3295 [06:00<27:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 583/3295 [06:00<27:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 583/3295 [06:00<27:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 584/3295 [06:02<28:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 584/3295 [06:02<28:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 585/3295 [06:02<27:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 585/3295 [06:02<27:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 586/3295 [06:02<27:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 586/3295 [06:02<27:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 587/3295 [06:04<27:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 587/3295 [06:04<27:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 588/3295 [06:04<27:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 588/3295 [06:04<27:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 589/3295 [06:04<27:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 589/3295 [06:04<27:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 590/3295 [06:05<27:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 590/3295 [06:05<27:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 591/3295 [06:05<27:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 591/3295 [06:05<27:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 592/3295 [06:05<27:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 592/3295 [06:05<27:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 593/3295 [06:07<27:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 593/3295 [06:07<27:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 594/3295 [06:07<27:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 594/3295 [06:07<27:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 595/3295 [06:07<27:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 595/3295 [06:07<27:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 596/3295 [06:09<27:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 596/3295 [06:09<27:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 597/3295 [06:09<27:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 597/3295 [06:09<27:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 598/3295 [06:09<27:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 598/3295 [06:09<27:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 599/3295 [06:11<27:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 599/3295 [06:11<27:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 600/3295 [06:11<27:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 600/3295 [06:11<27:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 601/3295 [06:11<27:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 601/3295 [06:11<27:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 602/3295 [06:13<27:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 602/3295 [06:13<27:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 603/3295 [06:13<27:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 603/3295 [06:13<27:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 604/3295 [06:13<27:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 604/3295 [06:13<27:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 605/3295 [06:15<27:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 605/3295 [06:15<27:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 606/3295 [06:15<27:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 606/3295 [06:15<27:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 607/3295 [06:15<27:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 607/3295 [06:15<27:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 608/3295 [06:17<27:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 608/3295 [06:17<27:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 609/3295 [06:17<27:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  18%|█▊        | 609/3295 [06:17<27:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 610/3295 [06:17<27:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 610/3295 [06:17<27:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 611/3295 [06:19<27:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 611/3295 [06:19<27:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 612/3295 [06:19<27:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 612/3295 [06:19<27:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 613/3295 [06:19<27:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 613/3295 [06:19<27:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 614/3295 [06:20<27:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 614/3295 [06:20<27:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 615/3295 [06:20<27:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 615/3295 [06:20<27:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 616/3295 [06:20<27:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 616/3295 [06:20<27:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 617/3295 [06:22<27:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▊        | 617/3295 [06:22<27:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 618/3295 [06:22<27:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 618/3295 [06:22<27:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 619/3295 [06:22<27:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 619/3295 [06:22<27:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 620/3295 [06:24<27:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 620/3295 [06:24<27:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 621/3295 [06:24<27:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 621/3295 [06:24<27:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 622/3295 [06:24<27:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 622/3295 [06:24<27:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 623/3295 [06:26<27:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 623/3295 [06:26<27:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 624/3295 [06:26<27:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 624/3295 [06:26<27:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 625/3295 [06:26<27:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 625/3295 [06:26<27:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 626/3295 [06:28<27:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 626/3295 [06:28<27:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 627/3295 [06:28<27:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 627/3295 [06:28<27:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 628/3295 [06:28<27:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 628/3295 [06:28<27:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 629/3295 [06:29<27:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 629/3295 [06:29<27:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 630/3295 [06:29<27:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 630/3295 [06:29<27:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 631/3295 [06:30<27:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 631/3295 [06:30<27:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 632/3295 [06:31<27:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 632/3295 [06:31<27:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 633/3295 [06:31<27:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 633/3295 [06:31<27:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 634/3295 [06:31<27:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 634/3295 [06:31<27:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 635/3295 [06:33<27:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 635/3295 [06:33<27:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 636/3295 [06:33<27:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 636/3295 [06:33<27:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 637/3295 [06:33<27:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 637/3295 [06:33<27:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 638/3295 [06:35<27:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 638/3295 [06:35<27:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 639/3295 [06:35<27:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 639/3295 [06:35<27:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 640/3295 [06:35<27:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 640/3295 [06:35<27:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 641/3295 [06:37<27:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 641/3295 [06:37<27:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 642/3295 [06:37<27:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  19%|█▉        | 642/3295 [06:37<27:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 643/3295 [06:37<27:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 643/3295 [06:37<27:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 644/3295 [06:39<27:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 644/3295 [06:39<27:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 645/3295 [06:39<27:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 645/3295 [06:39<27:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 646/3295 [06:39<27:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 646/3295 [06:39<27:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 647/3295 [06:41<27:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 647/3295 [06:41<27:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 648/3295 [06:41<27:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 648/3295 [06:41<27:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 649/3295 [06:41<27:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 649/3295 [06:41<27:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 650/3295 [06:43<27:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 650/3295 [06:43<27:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 651/3295 [06:43<27:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 651/3295 [06:43<27:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 652/3295 [06:43<27:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 652/3295 [06:43<27:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 653/3295 [06:45<27:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 653/3295 [06:45<27:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 654/3295 [06:45<27:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 654/3295 [06:45<27:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 655/3295 [06:45<27:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 655/3295 [06:45<27:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 656/3295 [06:47<27:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 656/3295 [06:47<27:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 657/3295 [06:47<27:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 657/3295 [06:47<27:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 658/3295 [06:47<27:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|█▉        | 658/3295 [06:47<27:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 659/3295 [06:48<27:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 659/3295 [06:48<27:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 660/3295 [06:48<27:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 660/3295 [06:48<27:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 661/3295 [06:49<27:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 661/3295 [06:49<27:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 662/3295 [06:50<27:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 662/3295 [06:50<27:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 663/3295 [06:50<27:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 663/3295 [06:50<27:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 664/3295 [06:50<27:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 664/3295 [06:50<27:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 665/3295 [06:52<27:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 665/3295 [06:52<27:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 666/3295 [06:52<27:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 666/3295 [06:52<27:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 667/3295 [06:52<27:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 667/3295 [06:52<27:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 668/3295 [06:54<27:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 668/3295 [06:54<27:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 669/3295 [06:54<27:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 669/3295 [06:54<27:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 670/3295 [06:54<27:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 670/3295 [06:54<27:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 671/3295 [06:56<27:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 671/3295 [06:56<27:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 672/3295 [06:56<27:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 672/3295 [06:56<27:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 673/3295 [06:56<27:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 673/3295 [06:56<27:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 674/3295 [06:58<27:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 674/3295 [06:58<27:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 675/3295 [06:58<27:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  20%|██        | 675/3295 [06:58<27:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 676/3295 [06:58<27:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 676/3295 [06:58<27:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 677/3295 [07:00<27:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 677/3295 [07:00<27:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 678/3295 [07:00<27:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 678/3295 [07:00<27:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 679/3295 [07:00<26:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 679/3295 [07:00<26:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 680/3295 [07:02<27:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 680/3295 [07:02<27:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 681/3295 [07:02<27:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 681/3295 [07:02<27:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 682/3295 [07:02<26:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 682/3295 [07:02<26:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 683/3295 [07:03<27:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 683/3295 [07:03<27:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 684/3295 [07:03<26:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 684/3295 [07:03<26:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 685/3295 [07:03<26:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 685/3295 [07:03<26:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 686/3295 [07:05<26:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 686/3295 [07:05<26:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 687/3295 [07:05<26:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 687/3295 [07:05<26:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 688/3295 [07:05<26:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 688/3295 [07:05<26:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 689/3295 [07:07<26:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 689/3295 [07:07<26:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 690/3295 [07:07<26:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 690/3295 [07:07<26:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 691/3295 [07:07<26:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 691/3295 [07:07<26:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 692/3295 [07:09<26:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 692/3295 [07:09<26:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 693/3295 [07:09<26:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 693/3295 [07:09<26:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 694/3295 [07:09<26:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 694/3295 [07:09<26:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 695/3295 [07:11<26:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 695/3295 [07:11<26:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 696/3295 [07:11<26:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 696/3295 [07:11<26:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 697/3295 [07:11<26:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 697/3295 [07:11<26:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 698/3295 [07:12<26:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 698/3295 [07:12<26:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 699/3295 [07:12<26:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 699/3295 [07:12<26:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 700/3295 [07:13<26:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██        | 700/3295 [07:13<26:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 701/3295 [07:14<26:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 701/3295 [07:14<26:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 702/3295 [07:14<26:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 702/3295 [07:14<26:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 703/3295 [07:14<26:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 703/3295 [07:14<26:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 704/3295 [07:16<26:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 704/3295 [07:16<26:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 705/3295 [07:16<26:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 705/3295 [07:16<26:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 706/3295 [07:16<26:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 706/3295 [07:16<26:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 707/3295 [07:18<26:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 707/3295 [07:18<26:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 708/3295 [07:18<26:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  21%|██▏       | 708/3295 [07:18<26:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 709/3295 [07:18<26:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 709/3295 [07:18<26:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 710/3295 [07:20<26:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 710/3295 [07:20<26:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 711/3295 [07:20<26:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 711/3295 [07:20<26:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 712/3295 [07:20<26:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 712/3295 [07:20<26:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 713/3295 [07:22<26:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 713/3295 [07:22<26:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 714/3295 [07:22<26:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 714/3295 [07:22<26:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 715/3295 [07:22<26:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 715/3295 [07:22<26:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 716/3295 [07:24<26:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 716/3295 [07:24<26:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 717/3295 [07:24<26:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 717/3295 [07:24<26:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 718/3295 [07:24<26:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 718/3295 [07:24<26:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 719/3295 [07:25<26:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 719/3295 [07:25<26:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 720/3295 [07:25<26:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 720/3295 [07:25<26:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 721/3295 [07:25<26:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 721/3295 [07:25<26:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 722/3295 [07:27<26:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 722/3295 [07:27<26:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 723/3295 [07:27<26:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 723/3295 [07:27<26:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 724/3295 [07:27<26:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 724/3295 [07:27<26:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 725/3295 [07:29<26:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 725/3295 [07:29<26:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 726/3295 [07:29<26:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 726/3295 [07:29<26:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 727/3295 [07:29<26:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 727/3295 [07:29<26:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 728/3295 [07:31<26:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 728/3295 [07:31<26:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 729/3295 [07:31<26:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 729/3295 [07:31<26:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 730/3295 [07:31<26:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 730/3295 [07:31<26:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 731/3295 [07:32<26:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 731/3295 [07:32<26:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 732/3295 [07:32<26:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 732/3295 [07:32<26:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 733/3295 [07:32<26:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 733/3295 [07:32<26:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 734/3295 [07:34<26:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 734/3295 [07:34<26:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 735/3295 [07:34<26:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 735/3295 [07:34<26:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 736/3295 [07:34<26:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 736/3295 [07:34<26:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 737/3295 [07:36<26:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 737/3295 [07:36<26:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 738/3295 [07:36<26:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 738/3295 [07:36<26:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 739/3295 [07:36<26:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 739/3295 [07:36<26:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 740/3295 [07:38<26:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 740/3295 [07:38<26:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 741/3295 [07:38<26:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  22%|██▏       | 741/3295 [07:38<26:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 742/3295 [07:38<26:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 742/3295 [07:38<26:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 743/3295 [07:40<26:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 743/3295 [07:40<26:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 744/3295 [07:40<26:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 744/3295 [07:40<26:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 745/3295 [07:40<26:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 745/3295 [07:40<26:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 746/3295 [07:42<26:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 746/3295 [07:42<26:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 747/3295 [07:42<26:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 747/3295 [07:42<26:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 748/3295 [07:42<26:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 748/3295 [07:42<26:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 749/3295 [07:44<26:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 749/3295 [07:44<26:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 750/3295 [07:44<26:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 750/3295 [07:44<26:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 751/3295 [07:44<26:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 751/3295 [07:44<26:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 752/3295 [07:46<26:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 752/3295 [07:46<26:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 753/3295 [07:46<26:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 753/3295 [07:46<26:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 754/3295 [07:46<26:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 754/3295 [07:46<26:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 755/3295 [07:47<26:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 755/3295 [07:47<26:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 756/3295 [07:47<26:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 756/3295 [07:47<26:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 757/3295 [07:47<26:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 757/3295 [07:47<26:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 758/3295 [07:49<26:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 758/3295 [07:49<26:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 759/3295 [07:49<26:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 759/3295 [07:49<26:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 760/3295 [07:49<26:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 760/3295 [07:49<26:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 761/3295 [07:51<26:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 761/3295 [07:51<26:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 762/3295 [07:51<26:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 762/3295 [07:51<26:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 763/3295 [07:51<26:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 763/3295 [07:51<26:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 764/3295 [07:53<26:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 764/3295 [07:53<26:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 765/3295 [07:53<26:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 765/3295 [07:53<26:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 766/3295 [07:53<26:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 766/3295 [07:53<26:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 767/3295 [07:55<26:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 767/3295 [07:55<26:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 768/3295 [07:55<26:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 768/3295 [07:55<26:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 769/3295 [07:55<26:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 769/3295 [07:55<26:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 770/3295 [07:57<26:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 770/3295 [07:57<26:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 771/3295 [07:57<26:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 771/3295 [07:57<26:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 772/3295 [07:57<26:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 772/3295 [07:57<26:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 773/3295 [07:59<26:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 773/3295 [07:59<26:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 774/3295 [07:59<26:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  23%|██▎       | 774/3295 [07:59<26:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 775/3295 [07:59<25:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 775/3295 [07:59<25:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 776/3295 [08:00<26:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 776/3295 [08:00<26:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 777/3295 [08:00<25:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 777/3295 [08:00<25:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 778/3295 [08:00<25:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 778/3295 [08:00<25:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 779/3295 [08:02<25:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 779/3295 [08:02<25:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 780/3295 [08:02<25:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 780/3295 [08:02<25:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 781/3295 [08:02<25:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 781/3295 [08:02<25:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 782/3295 [08:04<25:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▎       | 782/3295 [08:04<25:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 783/3295 [08:04<25:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 783/3295 [08:04<25:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 784/3295 [08:04<25:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 784/3295 [08:04<25:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 785/3295 [08:06<25:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 785/3295 [08:06<25:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 786/3295 [08:06<25:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 786/3295 [08:06<25:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 787/3295 [08:06<25:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 787/3295 [08:06<25:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 788/3295 [08:08<25:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 788/3295 [08:08<25:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 789/3295 [08:08<25:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 789/3295 [08:08<25:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 790/3295 [08:08<25:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 790/3295 [08:08<25:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 791/3295 [08:10<25:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 791/3295 [08:10<25:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 792/3295 [08:10<25:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 792/3295 [08:10<25:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 793/3295 [08:10<25:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 793/3295 [08:10<25:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 794/3295 [08:12<25:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 794/3295 [08:12<25:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 795/3295 [08:12<25:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 795/3295 [08:12<25:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 796/3295 [08:12<25:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 796/3295 [08:12<25:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 797/3295 [08:13<25:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 797/3295 [08:13<25:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 798/3295 [08:14<25:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 798/3295 [08:14<25:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 799/3295 [08:14<25:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 799/3295 [08:14<25:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 800/3295 [08:15<25:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 800/3295 [08:15<25:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 801/3295 [08:15<25:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 801/3295 [08:15<25:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 802/3295 [08:15<25:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 802/3295 [08:15<25:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 803/3295 [08:17<25:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 803/3295 [08:17<25:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 804/3295 [08:17<25:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 804/3295 [08:17<25:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 805/3295 [08:17<25:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 805/3295 [08:17<25:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 806/3295 [08:19<25:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 806/3295 [08:19<25:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 807/3295 [08:19<25:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  24%|██▍       | 807/3295 [08:19<25:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 808/3295 [08:19<25:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 808/3295 [08:19<25:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 809/3295 [08:21<25:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 809/3295 [08:21<25:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 810/3295 [08:21<25:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 810/3295 [08:21<25:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 811/3295 [08:21<25:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 811/3295 [08:21<25:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 812/3295 [08:23<25:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 812/3295 [08:23<25:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 813/3295 [08:23<25:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 813/3295 [08:23<25:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 814/3295 [08:23<25:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 814/3295 [08:23<25:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 815/3295 [08:25<25:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 815/3295 [08:25<25:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 816/3295 [08:25<25:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 816/3295 [08:25<25:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 817/3295 [08:25<25:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 817/3295 [08:25<25:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 818/3295 [08:26<25:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 818/3295 [08:26<25:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 819/3295 [08:26<25:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 819/3295 [08:26<25:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 820/3295 [08:26<25:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 820/3295 [08:26<25:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 821/3295 [08:28<25:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 821/3295 [08:28<25:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 822/3295 [08:28<25:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 822/3295 [08:28<25:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 823/3295 [08:28<25:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▍       | 823/3295 [08:28<25:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 824/3295 [08:30<25:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 824/3295 [08:30<25:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 825/3295 [08:30<25:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 825/3295 [08:30<25:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 826/3295 [08:30<25:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 826/3295 [08:30<25:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 827/3295 [08:32<25:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 827/3295 [08:32<25:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 828/3295 [08:32<25:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 828/3295 [08:32<25:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 829/3295 [08:32<25:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 829/3295 [08:32<25:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 830/3295 [08:34<25:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 830/3295 [08:34<25:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 831/3295 [08:34<25:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 831/3295 [08:34<25:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 832/3295 [08:34<25:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 832/3295 [08:34<25:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 833/3295 [08:36<25:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 833/3295 [08:36<25:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 834/3295 [08:36<25:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 834/3295 [08:36<25:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 835/3295 [08:36<25:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 835/3295 [08:36<25:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 836/3295 [08:38<25:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 836/3295 [08:38<25:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 837/3295 [08:38<25:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 837/3295 [08:38<25:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 838/3295 [08:38<25:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 838/3295 [08:38<25:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 839/3295 [08:40<25:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 839/3295 [08:40<25:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 840/3295 [08:40<25:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  25%|██▌       | 840/3295 [08:40<25:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 841/3295 [08:40<25:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 841/3295 [08:40<25:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 842/3295 [08:41<25:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 842/3295 [08:41<25:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 843/3295 [08:41<25:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 843/3295 [08:41<25:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 844/3295 [08:41<25:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 844/3295 [08:41<25:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 845/3295 [08:43<25:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 845/3295 [08:43<25:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 846/3295 [08:43<25:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 846/3295 [08:43<25:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 847/3295 [08:43<25:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 847/3295 [08:43<25:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 848/3295 [08:45<25:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 848/3295 [08:45<25:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 849/3295 [08:45<25:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 849/3295 [08:45<25:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 850/3295 [08:45<25:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 850/3295 [08:45<25:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 851/3295 [08:47<25:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 851/3295 [08:47<25:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 852/3295 [08:47<25:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 852/3295 [08:47<25:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 853/3295 [08:47<25:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 853/3295 [08:47<25:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 854/3295 [08:49<25:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 854/3295 [08:49<25:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 855/3295 [08:49<25:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 855/3295 [08:49<25:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 856/3295 [08:49<25:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 856/3295 [08:49<25:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 857/3295 [08:51<25:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 857/3295 [08:51<25:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 858/3295 [08:51<25:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 858/3295 [08:51<25:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 859/3295 [08:51<25:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 859/3295 [08:51<25:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 860/3295 [08:53<25:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 860/3295 [08:53<25:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 861/3295 [08:53<25:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 861/3295 [08:53<25:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 862/3295 [08:53<25:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 862/3295 [08:53<25:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 863/3295 [08:55<25:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 863/3295 [08:55<25:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 864/3295 [08:55<25:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▌       | 864/3295 [08:55<25:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 865/3295 [08:55<25:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 865/3295 [08:55<25:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 866/3295 [08:57<25:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 866/3295 [08:57<25:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 867/3295 [08:57<25:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 867/3295 [08:57<25:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 868/3295 [08:57<25:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 868/3295 [08:57<25:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 869/3295 [08:58<25:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 869/3295 [08:58<25:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 870/3295 [08:58<25:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 870/3295 [08:58<25:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 871/3295 [08:58<24:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 871/3295 [08:58<24:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 872/3295 [09:00<25:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 872/3295 [09:00<25:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 873/3295 [09:00<25:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  26%|██▋       | 873/3295 [09:00<25:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 874/3295 [09:00<24:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 874/3295 [09:00<24:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 875/3295 [09:02<25:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 875/3295 [09:02<25:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 876/3295 [09:02<24:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 876/3295 [09:02<24:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 877/3295 [09:02<24:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 877/3295 [09:02<24:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 878/3295 [09:04<24:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 878/3295 [09:04<24:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 879/3295 [09:04<24:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 879/3295 [09:04<24:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 880/3295 [09:04<24:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 880/3295 [09:04<24:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 881/3295 [09:06<24:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 881/3295 [09:06<24:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 882/3295 [09:06<24:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 882/3295 [09:06<24:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 883/3295 [09:06<24:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 883/3295 [09:06<24:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 884/3295 [09:07<24:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 884/3295 [09:07<24:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 885/3295 [09:07<24:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 885/3295 [09:07<24:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 886/3295 [09:07<24:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 886/3295 [09:07<24:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 887/3295 [09:09<24:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 887/3295 [09:09<24:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 888/3295 [09:09<24:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 888/3295 [09:09<24:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 889/3295 [09:09<24:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 889/3295 [09:09<24:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 890/3295 [09:11<24:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 890/3295 [09:11<24:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 891/3295 [09:11<24:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 891/3295 [09:11<24:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 892/3295 [09:11<24:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 892/3295 [09:11<24:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 893/3295 [09:13<24:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 893/3295 [09:13<24:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 894/3295 [09:13<24:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 894/3295 [09:13<24:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 895/3295 [09:13<24:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 895/3295 [09:13<24:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 896/3295 [09:15<24:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 896/3295 [09:15<24:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 897/3295 [09:15<24:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 897/3295 [09:15<24:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 898/3295 [09:15<24:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 898/3295 [09:15<24:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 899/3295 [09:17<24:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 899/3295 [09:17<24:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 900/3295 [09:17<24:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 900/3295 [09:17<24:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 901/3295 [09:17<24:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 901/3295 [09:17<24:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 902/3295 [09:18<24:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 902/3295 [09:18<24:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 903/3295 [09:18<24:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 903/3295 [09:18<24:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 904/3295 [09:18<24:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 904/3295 [09:18<24:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 905/3295 [09:20<24:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 905/3295 [09:20<24:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 906/3295 [09:20<24:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  27%|██▋       | 906/3295 [09:20<24:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 907/3295 [09:20<24:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 907/3295 [09:20<24:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 908/3295 [09:22<24:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 908/3295 [09:22<24:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 909/3295 [09:22<24:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 909/3295 [09:22<24:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 910/3295 [09:22<24:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 910/3295 [09:22<24:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 911/3295 [09:24<24:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 911/3295 [09:24<24:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 912/3295 [09:24<24:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 912/3295 [09:24<24:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 913/3295 [09:24<24:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 913/3295 [09:24<24:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 914/3295 [09:26<24:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 914/3295 [09:26<24:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 915/3295 [09:26<24:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 915/3295 [09:26<24:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 916/3295 [09:26<24:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 916/3295 [09:26<24:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 917/3295 [09:28<24:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 917/3295 [09:28<24:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 918/3295 [09:28<24:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 918/3295 [09:28<24:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 919/3295 [09:28<24:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 919/3295 [09:28<24:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 920/3295 [09:30<24:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 920/3295 [09:30<24:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 921/3295 [09:30<24:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 921/3295 [09:30<24:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 922/3295 [09:30<24:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 922/3295 [09:30<24:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 923/3295 [09:31<24:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 923/3295 [09:31<24:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 924/3295 [09:31<24:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 924/3295 [09:31<24:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 925/3295 [09:31<24:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 925/3295 [09:31<24:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 926/3295 [09:33<24:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 926/3295 [09:33<24:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 927/3295 [09:33<24:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 927/3295 [09:33<24:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 928/3295 [09:33<24:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 928/3295 [09:33<24:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 929/3295 [09:35<24:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 929/3295 [09:35<24:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 930/3295 [09:35<24:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 930/3295 [09:35<24:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 931/3295 [09:35<24:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 931/3295 [09:35<24:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 932/3295 [09:37<24:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 932/3295 [09:37<24:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 933/3295 [09:37<24:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 933/3295 [09:37<24:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 934/3295 [09:37<24:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 934/3295 [09:37<24:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 935/3295 [09:39<24:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 935/3295 [09:39<24:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 936/3295 [09:39<24:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 936/3295 [09:39<24:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 937/3295 [09:39<24:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 937/3295 [09:39<24:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 938/3295 [09:40<24:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 938/3295 [09:40<24:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 939/3295 [09:40<24:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  28%|██▊       | 939/3295 [09:40<24:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 940/3295 [09:40<24:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 940/3295 [09:40<24:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 941/3295 [09:42<24:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 941/3295 [09:42<24:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 942/3295 [09:42<24:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 942/3295 [09:42<24:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 943/3295 [09:42<24:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 943/3295 [09:42<24:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 944/3295 [09:44<24:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 944/3295 [09:44<24:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 945/3295 [09:44<24:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 945/3295 [09:44<24:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 946/3295 [09:44<24:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 946/3295 [09:44<24:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 947/3295 [09:46<24:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▊       | 947/3295 [09:46<24:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 948/3295 [09:46<24:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 948/3295 [09:46<24:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 949/3295 [09:46<24:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 949/3295 [09:46<24:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 950/3295 [09:48<24:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 950/3295 [09:48<24:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 951/3295 [09:48<24:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 951/3295 [09:48<24:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 952/3295 [09:48<24:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 952/3295 [09:48<24:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 953/3295 [09:50<24:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 953/3295 [09:50<24:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 954/3295 [09:50<24:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 954/3295 [09:50<24:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 955/3295 [09:50<24:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 955/3295 [09:50<24:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 956/3295 [09:51<24:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 956/3295 [09:51<24:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 957/3295 [09:51<24:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 957/3295 [09:51<24:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 958/3295 [09:51<24:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 958/3295 [09:51<24:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 959/3295 [09:53<24:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 959/3295 [09:53<24:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 960/3295 [09:53<24:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 960/3295 [09:53<24:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 961/3295 [09:53<24:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 961/3295 [09:53<24:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 962/3295 [09:55<24:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 962/3295 [09:55<24:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 963/3295 [09:55<24:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 963/3295 [09:55<24:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 964/3295 [09:55<24:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 964/3295 [09:55<24:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 965/3295 [09:57<24:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 965/3295 [09:57<24:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 966/3295 [09:57<24:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 966/3295 [09:57<24:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 967/3295 [09:57<23:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 967/3295 [09:57<23:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 968/3295 [09:59<24:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 968/3295 [09:59<24:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 969/3295 [09:59<23:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 969/3295 [09:59<23:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 970/3295 [09:59<23:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 970/3295 [09:59<23:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 971/3295 [10:01<23:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 971/3295 [10:01<23:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 972/3295 [10:01<23:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  29%|██▉       | 972/3295 [10:01<23:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 973/3295 [10:01<23:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 973/3295 [10:01<23:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 974/3295 [10:02<23:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 974/3295 [10:02<23:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 975/3295 [10:02<23:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 975/3295 [10:02<23:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 976/3295 [10:03<23:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 976/3295 [10:03<23:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 977/3295 [10:04<23:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 977/3295 [10:04<23:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 978/3295 [10:04<23:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 978/3295 [10:04<23:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 979/3295 [10:04<23:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 979/3295 [10:04<23:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 980/3295 [10:06<23:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 980/3295 [10:06<23:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 981/3295 [10:06<23:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 981/3295 [10:06<23:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 982/3295 [10:06<23:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 982/3295 [10:06<23:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 983/3295 [10:08<23:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 983/3295 [10:08<23:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 984/3295 [10:08<23:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 984/3295 [10:08<23:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 985/3295 [10:08<23:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 985/3295 [10:08<23:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 986/3295 [10:10<23:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 986/3295 [10:10<23:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 987/3295 [10:10<23:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 987/3295 [10:10<23:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 988/3295 [10:10<23:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|██▉       | 988/3295 [10:10<23:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 989/3295 [10:12<23:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 989/3295 [10:12<23:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 990/3295 [10:12<23:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 990/3295 [10:12<23:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 991/3295 [10:12<23:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 991/3295 [10:12<23:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 992/3295 [10:14<23:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 992/3295 [10:14<23:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 993/3295 [10:14<23:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 993/3295 [10:14<23:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 994/3295 [10:14<23:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 994/3295 [10:14<23:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 995/3295 [10:15<23:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 995/3295 [10:15<23:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 996/3295 [10:15<23:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 996/3295 [10:15<23:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 997/3295 [10:16<23:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 997/3295 [10:16<23:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 998/3295 [10:17<23:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 998/3295 [10:17<23:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 999/3295 [10:17<23:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 999/3295 [10:17<23:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1000/3295 [10:17<23:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1000/3295 [10:17<23:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1001/3295 [10:19<23:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1001/3295 [10:19<23:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1002/3295 [10:19<23:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1002/3295 [10:19<23:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1003/3295 [10:19<23:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1003/3295 [10:19<23:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1004/3295 [10:21<23:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  30%|███       | 1004/3295 [10:21<23:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1005/3295 [10:21<23:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1005/3295 [10:21<23:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1006/3295 [10:21<23:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1006/3295 [10:21<23:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1007/3295 [10:23<23:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1007/3295 [10:23<23:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1008/3295 [10:23<23:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1008/3295 [10:23<23:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1009/3295 [10:23<23:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1009/3295 [10:23<23:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1010/3295 [10:25<23:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1010/3295 [10:25<23:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1011/3295 [10:25<23:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1011/3295 [10:25<23:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1012/3295 [10:25<23:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1012/3295 [10:25<23:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1013/3295 [10:26<23:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1013/3295 [10:26<23:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1014/3295 [10:26<23:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1014/3295 [10:26<23:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1015/3295 [10:26<23:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1015/3295 [10:26<23:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1016/3295 [10:28<23:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1016/3295 [10:28<23:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1017/3295 [10:28<23:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1017/3295 [10:28<23:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1018/3295 [10:28<23:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1018/3295 [10:28<23:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1019/3295 [10:30<23:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1019/3295 [10:30<23:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1020/3295 [10:30<23:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1020/3295 [10:30<23:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1021/3295 [10:30<23:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1021/3295 [10:30<23:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1022/3295 [10:32<23:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1022/3295 [10:32<23:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1023/3295 [10:32<23:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1023/3295 [10:32<23:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1024/3295 [10:32<23:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1024/3295 [10:32<23:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1025/3295 [10:34<23:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1025/3295 [10:34<23:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1026/3295 [10:34<23:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1026/3295 [10:34<23:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1027/3295 [10:34<23:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1027/3295 [10:34<23:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1028/3295 [10:36<23:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1028/3295 [10:36<23:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1029/3295 [10:36<23:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███       | 1029/3295 [10:36<23:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1030/3295 [10:36<23:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1030/3295 [10:36<23:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1031/3295 [10:38<23:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1031/3295 [10:38<23:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1032/3295 [10:38<23:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1032/3295 [10:38<23:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1033/3295 [10:38<23:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1033/3295 [10:38<23:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1034/3295 [10:40<23:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1034/3295 [10:40<23:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1035/3295 [10:40<23:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1035/3295 [10:40<23:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1036/3295 [10:40<23:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1036/3295 [10:40<23:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1037/3295 [10:41<23:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  31%|███▏      | 1037/3295 [10:41<23:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1038/3295 [10:42<23:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1038/3295 [10:42<23:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1039/3295 [10:42<23:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1039/3295 [10:42<23:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1040/3295 [10:43<23:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1040/3295 [10:43<23:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1041/3295 [10:43<23:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1041/3295 [10:43<23:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1042/3295 [10:43<23:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1042/3295 [10:43<23:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1043/3295 [10:45<23:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1043/3295 [10:45<23:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1044/3295 [10:45<23:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1044/3295 [10:45<23:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1045/3295 [10:45<23:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1045/3295 [10:45<23:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1046/3295 [10:47<23:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1046/3295 [10:47<23:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1047/3295 [10:47<23:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1047/3295 [10:47<23:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1048/3295 [10:47<23:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1048/3295 [10:47<23:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1049/3295 [10:49<23:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1049/3295 [10:49<23:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1050/3295 [10:49<23:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1050/3295 [10:49<23:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1051/3295 [10:49<23:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1051/3295 [10:49<23:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1052/3295 [10:51<23:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1052/3295 [10:51<23:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1053/3295 [10:51<23:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1053/3295 [10:51<23:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1054/3295 [10:51<23:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1054/3295 [10:51<23:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1055/3295 [10:52<23:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1055/3295 [10:52<23:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1056/3295 [10:52<23:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1056/3295 [10:52<23:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1057/3295 [10:52<23:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1057/3295 [10:52<23:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1058/3295 [10:54<23:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1058/3295 [10:54<23:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1059/3295 [10:54<23:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1059/3295 [10:54<23:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1060/3295 [10:54<23:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1060/3295 [10:54<23:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1061/3295 [10:56<23:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1061/3295 [10:56<23:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1062/3295 [10:56<22:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1062/3295 [10:56<22:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1063/3295 [10:56<22:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1063/3295 [10:56<22:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1064/3295 [10:58<22:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1064/3295 [10:58<22:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1065/3295 [10:58<22:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1065/3295 [10:58<22:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1066/3295 [10:58<22:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1066/3295 [10:58<22:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1067/3295 [10:59<22:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1067/3295 [10:59<22:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1068/3295 [10:59<22:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1068/3295 [10:59<22:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1069/3295 [10:59<22:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1069/3295 [10:59<22:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1070/3295 [11:01<22:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  32%|███▏      | 1070/3295 [11:01<22:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1071/3295 [11:01<22:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1071/3295 [11:01<22:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1072/3295 [11:01<22:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1072/3295 [11:01<22:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1073/3295 [11:03<22:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1073/3295 [11:03<22:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1074/3295 [11:03<22:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1074/3295 [11:03<22:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1075/3295 [11:03<22:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1075/3295 [11:03<22:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1076/3295 [11:05<22:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1076/3295 [11:05<22:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1077/3295 [11:05<22:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1077/3295 [11:05<22:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1078/3295 [11:05<22:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1078/3295 [11:05<22:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1079/3295 [11:07<22:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1079/3295 [11:07<22:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1080/3295 [11:07<22:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1080/3295 [11:07<22:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1081/3295 [11:07<22:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1081/3295 [11:07<22:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1082/3295 [11:09<22:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1082/3295 [11:09<22:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1083/3295 [11:09<22:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1083/3295 [11:09<22:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1084/3295 [11:09<22:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1084/3295 [11:09<22:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1085/3295 [11:11<22:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1085/3295 [11:11<22:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1086/3295 [11:11<22:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1086/3295 [11:11<22:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1087/3295 [11:11<22:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1087/3295 [11:11<22:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1088/3295 [11:13<22:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1088/3295 [11:13<22:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1089/3295 [11:13<22:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1089/3295 [11:13<22:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1090/3295 [11:13<22:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1090/3295 [11:13<22:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1091/3295 [11:14<22:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1091/3295 [11:14<22:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1092/3295 [11:14<22:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1092/3295 [11:14<22:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1093/3295 [11:15<22:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1093/3295 [11:15<22:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1094/3295 [11:16<22:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1094/3295 [11:16<22:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1095/3295 [11:16<22:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1095/3295 [11:16<22:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1096/3295 [11:17<22:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1096/3295 [11:17<22:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1097/3295 [11:18<22:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1097/3295 [11:18<22:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1098/3295 [11:18<22:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1098/3295 [11:18<22:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1099/3295 [11:18<22:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1099/3295 [11:18<22:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1100/3295 [11:20<22:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1100/3295 [11:20<22:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1101/3295 [11:20<22:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1101/3295 [11:20<22:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1102/3295 [11:20<22:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1102/3295 [11:20<22:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1103/3295 [11:22<22:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  33%|███▎      | 1103/3295 [11:22<22:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1104/3295 [11:22<22:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1104/3295 [11:22<22:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1105/3295 [11:22<22:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1105/3295 [11:22<22:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1106/3295 [11:24<22:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1106/3295 [11:24<22:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1107/3295 [11:24<22:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1107/3295 [11:24<22:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1108/3295 [11:24<22:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1108/3295 [11:24<22:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1109/3295 [11:26<22:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1109/3295 [11:26<22:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1110/3295 [11:26<22:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1110/3295 [11:26<22:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1111/3295 [11:26<22:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1111/3295 [11:26<22:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1112/3295 [11:28<22:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▎      | 1112/3295 [11:28<22:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1113/3295 [11:28<22:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1113/3295 [11:28<22:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1114/3295 [11:28<22:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1114/3295 [11:28<22:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1115/3295 [11:29<22:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1115/3295 [11:29<22:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1116/3295 [11:29<22:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1116/3295 [11:29<22:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1117/3295 [11:29<22:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1117/3295 [11:29<22:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1118/3295 [11:31<22:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1118/3295 [11:31<22:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1119/3295 [11:31<22:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1119/3295 [11:31<22:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1120/3295 [11:31<22:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1120/3295 [11:31<22:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1121/3295 [11:33<22:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1121/3295 [11:33<22:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1122/3295 [11:33<22:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1122/3295 [11:33<22:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1123/3295 [11:33<22:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1123/3295 [11:33<22:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1124/3295 [11:35<22:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1124/3295 [11:35<22:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1125/3295 [11:35<22:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1125/3295 [11:35<22:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1126/3295 [11:35<22:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1126/3295 [11:35<22:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1127/3295 [11:37<22:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1127/3295 [11:37<22:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1128/3295 [11:37<22:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1128/3295 [11:37<22:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1129/3295 [11:37<22:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1129/3295 [11:37<22:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1130/3295 [11:38<22:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1130/3295 [11:38<22:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1131/3295 [11:38<22:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1131/3295 [11:38<22:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1132/3295 [11:38<22:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1132/3295 [11:38<22:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1133/3295 [11:40<22:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1133/3295 [11:40<22:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1134/3295 [11:40<22:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1134/3295 [11:40<22:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1135/3295 [11:40<22:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1135/3295 [11:40<22:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1136/3295 [11:42<22:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  34%|███▍      | 1136/3295 [11:42<22:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1137/3295 [11:42<22:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1137/3295 [11:42<22:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1138/3295 [11:42<22:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1138/3295 [11:42<22:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1139/3295 [11:44<22:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1139/3295 [11:44<22:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1140/3295 [11:44<22:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1140/3295 [11:44<22:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1141/3295 [11:44<22:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1141/3295 [11:44<22:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1142/3295 [11:46<22:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1142/3295 [11:46<22:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1143/3295 [11:46<22:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1143/3295 [11:46<22:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1144/3295 [11:46<22:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1144/3295 [11:46<22:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1145/3295 [11:48<22:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1145/3295 [11:48<22:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1146/3295 [11:48<22:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1146/3295 [11:48<22:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1147/3295 [11:48<22:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1147/3295 [11:48<22:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1148/3295 [11:50<22:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1148/3295 [11:50<22:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1149/3295 [11:50<22:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1149/3295 [11:50<22:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1150/3295 [11:50<22:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1150/3295 [11:50<22:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1151/3295 [11:52<22:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1151/3295 [11:52<22:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1152/3295 [11:52<22:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1152/3295 [11:52<22:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1153/3295 [11:52<22:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▍      | 1153/3295 [11:52<22:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1154/3295 [11:53<22:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1154/3295 [11:53<22:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1155/3295 [11:53<22:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1155/3295 [11:53<22:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1156/3295 [11:53<22:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1156/3295 [11:53<22:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1157/3295 [11:55<22:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1157/3295 [11:55<22:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1158/3295 [11:55<22:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1158/3295 [11:55<22:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1159/3295 [11:55<21:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1159/3295 [11:55<21:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1160/3295 [11:57<22:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1160/3295 [11:57<22:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1161/3295 [11:57<21:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1161/3295 [11:57<21:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1162/3295 [11:57<21:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1162/3295 [11:57<21:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1163/3295 [11:59<21:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1163/3295 [11:59<21:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1164/3295 [11:59<21:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1164/3295 [11:59<21:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1165/3295 [11:59<21:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1165/3295 [11:59<21:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1166/3295 [12:01<21:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1166/3295 [12:01<21:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1167/3295 [12:01<21:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1167/3295 [12:01<21:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1168/3295 [12:01<21:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1168/3295 [12:01<21:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1169/3295 [12:03<21:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  35%|███▌      | 1169/3295 [12:03<21:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1170/3295 [12:03<21:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1170/3295 [12:03<21:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1171/3295 [12:03<21:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1171/3295 [12:03<21:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1172/3295 [12:04<21:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1172/3295 [12:04<21:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1173/3295 [12:04<21:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1173/3295 [12:04<21:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1174/3295 [12:05<21:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1174/3295 [12:05<21:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1175/3295 [12:06<21:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1175/3295 [12:06<21:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1176/3295 [12:06<21:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1176/3295 [12:06<21:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1177/3295 [12:06<21:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1177/3295 [12:06<21:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1178/3295 [12:08<21:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1178/3295 [12:08<21:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1179/3295 [12:08<21:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1179/3295 [12:08<21:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1180/3295 [12:08<21:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1180/3295 [12:08<21:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1181/3295 [12:10<21:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1181/3295 [12:10<21:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1182/3295 [12:10<21:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1182/3295 [12:10<21:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1183/3295 [12:10<21:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1183/3295 [12:10<21:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1184/3295 [12:12<21:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1184/3295 [12:12<21:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1185/3295 [12:12<21:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1185/3295 [12:12<21:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1186/3295 [12:12<21:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1186/3295 [12:12<21:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1187/3295 [12:14<21:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1187/3295 [12:14<21:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1188/3295 [12:14<21:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1188/3295 [12:14<21:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1189/3295 [12:14<21:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1189/3295 [12:14<21:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1190/3295 [12:16<21:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1190/3295 [12:16<21:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1191/3295 [12:16<21:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1191/3295 [12:16<21:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1192/3295 [12:16<21:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1192/3295 [12:16<21:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1193/3295 [12:17<21:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1193/3295 [12:17<21:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1194/3295 [12:17<21:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▌      | 1194/3295 [12:17<21:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1195/3295 [12:17<21:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1195/3295 [12:17<21:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1196/3295 [12:19<21:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1196/3295 [12:19<21:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1197/3295 [12:19<21:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1197/3295 [12:19<21:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1198/3295 [12:19<21:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1198/3295 [12:19<21:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1199/3295 [12:21<21:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1199/3295 [12:21<21:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1200/3295 [12:21<21:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1200/3295 [12:21<21:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1201/3295 [12:21<21:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1201/3295 [12:21<21:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1202/3295 [12:23<21:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  36%|███▋      | 1202/3295 [12:23<21:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1203/3295 [12:23<21:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1203/3295 [12:23<21:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1204/3295 [12:23<21:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1204/3295 [12:23<21:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1205/3295 [12:25<21:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1205/3295 [12:25<21:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1206/3295 [12:25<21:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1206/3295 [12:25<21:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1207/3295 [12:25<21:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1207/3295 [12:25<21:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1208/3295 [12:26<21:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1208/3295 [12:26<21:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1209/3295 [12:26<21:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1209/3295 [12:26<21:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1210/3295 [12:26<21:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1210/3295 [12:26<21:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1211/3295 [12:28<21:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1211/3295 [12:28<21:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1212/3295 [12:28<21:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1212/3295 [12:28<21:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1213/3295 [12:28<21:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1213/3295 [12:28<21:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1214/3295 [12:30<21:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1214/3295 [12:30<21:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1215/3295 [12:30<21:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1215/3295 [12:30<21:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1216/3295 [12:30<21:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1216/3295 [12:30<21:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1217/3295 [12:32<21:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1217/3295 [12:32<21:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1218/3295 [12:32<21:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1218/3295 [12:32<21:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1219/3295 [12:32<21:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1219/3295 [12:32<21:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1220/3295 [12:34<21:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1220/3295 [12:34<21:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1221/3295 [12:34<21:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1221/3295 [12:34<21:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1222/3295 [12:34<21:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1222/3295 [12:34<21:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1223/3295 [12:36<21:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1223/3295 [12:36<21:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1224/3295 [12:36<21:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1224/3295 [12:36<21:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1225/3295 [12:36<21:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1225/3295 [12:36<21:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1226/3295 [12:37<21:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1226/3295 [12:37<21:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1227/3295 [12:37<21:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1227/3295 [12:37<21:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1228/3295 [12:37<21:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1228/3295 [12:37<21:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1229/3295 [12:39<21:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1229/3295 [12:39<21:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1230/3295 [12:39<21:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1230/3295 [12:39<21:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1231/3295 [12:39<21:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1231/3295 [12:39<21:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1232/3295 [12:41<21:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1232/3295 [12:41<21:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1233/3295 [12:41<21:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1233/3295 [12:41<21:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1234/3295 [12:41<21:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1234/3295 [12:41<21:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1235/3295 [12:43<21:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  37%|███▋      | 1235/3295 [12:43<21:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1236/3295 [12:43<21:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1236/3295 [12:43<21:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1237/3295 [12:43<21:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1237/3295 [12:43<21:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1238/3295 [12:45<21:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1238/3295 [12:45<21:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1239/3295 [12:45<21:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1239/3295 [12:45<21:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1240/3295 [12:45<21:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1240/3295 [12:45<21:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1241/3295 [12:47<21:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1241/3295 [12:47<21:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1242/3295 [12:47<21:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1242/3295 [12:47<21:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1243/3295 [12:47<21:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1243/3295 [12:47<21:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1244/3295 [12:49<21:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1244/3295 [12:49<21:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1245/3295 [12:49<21:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1245/3295 [12:49<21:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1246/3295 [12:49<21:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1246/3295 [12:49<21:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1247/3295 [12:51<21:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1247/3295 [12:51<21:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1248/3295 [12:51<21:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1248/3295 [12:51<21:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1249/3295 [12:51<21:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1249/3295 [12:51<21:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1250/3295 [12:52<21:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1250/3295 [12:52<21:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1251/3295 [12:52<21:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1251/3295 [12:52<21:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1252/3295 [12:52<21:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1252/3295 [12:52<21:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1253/3295 [12:54<21:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1253/3295 [12:54<21:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1254/3295 [12:54<21:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1254/3295 [12:54<21:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1255/3295 [12:54<20:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1255/3295 [12:54<20:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1256/3295 [12:56<21:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1256/3295 [12:56<21:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1257/3295 [12:56<20:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1257/3295 [12:56<20:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1258/3295 [12:56<20:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1258/3295 [12:56<20:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1259/3295 [12:58<20:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1259/3295 [12:58<20:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1260/3295 [12:58<20:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1260/3295 [12:58<20:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1261/3295 [12:58<20:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1261/3295 [12:58<20:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1262/3295 [13:00<20:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1262/3295 [13:00<20:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1263/3295 [13:00<20:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1263/3295 [13:00<20:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1264/3295 [13:00<20:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1264/3295 [13:00<20:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1265/3295 [13:01<20:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1265/3295 [13:01<20:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1266/3295 [13:02<20:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1266/3295 [13:02<20:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1267/3295 [13:02<20:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1267/3295 [13:02<20:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1268/3295 [13:03<20:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  38%|███▊      | 1268/3295 [13:03<20:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1269/3295 [13:03<20:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1269/3295 [13:03<20:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1270/3295 [13:03<20:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1270/3295 [13:03<20:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1271/3295 [13:05<20:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1271/3295 [13:05<20:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1272/3295 [13:05<20:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1272/3295 [13:05<20:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1273/3295 [13:05<20:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1273/3295 [13:05<20:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1274/3295 [13:07<20:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1274/3295 [13:07<20:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1275/3295 [13:07<20:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1275/3295 [13:07<20:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1276/3295 [13:07<20:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▊      | 1276/3295 [13:07<20:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1277/3295 [13:09<20:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1277/3295 [13:09<20:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1278/3295 [13:09<20:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1278/3295 [13:09<20:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1279/3295 [13:09<20:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1279/3295 [13:09<20:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1280/3295 [13:11<20:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1280/3295 [13:11<20:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1281/3295 [13:11<20:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1281/3295 [13:11<20:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1282/3295 [13:11<20:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1282/3295 [13:11<20:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1283/3295 [13:13<20:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1283/3295 [13:13<20:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1284/3295 [13:13<20:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1284/3295 [13:13<20:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1285/3295 [13:13<20:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1285/3295 [13:13<20:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1286/3295 [13:14<20:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1286/3295 [13:14<20:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1287/3295 [13:14<20:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1287/3295 [13:14<20:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1288/3295 [13:15<20:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1288/3295 [13:15<20:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1289/3295 [13:16<20:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1289/3295 [13:16<20:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1290/3295 [13:16<20:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1290/3295 [13:16<20:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1291/3295 [13:16<20:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1291/3295 [13:16<20:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1292/3295 [13:18<20:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1292/3295 [13:18<20:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1293/3295 [13:18<20:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1293/3295 [13:18<20:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1294/3295 [13:18<20:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1294/3295 [13:18<20:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1295/3295 [13:20<20:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1295/3295 [13:20<20:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1296/3295 [13:20<20:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1296/3295 [13:20<20:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1297/3295 [13:20<20:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1297/3295 [13:20<20:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1298/3295 [13:22<20:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1298/3295 [13:22<20:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1299/3295 [13:22<20:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1299/3295 [13:22<20:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1300/3295 [13:22<20:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1300/3295 [13:22<20:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1301/3295 [13:24<20:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  39%|███▉      | 1301/3295 [13:24<20:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1302/3295 [13:24<20:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1302/3295 [13:24<20:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1303/3295 [13:24<20:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1303/3295 [13:24<20:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1304/3295 [13:26<20:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1304/3295 [13:26<20:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1305/3295 [13:26<20:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1305/3295 [13:26<20:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1306/3295 [13:26<20:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1306/3295 [13:26<20:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1307/3295 [13:27<20:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1307/3295 [13:27<20:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1308/3295 [13:27<20:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1308/3295 [13:27<20:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1309/3295 [13:27<20:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1309/3295 [13:27<20:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1310/3295 [13:29<20:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1310/3295 [13:29<20:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1311/3295 [13:29<20:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1311/3295 [13:29<20:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1312/3295 [13:29<20:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1312/3295 [13:29<20:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1313/3295 [13:31<20:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1313/3295 [13:31<20:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1314/3295 [13:31<20:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1314/3295 [13:31<20:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1315/3295 [13:31<20:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1315/3295 [13:31<20:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1316/3295 [13:33<20:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1316/3295 [13:33<20:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1317/3295 [13:33<20:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|███▉      | 1317/3295 [13:33<20:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1318/3295 [13:33<20:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1318/3295 [13:33<20:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1319/3295 [13:35<20:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1319/3295 [13:35<20:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1320/3295 [13:35<20:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1320/3295 [13:35<20:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1321/3295 [13:35<20:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1321/3295 [13:35<20:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1322/3295 [13:37<20:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1322/3295 [13:37<20:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1323/3295 [13:37<20:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1323/3295 [13:37<20:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1324/3295 [13:37<20:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1324/3295 [13:37<20:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1325/3295 [13:39<20:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1325/3295 [13:39<20:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1326/3295 [13:39<20:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1326/3295 [13:39<20:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1327/3295 [13:39<20:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1327/3295 [13:39<20:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1328/3295 [13:40<20:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1328/3295 [13:40<20:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1329/3295 [13:40<20:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1329/3295 [13:40<20:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1330/3295 [13:40<20:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1330/3295 [13:40<20:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1331/3295 [13:42<20:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1331/3295 [13:42<20:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1332/3295 [13:42<20:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1332/3295 [13:42<20:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1333/3295 [13:42<20:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1333/3295 [13:42<20:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1334/3295 [13:44<20:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  40%|████      | 1334/3295 [13:44<20:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1335/3295 [13:44<20:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1335/3295 [13:44<20:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1336/3295 [13:44<20:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1336/3295 [13:44<20:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1337/3295 [13:46<20:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1337/3295 [13:46<20:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1338/3295 [13:46<20:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1338/3295 [13:46<20:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1339/3295 [13:46<20:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1339/3295 [13:46<20:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1340/3295 [13:48<20:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1340/3295 [13:48<20:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1341/3295 [13:48<20:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1341/3295 [13:48<20:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1342/3295 [13:48<20:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1342/3295 [13:48<20:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1343/3295 [13:49<20:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1343/3295 [13:49<20:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1344/3295 [13:49<20:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1344/3295 [13:49<20:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1345/3295 [13:49<20:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1345/3295 [13:49<20:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1346/3295 [13:51<20:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1346/3295 [13:51<20:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1347/3295 [13:51<20:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1347/3295 [13:51<20:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1348/3295 [13:51<20:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1348/3295 [13:51<20:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1349/3295 [13:53<20:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1349/3295 [13:53<20:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1350/3295 [13:53<20:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1350/3295 [13:53<20:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1351/3295 [13:53<19:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1351/3295 [13:53<19:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1352/3295 [13:55<20:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1352/3295 [13:55<20:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1353/3295 [13:55<19:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1353/3295 [13:55<19:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1354/3295 [13:55<19:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1354/3295 [13:55<19:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1355/3295 [13:57<19:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1355/3295 [13:57<19:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1356/3295 [13:57<19:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1356/3295 [13:57<19:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1357/3295 [13:57<19:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1357/3295 [13:57<19:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1358/3295 [13:59<19:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1358/3295 [13:59<19:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1359/3295 [13:59<19:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████      | 1359/3295 [13:59<19:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1360/3295 [13:59<19:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1360/3295 [13:59<19:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1361/3295 [14:01<19:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1361/3295 [14:01<19:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1362/3295 [14:01<19:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1362/3295 [14:01<19:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1363/3295 [14:01<19:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1363/3295 [14:01<19:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1364/3295 [14:03<19:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1364/3295 [14:03<19:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1365/3295 [14:03<19:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1365/3295 [14:03<19:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1366/3295 [14:03<19:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1366/3295 [14:03<19:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1367/3295 [14:05<19:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  41%|████▏     | 1367/3295 [14:05<19:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1368/3295 [14:05<19:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1368/3295 [14:05<19:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1369/3295 [14:05<19:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1369/3295 [14:05<19:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1370/3295 [14:07<19:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1370/3295 [14:07<19:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1371/3295 [14:07<19:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1371/3295 [14:07<19:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1372/3295 [14:07<19:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1372/3295 [14:07<19:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1373/3295 [14:09<19:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1373/3295 [14:09<19:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1374/3295 [14:09<19:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1374/3295 [14:09<19:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1375/3295 [14:09<19:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1375/3295 [14:09<19:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1376/3295 [14:11<19:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1376/3295 [14:11<19:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1377/3295 [14:11<19:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1377/3295 [14:11<19:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1378/3295 [14:11<19:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1378/3295 [14:11<19:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1379/3295 [14:12<19:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1379/3295 [14:12<19:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1380/3295 [14:13<19:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1380/3295 [14:13<19:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1381/3295 [14:13<19:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1381/3295 [14:13<19:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1382/3295 [14:14<19:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1382/3295 [14:14<19:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1383/3295 [14:14<19:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1383/3295 [14:14<19:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1384/3295 [14:14<19:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1384/3295 [14:14<19:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1385/3295 [14:16<19:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1385/3295 [14:16<19:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1386/3295 [14:16<19:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1386/3295 [14:16<19:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1387/3295 [14:16<19:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1387/3295 [14:16<19:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1388/3295 [14:18<19:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1388/3295 [14:18<19:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1389/3295 [14:18<19:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1389/3295 [14:18<19:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1390/3295 [14:18<19:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1390/3295 [14:18<19:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1391/3295 [14:20<19:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1391/3295 [14:20<19:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1392/3295 [14:20<19:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1392/3295 [14:20<19:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1393/3295 [14:20<19:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1393/3295 [14:20<19:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1394/3295 [14:22<19:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1394/3295 [14:22<19:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1395/3295 [14:22<19:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1395/3295 [14:22<19:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1396/3295 [14:22<19:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1396/3295 [14:22<19:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1397/3295 [14:24<19:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1397/3295 [14:24<19:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1398/3295 [14:24<19:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1398/3295 [14:24<19:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1399/3295 [14:24<19:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1399/3295 [14:24<19:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1400/3295 [14:26<19:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  42%|████▏     | 1400/3295 [14:26<19:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1401/3295 [14:26<19:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1401/3295 [14:26<19:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1402/3295 [14:26<19:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1402/3295 [14:26<19:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1403/3295 [14:28<19:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1403/3295 [14:28<19:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1404/3295 [14:28<19:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1404/3295 [14:28<19:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1405/3295 [14:28<19:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1405/3295 [14:28<19:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1406/3295 [14:30<19:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1406/3295 [14:30<19:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1407/3295 [14:30<19:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1407/3295 [14:30<19:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1408/3295 [14:30<19:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1408/3295 [14:30<19:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1409/3295 [14:31<19:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1409/3295 [14:31<19:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1410/3295 [14:31<19:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1410/3295 [14:31<19:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1411/3295 [14:31<19:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1411/3295 [14:31<19:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1412/3295 [14:33<19:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1412/3295 [14:33<19:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1413/3295 [14:33<19:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1413/3295 [14:33<19:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1414/3295 [14:33<19:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1414/3295 [14:33<19:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1415/3295 [14:35<19:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1415/3295 [14:35<19:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1416/3295 [14:35<19:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1416/3295 [14:35<19:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1417/3295 [14:35<19:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1417/3295 [14:35<19:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1418/3295 [14:37<19:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1418/3295 [14:37<19:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1419/3295 [14:37<19:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1419/3295 [14:37<19:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1420/3295 [14:37<19:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1420/3295 [14:37<19:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1421/3295 [14:39<19:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1421/3295 [14:39<19:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1422/3295 [14:39<19:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1422/3295 [14:39<19:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1423/3295 [14:39<19:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1423/3295 [14:39<19:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1424/3295 [14:41<19:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1424/3295 [14:41<19:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1425/3295 [14:41<19:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1425/3295 [14:41<19:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1426/3295 [14:41<19:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1426/3295 [14:41<19:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1427/3295 [14:43<19:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1427/3295 [14:43<19:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1428/3295 [14:43<19:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1428/3295 [14:43<19:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1429/3295 [14:43<19:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1429/3295 [14:43<19:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1430/3295 [14:44<19:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1430/3295 [14:44<19:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1431/3295 [14:44<19:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1431/3295 [14:44<19:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1432/3295 [14:45<19:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1432/3295 [14:45<19:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1433/3295 [14:46<19:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  43%|████▎     | 1433/3295 [14:46<19:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1434/3295 [14:46<19:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1434/3295 [14:46<19:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1435/3295 [14:46<19:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1435/3295 [14:46<19:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1436/3295 [14:48<19:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1436/3295 [14:48<19:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1437/3295 [14:48<19:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1437/3295 [14:48<19:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1438/3295 [14:48<19:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1438/3295 [14:48<19:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1439/3295 [14:50<19:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1439/3295 [14:50<19:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1440/3295 [14:50<19:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1440/3295 [14:50<19:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1441/3295 [14:50<19:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▎     | 1441/3295 [14:50<19:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1442/3295 [14:52<19:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1442/3295 [14:52<19:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1443/3295 [14:52<19:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1443/3295 [14:52<19:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1444/3295 [14:52<19:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1444/3295 [14:52<19:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1445/3295 [14:53<19:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1445/3295 [14:53<19:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1446/3295 [14:53<19:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1446/3295 [14:53<19:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1447/3295 [14:54<19:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1447/3295 [14:54<19:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1448/3295 [14:55<19:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1448/3295 [14:55<19:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1449/3295 [14:55<19:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1449/3295 [14:55<19:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1450/3295 [14:55<19:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1450/3295 [14:55<19:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1451/3295 [14:57<19:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1451/3295 [14:57<19:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1452/3295 [14:57<18:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1452/3295 [14:57<18:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1453/3295 [14:57<18:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1453/3295 [14:57<18:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1454/3295 [14:59<18:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1454/3295 [14:59<18:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1455/3295 [14:59<18:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1455/3295 [14:59<18:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1456/3295 [14:59<18:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1456/3295 [14:59<18:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1457/3295 [15:01<18:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1457/3295 [15:01<18:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1458/3295 [15:01<18:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1458/3295 [15:01<18:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1459/3295 [15:01<18:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1459/3295 [15:01<18:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1460/3295 [15:03<18:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1460/3295 [15:03<18:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1461/3295 [15:03<18:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1461/3295 [15:03<18:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1462/3295 [15:03<18:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1462/3295 [15:03<18:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1463/3295 [15:05<18:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1463/3295 [15:05<18:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1464/3295 [15:05<18:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1464/3295 [15:05<18:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1465/3295 [15:05<18:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1465/3295 [15:05<18:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1466/3295 [15:07<18:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  44%|████▍     | 1466/3295 [15:07<18:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1467/3295 [15:07<18:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1467/3295 [15:07<18:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1468/3295 [15:07<18:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1468/3295 [15:07<18:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1469/3295 [15:09<18:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1469/3295 [15:09<18:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1470/3295 [15:09<18:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1470/3295 [15:09<18:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1471/3295 [15:09<18:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1471/3295 [15:09<18:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1472/3295 [15:10<18:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1472/3295 [15:10<18:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1473/3295 [15:10<18:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1473/3295 [15:10<18:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1474/3295 [15:10<18:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1474/3295 [15:10<18:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1475/3295 [15:12<18:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1475/3295 [15:12<18:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1476/3295 [15:12<18:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1476/3295 [15:12<18:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1477/3295 [15:12<18:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1477/3295 [15:12<18:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1478/3295 [15:14<18:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1478/3295 [15:14<18:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1479/3295 [15:14<18:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1479/3295 [15:14<18:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1480/3295 [15:14<18:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1480/3295 [15:14<18:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1481/3295 [15:16<18:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1481/3295 [15:16<18:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1482/3295 [15:16<18:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▍     | 1482/3295 [15:16<18:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1483/3295 [15:16<18:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1483/3295 [15:16<18:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1484/3295 [15:18<18:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1484/3295 [15:18<18:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1485/3295 [15:18<18:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1485/3295 [15:18<18:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1486/3295 [15:18<18:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1486/3295 [15:18<18:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1487/3295 [15:20<18:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1487/3295 [15:20<18:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1488/3295 [15:20<18:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1488/3295 [15:20<18:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1489/3295 [15:20<18:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1489/3295 [15:20<18:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1490/3295 [15:22<18:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1490/3295 [15:22<18:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1491/3295 [15:22<18:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1491/3295 [15:22<18:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1492/3295 [15:22<18:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1492/3295 [15:22<18:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1493/3295 [15:24<18:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1493/3295 [15:24<18:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1494/3295 [15:24<18:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1494/3295 [15:24<18:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1495/3295 [15:24<18:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1495/3295 [15:24<18:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1496/3295 [15:25<18:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1496/3295 [15:25<18:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1497/3295 [15:25<18:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1497/3295 [15:25<18:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1498/3295 [15:26<18:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1498/3295 [15:26<18:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1499/3295 [15:27<18:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  45%|████▌     | 1499/3295 [15:27<18:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1500/3295 [15:27<18:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1500/3295 [15:27<18:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1501/3295 [15:27<18:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1501/3295 [15:27<18:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1502/3295 [15:29<18:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1502/3295 [15:29<18:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1503/3295 [15:29<18:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1503/3295 [15:29<18:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1504/3295 [15:29<18:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1504/3295 [15:29<18:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1505/3295 [15:31<18:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1505/3295 [15:31<18:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1506/3295 [15:31<18:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1506/3295 [15:31<18:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1507/3295 [15:31<18:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1507/3295 [15:31<18:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1508/3295 [15:33<18:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1508/3295 [15:33<18:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1509/3295 [15:33<18:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1509/3295 [15:33<18:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1510/3295 [15:33<18:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1510/3295 [15:33<18:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1511/3295 [15:34<18:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1511/3295 [15:34<18:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1512/3295 [15:34<18:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1512/3295 [15:34<18:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1513/3295 [15:34<18:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1513/3295 [15:34<18:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1514/3295 [15:36<18:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1514/3295 [15:36<18:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1515/3295 [15:36<18:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1515/3295 [15:36<18:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1516/3295 [15:36<18:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1516/3295 [15:36<18:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1517/3295 [15:38<18:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1517/3295 [15:38<18:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1518/3295 [15:38<18:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1518/3295 [15:38<18:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1519/3295 [15:38<18:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1519/3295 [15:38<18:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1520/3295 [15:40<18:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1520/3295 [15:40<18:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1521/3295 [15:40<18:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1521/3295 [15:40<18:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1522/3295 [15:40<18:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1522/3295 [15:40<18:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1523/3295 [15:42<18:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▌     | 1523/3295 [15:42<18:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1524/3295 [15:42<18:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1524/3295 [15:42<18:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1525/3295 [15:42<18:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1525/3295 [15:42<18:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1526/3295 [15:44<18:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1526/3295 [15:44<18:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1527/3295 [15:44<18:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1527/3295 [15:44<18:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1528/3295 [15:44<18:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1528/3295 [15:44<18:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1529/3295 [15:46<18:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1529/3295 [15:46<18:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1530/3295 [15:46<18:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1530/3295 [15:46<18:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1531/3295 [15:46<18:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1531/3295 [15:46<18:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1532/3295 [15:47<18:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  46%|████▋     | 1532/3295 [15:47<18:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1533/3295 [15:47<18:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1533/3295 [15:47<18:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1534/3295 [15:47<18:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1534/3295 [15:47<18:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1535/3295 [15:49<18:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1535/3295 [15:49<18:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1536/3295 [15:49<18:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1536/3295 [15:49<18:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1537/3295 [15:49<18:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1537/3295 [15:49<18:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1538/3295 [15:51<18:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1538/3295 [15:51<18:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1539/3295 [15:51<18:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1539/3295 [15:51<18:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1540/3295 [15:51<18:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1540/3295 [15:51<18:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1541/3295 [15:53<18:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1541/3295 [15:53<18:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1542/3295 [15:53<18:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1542/3295 [15:53<18:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1543/3295 [15:53<18:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1543/3295 [15:53<18:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1544/3295 [15:55<18:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1544/3295 [15:55<18:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1545/3295 [15:55<18:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1545/3295 [15:55<18:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1546/3295 [15:55<18:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1546/3295 [15:55<18:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1547/3295 [15:56<18:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1547/3295 [15:56<18:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1548/3295 [15:56<17:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1548/3295 [15:56<17:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1549/3295 [15:56<17:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1549/3295 [15:56<17:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1550/3295 [15:59<17:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1550/3295 [15:59<17:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1551/3295 [15:59<17:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1551/3295 [15:59<17:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1552/3295 [15:59<17:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1552/3295 [15:59<17:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1553/3295 [16:00<17:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1553/3295 [16:00<17:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1554/3295 [16:00<17:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1554/3295 [16:00<17:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1555/3295 [16:00<17:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1555/3295 [16:00<17:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1556/3295 [16:02<17:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1556/3295 [16:02<17:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1557/3295 [16:02<17:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1557/3295 [16:02<17:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1558/3295 [16:02<17:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1558/3295 [16:02<17:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1559/3295 [16:04<17:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1559/3295 [16:04<17:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1560/3295 [16:04<17:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1560/3295 [16:04<17:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1561/3295 [16:04<17:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1561/3295 [16:04<17:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1562/3295 [16:06<17:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1562/3295 [16:06<17:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1563/3295 [16:06<17:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1563/3295 [16:06<17:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1564/3295 [16:06<17:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1564/3295 [16:06<17:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1565/3295 [16:08<17:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  47%|████▋     | 1565/3295 [16:08<17:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1566/3295 [16:08<17:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1566/3295 [16:08<17:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1567/3295 [16:08<17:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1567/3295 [16:08<17:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1568/3295 [16:10<17:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1568/3295 [16:10<17:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1569/3295 [16:10<17:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1569/3295 [16:10<17:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1570/3295 [16:10<17:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1570/3295 [16:10<17:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1571/3295 [16:12<17:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1571/3295 [16:12<17:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1572/3295 [16:12<17:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1572/3295 [16:12<17:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1573/3295 [16:12<17:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1573/3295 [16:12<17:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1574/3295 [16:14<17:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1574/3295 [16:14<17:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1575/3295 [16:14<17:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1575/3295 [16:14<17:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1576/3295 [16:14<17:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1576/3295 [16:14<17:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1577/3295 [16:16<17:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1577/3295 [16:16<17:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1578/3295 [16:16<17:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1578/3295 [16:16<17:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1579/3295 [16:16<17:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1579/3295 [16:16<17:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1580/3295 [16:18<17:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1580/3295 [16:18<17:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1581/3295 [16:18<17:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1581/3295 [16:18<17:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1582/3295 [16:18<17:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1582/3295 [16:18<17:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1583/3295 [16:20<17:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1583/3295 [16:20<17:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1584/3295 [16:20<17:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1584/3295 [16:20<17:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1585/3295 [16:20<17:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1585/3295 [16:20<17:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1586/3295 [16:21<17:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1586/3295 [16:21<17:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1587/3295 [16:22<17:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1587/3295 [16:22<17:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1588/3295 [16:22<17:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1588/3295 [16:22<17:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1589/3295 [16:23<17:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1589/3295 [16:23<17:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1590/3295 [16:23<17:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1590/3295 [16:23<17:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1591/3295 [16:23<17:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1591/3295 [16:23<17:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1592/3295 [16:25<17:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1592/3295 [16:25<17:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1593/3295 [16:25<17:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1593/3295 [16:25<17:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1594/3295 [16:25<17:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1594/3295 [16:25<17:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1595/3295 [16:27<17:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1595/3295 [16:27<17:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1596/3295 [16:27<17:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1596/3295 [16:27<17:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1597/3295 [16:27<17:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1597/3295 [16:27<17:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1598/3295 [16:29<17:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  48%|████▊     | 1598/3295 [16:29<17:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1599/3295 [16:29<17:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1599/3295 [16:29<17:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1600/3295 [16:29<17:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1600/3295 [16:29<17:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1601/3295 [16:31<17:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1601/3295 [16:31<17:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1602/3295 [16:31<17:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1602/3295 [16:31<17:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1603/3295 [16:31<17:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1603/3295 [16:31<17:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1604/3295 [16:32<17:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1604/3295 [16:32<17:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1605/3295 [16:32<17:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1605/3295 [16:32<17:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1606/3295 [16:32<17:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▊     | 1606/3295 [16:32<17:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1607/3295 [16:34<17:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1607/3295 [16:34<17:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1608/3295 [16:34<17:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1608/3295 [16:34<17:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1609/3295 [16:34<17:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1609/3295 [16:34<17:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1610/3295 [16:36<17:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1610/3295 [16:36<17:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1611/3295 [16:36<17:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1611/3295 [16:36<17:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1612/3295 [16:36<17:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1612/3295 [16:36<17:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1613/3295 [16:38<17:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1613/3295 [16:38<17:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1614/3295 [16:38<17:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1614/3295 [16:38<17:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1615/3295 [16:38<17:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1615/3295 [16:38<17:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1616/3295 [16:40<17:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1616/3295 [16:40<17:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1617/3295 [16:40<17:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1617/3295 [16:40<17:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1618/3295 [16:40<17:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1618/3295 [16:40<17:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1619/3295 [16:42<17:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1619/3295 [16:42<17:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1620/3295 [16:42<17:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1620/3295 [16:42<17:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1621/3295 [16:42<17:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1621/3295 [16:42<17:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1622/3295 [16:44<17:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1622/3295 [16:44<17:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1623/3295 [16:44<17:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1623/3295 [16:44<17:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1624/3295 [16:44<17:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1624/3295 [16:44<17:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1625/3295 [16:46<17:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1625/3295 [16:46<17:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1626/3295 [16:46<17:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1626/3295 [16:46<17:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1627/3295 [16:46<17:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1627/3295 [16:46<17:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1628/3295 [16:47<17:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1628/3295 [16:47<17:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1629/3295 [16:47<17:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1629/3295 [16:47<17:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1630/3295 [16:47<17:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1630/3295 [16:47<17:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1631/3295 [16:49<17:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  49%|████▉     | 1631/3295 [16:49<17:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1632/3295 [16:49<17:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1632/3295 [16:49<17:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1633/3295 [16:49<17:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1633/3295 [16:49<17:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1634/3295 [16:51<17:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1634/3295 [16:51<17:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1635/3295 [16:51<17:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1635/3295 [16:51<17:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1636/3295 [16:51<17:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1636/3295 [16:51<17:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1637/3295 [16:53<17:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1637/3295 [16:53<17:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1638/3295 [16:53<17:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1638/3295 [16:53<17:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1639/3295 [16:53<17:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1639/3295 [16:53<17:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1640/3295 [16:55<17:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1640/3295 [16:55<17:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1641/3295 [16:55<17:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1641/3295 [16:55<17:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1642/3295 [16:55<17:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1642/3295 [16:55<17:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1643/3295 [16:56<17:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1643/3295 [16:56<17:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1644/3295 [16:56<17:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1644/3295 [16:56<17:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1645/3295 [16:57<17:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1645/3295 [16:57<17:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1646/3295 [16:58<17:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1646/3295 [16:58<17:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1647/3295 [16:58<16:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|████▉     | 1647/3295 [16:58<16:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1648/3295 [16:58<16:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1648/3295 [16:58<16:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1649/3295 [17:00<16:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1649/3295 [17:00<16:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1650/3295 [17:00<16:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1650/3295 [17:00<16:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1651/3295 [17:00<16:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1651/3295 [17:00<16:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1652/3295 [17:02<16:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1652/3295 [17:02<16:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1653/3295 [17:02<16:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1653/3295 [17:02<16:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1654/3295 [17:02<16:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1654/3295 [17:02<16:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1655/3295 [17:04<16:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1655/3295 [17:04<16:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1656/3295 [17:04<16:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1656/3295 [17:04<16:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1657/3295 [17:04<16:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1657/3295 [17:04<16:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1658/3295 [17:06<16:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1658/3295 [17:06<16:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1659/3295 [17:06<16:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1659/3295 [17:06<16:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1660/3295 [17:06<16:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1660/3295 [17:06<16:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1661/3295 [17:08<16:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1661/3295 [17:08<16:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1662/3295 [17:08<16:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1662/3295 [17:08<16:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1663/3295 [17:08<16:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  50%|█████     | 1663/3295 [17:08<16:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1664/3295 [17:09<16:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1664/3295 [17:09<16:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1665/3295 [17:09<16:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1665/3295 [17:09<16:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1666/3295 [17:09<16:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1666/3295 [17:09<16:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1667/3295 [17:11<16:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1667/3295 [17:11<16:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1668/3295 [17:11<16:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1668/3295 [17:11<16:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1669/3295 [17:11<16:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1669/3295 [17:11<16:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1670/3295 [17:13<16:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1670/3295 [17:13<16:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1671/3295 [17:13<16:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1671/3295 [17:13<16:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1672/3295 [17:13<16:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1672/3295 [17:13<16:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1673/3295 [17:15<16:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1673/3295 [17:15<16:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1674/3295 [17:15<16:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1674/3295 [17:15<16:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1675/3295 [17:15<16:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1675/3295 [17:15<16:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1676/3295 [17:17<16:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1676/3295 [17:17<16:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1677/3295 [17:17<16:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1677/3295 [17:17<16:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1678/3295 [17:17<16:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1678/3295 [17:17<16:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1679/3295 [17:19<16:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1679/3295 [17:19<16:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1680/3295 [17:19<16:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1680/3295 [17:19<16:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1681/3295 [17:19<16:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1681/3295 [17:19<16:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1682/3295 [17:21<16:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1682/3295 [17:21<16:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1683/3295 [17:21<16:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1683/3295 [17:21<16:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1684/3295 [17:21<16:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1684/3295 [17:21<16:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1685/3295 [17:22<16:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1685/3295 [17:22<16:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1686/3295 [17:22<16:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1686/3295 [17:22<16:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1687/3295 [17:22<16:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1687/3295 [17:22<16:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1688/3295 [17:24<16:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████     | 1688/3295 [17:24<16:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1689/3295 [17:24<16:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1689/3295 [17:24<16:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1690/3295 [17:24<16:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1690/3295 [17:24<16:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1691/3295 [17:26<16:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1691/3295 [17:26<16:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1692/3295 [17:26<16:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1692/3295 [17:26<16:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1693/3295 [17:26<16:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1693/3295 [17:26<16:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1694/3295 [17:28<16:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1694/3295 [17:28<16:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1695/3295 [17:28<16:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1695/3295 [17:28<16:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1696/3295 [17:28<16:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  51%|█████▏    | 1696/3295 [17:28<16:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1697/3295 [17:29<16:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1697/3295 [17:29<16:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1698/3295 [17:29<16:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1698/3295 [17:29<16:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1699/3295 [17:29<16:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1699/3295 [17:29<16:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1700/3295 [17:32<16:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1700/3295 [17:32<16:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1701/3295 [17:32<16:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1701/3295 [17:32<16:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1702/3295 [17:32<16:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1702/3295 [17:32<16:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1703/3295 [17:33<16:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1703/3295 [17:33<16:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1704/3295 [17:34<16:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1704/3295 [17:34<16:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1705/3295 [17:34<16:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1705/3295 [17:34<16:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1706/3295 [17:35<16:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1706/3295 [17:35<16:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1707/3295 [17:35<16:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1707/3295 [17:35<16:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1708/3295 [17:35<16:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1708/3295 [17:35<16:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1709/3295 [17:37<16:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1709/3295 [17:37<16:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1710/3295 [17:38<16:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1710/3295 [17:38<16:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1711/3295 [17:38<16:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1711/3295 [17:38<16:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1712/3295 [17:39<16:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1712/3295 [17:39<16:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1713/3295 [17:39<16:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1713/3295 [17:39<16:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1714/3295 [17:39<16:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1714/3295 [17:39<16:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1715/3295 [17:41<16:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1715/3295 [17:41<16:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1716/3295 [17:41<16:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1716/3295 [17:41<16:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1717/3295 [17:41<16:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1717/3295 [17:41<16:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1718/3295 [17:43<16:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1718/3295 [17:43<16:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1719/3295 [17:43<16:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1719/3295 [17:43<16:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1720/3295 [17:43<16:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1720/3295 [17:43<16:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1721/3295 [17:45<16:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1721/3295 [17:45<16:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1722/3295 [17:45<16:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1722/3295 [17:45<16:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1723/3295 [17:45<16:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1723/3295 [17:45<16:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1724/3295 [17:47<16:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1724/3295 [17:47<16:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1725/3295 [17:47<16:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1725/3295 [17:47<16:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1726/3295 [17:47<16:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1726/3295 [17:47<16:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1727/3295 [17:49<16:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1727/3295 [17:49<16:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1728/3295 [17:49<16:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1728/3295 [17:49<16:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1729/3295 [17:49<16:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  52%|█████▏    | 1729/3295 [17:49<16:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1730/3295 [17:51<16:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1730/3295 [17:51<16:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1731/3295 [17:51<16:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1731/3295 [17:51<16:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1732/3295 [17:51<16:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1732/3295 [17:51<16:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1733/3295 [17:52<16:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1733/3295 [17:52<16:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1734/3295 [17:52<16:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1734/3295 [17:52<16:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1735/3295 [17:53<16:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1735/3295 [17:53<16:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1736/3295 [17:54<16:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1736/3295 [17:54<16:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1737/3295 [17:54<16:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1737/3295 [17:54<16:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1738/3295 [17:54<16:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1738/3295 [17:54<16:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1739/3295 [17:56<16:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1739/3295 [17:56<16:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1740/3295 [17:56<16:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1740/3295 [17:56<16:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1741/3295 [17:56<16:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1741/3295 [17:56<16:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1742/3295 [17:58<16:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1742/3295 [17:58<16:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1743/3295 [17:58<16:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1743/3295 [17:58<16:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1744/3295 [17:58<15:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1744/3295 [17:58<15:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1745/3295 [18:00<15:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1745/3295 [18:00<15:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1746/3295 [18:00<15:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1746/3295 [18:00<15:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1747/3295 [18:00<15:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1747/3295 [18:00<15:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1748/3295 [18:02<15:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1748/3295 [18:02<15:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1749/3295 [18:02<15:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1749/3295 [18:02<15:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1750/3295 [18:02<15:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1750/3295 [18:02<15:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1751/3295 [18:04<15:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1751/3295 [18:04<15:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1752/3295 [18:04<15:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1752/3295 [18:04<15:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1753/3295 [18:04<15:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1753/3295 [18:04<15:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1754/3295 [18:06<15:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1754/3295 [18:06<15:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1755/3295 [18:06<15:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1755/3295 [18:06<15:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1756/3295 [18:06<15:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1756/3295 [18:06<15:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1757/3295 [18:08<15:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1757/3295 [18:08<15:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1758/3295 [18:08<15:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1758/3295 [18:08<15:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1759/3295 [18:08<15:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1759/3295 [18:08<15:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1760/3295 [18:10<15:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1760/3295 [18:10<15:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1761/3295 [18:10<15:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1761/3295 [18:10<15:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1762/3295 [18:10<15:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  53%|█████▎    | 1762/3295 [18:10<15:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1763/3295 [18:11<15:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1763/3295 [18:11<15:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1764/3295 [18:11<15:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1764/3295 [18:11<15:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1765/3295 [18:11<15:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1765/3295 [18:11<15:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1766/3295 [18:13<15:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1766/3295 [18:13<15:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1767/3295 [18:13<15:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1767/3295 [18:13<15:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1768/3295 [18:13<15:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1768/3295 [18:13<15:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1769/3295 [18:15<15:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1769/3295 [18:15<15:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1770/3295 [18:15<15:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1770/3295 [18:15<15:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1771/3295 [18:15<15:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▎    | 1771/3295 [18:15<15:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1772/3295 [18:17<15:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1772/3295 [18:17<15:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1773/3295 [18:17<15:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1773/3295 [18:17<15:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1774/3295 [18:17<15:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1774/3295 [18:17<15:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1775/3295 [18:19<15:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1775/3295 [18:19<15:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1776/3295 [18:19<15:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1776/3295 [18:19<15:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1777/3295 [18:19<15:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1777/3295 [18:19<15:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1778/3295 [18:21<15:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1778/3295 [18:21<15:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1779/3295 [18:21<15:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1779/3295 [18:21<15:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1780/3295 [18:21<15:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1780/3295 [18:21<15:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1781/3295 [18:23<15:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1781/3295 [18:23<15:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1782/3295 [18:23<15:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1782/3295 [18:23<15:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1783/3295 [18:23<15:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1783/3295 [18:23<15:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1784/3295 [18:25<15:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1784/3295 [18:25<15:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1785/3295 [18:25<15:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1785/3295 [18:25<15:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1786/3295 [18:25<15:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1786/3295 [18:25<15:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1787/3295 [18:26<15:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1787/3295 [18:26<15:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1788/3295 [18:26<15:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1788/3295 [18:26<15:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1789/3295 [18:26<15:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1789/3295 [18:26<15:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1790/3295 [18:28<15:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1790/3295 [18:28<15:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1791/3295 [18:28<15:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1791/3295 [18:28<15:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1792/3295 [18:28<15:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1792/3295 [18:28<15:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1793/3295 [18:30<15:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1793/3295 [18:30<15:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1794/3295 [18:30<15:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1794/3295 [18:30<15:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1795/3295 [18:30<15:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  54%|█████▍    | 1795/3295 [18:30<15:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1796/3295 [18:32<15:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1796/3295 [18:32<15:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1797/3295 [18:32<15:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1797/3295 [18:32<15:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1798/3295 [18:32<15:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1798/3295 [18:32<15:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1799/3295 [18:34<15:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1799/3295 [18:34<15:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1800/3295 [18:34<15:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1800/3295 [18:34<15:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1801/3295 [18:34<15:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1801/3295 [18:34<15:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1802/3295 [18:36<15:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1802/3295 [18:36<15:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1803/3295 [18:36<15:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1803/3295 [18:36<15:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1804/3295 [18:36<15:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1804/3295 [18:36<15:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1805/3295 [18:38<15:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1805/3295 [18:38<15:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1806/3295 [18:38<15:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1806/3295 [18:38<15:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1807/3295 [18:38<15:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1807/3295 [18:38<15:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1808/3295 [18:39<15:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1808/3295 [18:39<15:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1809/3295 [18:39<15:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1809/3295 [18:39<15:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1810/3295 [18:39<15:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1810/3295 [18:39<15:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1811/3295 [18:41<15:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1811/3295 [18:41<15:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1812/3295 [18:41<15:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▍    | 1812/3295 [18:41<15:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1813/3295 [18:41<15:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1813/3295 [18:41<15:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1814/3295 [18:43<15:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1814/3295 [18:43<15:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1815/3295 [18:43<15:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1815/3295 [18:43<15:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1816/3295 [18:43<15:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1816/3295 [18:43<15:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1817/3295 [18:45<15:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1817/3295 [18:45<15:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1818/3295 [18:45<15:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1818/3295 [18:45<15:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1819/3295 [18:45<15:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1819/3295 [18:45<15:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1820/3295 [18:47<15:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1820/3295 [18:47<15:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1821/3295 [18:47<15:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1821/3295 [18:47<15:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1822/3295 [18:47<15:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1822/3295 [18:47<15:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1823/3295 [18:48<15:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1823/3295 [18:48<15:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1824/3295 [18:48<15:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1824/3295 [18:48<15:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1825/3295 [18:49<15:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1825/3295 [18:49<15:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1826/3295 [18:50<15:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1826/3295 [18:50<15:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1827/3295 [18:50<15:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1827/3295 [18:50<15:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1828/3295 [18:50<15:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  55%|█████▌    | 1828/3295 [18:50<15:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1829/3295 [18:52<15:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1829/3295 [18:52<15:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1830/3295 [18:52<15:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1830/3295 [18:52<15:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1831/3295 [18:52<15:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1831/3295 [18:52<15:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1832/3295 [18:54<15:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1832/3295 [18:54<15:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1833/3295 [18:54<15:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1833/3295 [18:54<15:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1834/3295 [18:54<15:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1834/3295 [18:54<15:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1835/3295 [18:56<15:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1835/3295 [18:56<15:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1836/3295 [18:56<15:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1836/3295 [18:56<15:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1837/3295 [18:56<15:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1837/3295 [18:56<15:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1838/3295 [18:58<15:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1838/3295 [18:58<15:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1839/3295 [18:58<15:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1839/3295 [18:58<15:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1840/3295 [18:58<15:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1840/3295 [18:58<15:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1841/3295 [19:00<15:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1841/3295 [19:00<15:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1842/3295 [19:00<14:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1842/3295 [19:00<14:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1843/3295 [19:00<14:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1843/3295 [19:00<14:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1844/3295 [19:02<14:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1844/3295 [19:02<14:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1845/3295 [19:02<14:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1845/3295 [19:02<14:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1846/3295 [19:02<14:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1846/3295 [19:02<14:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1847/3295 [19:03<14:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1847/3295 [19:03<14:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1848/3295 [19:03<14:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1848/3295 [19:03<14:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1849/3295 [19:03<14:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1849/3295 [19:03<14:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1850/3295 [19:05<14:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1850/3295 [19:05<14:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1851/3295 [19:05<14:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1851/3295 [19:05<14:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1852/3295 [19:05<14:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1852/3295 [19:05<14:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1853/3295 [19:07<14:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▌    | 1853/3295 [19:07<14:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1854/3295 [19:07<14:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1854/3295 [19:07<14:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1855/3295 [19:07<14:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1855/3295 [19:07<14:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1856/3295 [19:09<14:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1856/3295 [19:09<14:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1857/3295 [19:09<14:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1857/3295 [19:09<14:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1858/3295 [19:09<14:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1858/3295 [19:09<14:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1859/3295 [19:11<14:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1859/3295 [19:11<14:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1860/3295 [19:11<14:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1860/3295 [19:11<14:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1861/3295 [19:11<14:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  56%|█████▋    | 1861/3295 [19:11<14:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1862/3295 [19:13<14:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1862/3295 [19:13<14:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1863/3295 [19:13<14:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1863/3295 [19:13<14:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1864/3295 [19:13<14:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1864/3295 [19:13<14:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1865/3295 [19:15<14:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1865/3295 [19:15<14:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1866/3295 [19:15<14:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1866/3295 [19:15<14:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1867/3295 [19:15<14:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1867/3295 [19:15<14:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1868/3295 [19:17<14:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1868/3295 [19:17<14:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1869/3295 [19:17<14:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1869/3295 [19:17<14:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1870/3295 [19:17<14:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1870/3295 [19:17<14:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1871/3295 [19:18<14:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1871/3295 [19:18<14:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1872/3295 [19:18<14:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1872/3295 [19:18<14:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1873/3295 [19:18<14:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1873/3295 [19:18<14:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1874/3295 [19:20<14:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1874/3295 [19:20<14:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1875/3295 [19:20<14:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1875/3295 [19:20<14:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1876/3295 [19:20<14:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1876/3295 [19:20<14:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1877/3295 [19:22<14:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1877/3295 [19:22<14:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1878/3295 [19:22<14:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1878/3295 [19:22<14:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1879/3295 [19:22<14:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1879/3295 [19:22<14:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1880/3295 [19:24<14:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1880/3295 [19:24<14:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1881/3295 [19:24<14:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1881/3295 [19:24<14:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1882/3295 [19:24<14:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1882/3295 [19:24<14:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1883/3295 [19:26<14:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1883/3295 [19:26<14:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1884/3295 [19:26<14:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1884/3295 [19:26<14:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1885/3295 [19:26<14:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1885/3295 [19:26<14:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1886/3295 [19:28<14:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1886/3295 [19:28<14:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1887/3295 [19:28<14:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1887/3295 [19:28<14:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1888/3295 [19:28<14:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1888/3295 [19:28<14:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1889/3295 [19:29<14:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1889/3295 [19:29<14:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1890/3295 [19:29<14:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1890/3295 [19:29<14:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1891/3295 [19:29<14:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1891/3295 [19:29<14:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1892/3295 [19:31<14:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1892/3295 [19:31<14:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1893/3295 [19:31<14:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1893/3295 [19:31<14:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1894/3295 [19:31<14:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  57%|█████▋    | 1894/3295 [19:31<14:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1895/3295 [19:33<14:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1895/3295 [19:33<14:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1896/3295 [19:33<14:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1896/3295 [19:33<14:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1897/3295 [19:33<14:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1897/3295 [19:33<14:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1898/3295 [19:35<14:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1898/3295 [19:35<14:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1899/3295 [19:35<14:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1899/3295 [19:35<14:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1900/3295 [19:35<14:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1900/3295 [19:35<14:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1901/3295 [19:37<14:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1901/3295 [19:37<14:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1902/3295 [19:37<14:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1902/3295 [19:37<14:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1903/3295 [19:37<14:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1903/3295 [19:37<14:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1904/3295 [19:39<14:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1904/3295 [19:39<14:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1905/3295 [19:39<14:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1905/3295 [19:39<14:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1906/3295 [19:39<14:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1906/3295 [19:39<14:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1907/3295 [19:40<14:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1907/3295 [19:40<14:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1908/3295 [19:40<14:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1908/3295 [19:40<14:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1909/3295 [19:40<14:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1909/3295 [19:40<14:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1910/3295 [19:42<14:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1910/3295 [19:42<14:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1911/3295 [19:42<14:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1911/3295 [19:42<14:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1912/3295 [19:42<14:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1912/3295 [19:42<14:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1913/3295 [19:44<14:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1913/3295 [19:44<14:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1914/3295 [19:44<14:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1914/3295 [19:44<14:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1915/3295 [19:44<14:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1915/3295 [19:44<14:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1916/3295 [19:46<14:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1916/3295 [19:46<14:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1917/3295 [19:46<14:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1917/3295 [19:46<14:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1918/3295 [19:46<14:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1918/3295 [19:46<14:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1919/3295 [19:47<14:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1919/3295 [19:47<14:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1920/3295 [19:47<14:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1920/3295 [19:47<14:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1921/3295 [19:48<14:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1921/3295 [19:48<14:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1922/3295 [19:49<14:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1922/3295 [19:49<14:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1923/3295 [19:50<14:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1923/3295 [19:50<14:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1924/3295 [19:50<14:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1924/3295 [19:50<14:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1925/3295 [19:51<14:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1925/3295 [19:51<14:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1926/3295 [19:51<14:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1926/3295 [19:51<14:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1927/3295 [19:51<14:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  58%|█████▊    | 1927/3295 [19:51<14:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1928/3295 [19:53<14:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1928/3295 [19:53<14:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1929/3295 [19:53<14:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1929/3295 [19:53<14:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1930/3295 [19:53<14:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1930/3295 [19:53<14:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1931/3295 [19:55<14:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1931/3295 [19:55<14:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1932/3295 [19:55<14:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1932/3295 [19:55<14:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1933/3295 [19:55<14:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1933/3295 [19:55<14:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1934/3295 [19:57<14:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1934/3295 [19:57<14:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1935/3295 [19:57<14:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▊    | 1935/3295 [19:57<14:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1936/3295 [19:57<14:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1936/3295 [19:57<14:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1937/3295 [19:59<14:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1937/3295 [19:59<14:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1938/3295 [19:59<13:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1938/3295 [19:59<13:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1939/3295 [19:59<13:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1939/3295 [19:59<13:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1940/3295 [20:00<13:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1940/3295 [20:00<13:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1941/3295 [20:01<13:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1941/3295 [20:01<13:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1942/3295 [20:01<13:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1942/3295 [20:01<13:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1943/3295 [20:02<13:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1943/3295 [20:02<13:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1944/3295 [20:02<13:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1944/3295 [20:02<13:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1945/3295 [20:02<13:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1945/3295 [20:02<13:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1946/3295 [20:04<13:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1946/3295 [20:04<13:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1947/3295 [20:04<13:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1947/3295 [20:04<13:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1948/3295 [20:04<13:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1948/3295 [20:04<13:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1949/3295 [20:06<13:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1949/3295 [20:06<13:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1950/3295 [20:06<13:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1950/3295 [20:06<13:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1951/3295 [20:06<13:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1951/3295 [20:06<13:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1952/3295 [20:08<13:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1952/3295 [20:08<13:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1953/3295 [20:08<13:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1953/3295 [20:08<13:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1954/3295 [20:08<13:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1954/3295 [20:08<13:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1955/3295 [20:10<13:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1955/3295 [20:10<13:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1956/3295 [20:10<13:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1956/3295 [20:10<13:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1957/3295 [20:10<13:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1957/3295 [20:10<13:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1958/3295 [20:12<13:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1958/3295 [20:12<13:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1959/3295 [20:12<13:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1959/3295 [20:12<13:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1960/3295 [20:12<13:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  59%|█████▉    | 1960/3295 [20:12<13:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1961/3295 [20:14<13:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1961/3295 [20:14<13:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1962/3295 [20:14<13:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1962/3295 [20:14<13:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1963/3295 [20:14<13:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1963/3295 [20:14<13:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1964/3295 [20:16<13:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1964/3295 [20:16<13:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1965/3295 [20:16<13:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1965/3295 [20:16<13:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1966/3295 [20:16<13:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1966/3295 [20:16<13:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1967/3295 [20:18<13:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1967/3295 [20:18<13:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1968/3295 [20:18<13:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1968/3295 [20:18<13:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1969/3295 [20:18<13:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1969/3295 [20:18<13:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1970/3295 [20:19<13:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1970/3295 [20:19<13:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1971/3295 [20:19<13:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1971/3295 [20:19<13:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1972/3295 [20:19<13:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1972/3295 [20:19<13:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1973/3295 [20:21<13:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1973/3295 [20:21<13:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1974/3295 [20:21<13:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1974/3295 [20:21<13:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1975/3295 [20:21<13:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1975/3295 [20:21<13:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1976/3295 [20:23<13:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|█████▉    | 1976/3295 [20:23<13:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1977/3295 [20:23<13:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1977/3295 [20:23<13:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1978/3295 [20:23<13:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1978/3295 [20:23<13:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1979/3295 [20:25<13:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1979/3295 [20:25<13:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1980/3295 [20:25<13:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1980/3295 [20:25<13:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1981/3295 [20:25<13:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1981/3295 [20:25<13:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1982/3295 [20:27<13:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1982/3295 [20:27<13:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1983/3295 [20:27<13:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1983/3295 [20:27<13:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1984/3295 [20:27<13:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1984/3295 [20:27<13:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1985/3295 [20:29<13:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1985/3295 [20:29<13:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1986/3295 [20:29<13:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1986/3295 [20:29<13:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1987/3295 [20:29<13:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1987/3295 [20:29<13:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1988/3295 [20:31<13:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1988/3295 [20:31<13:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1989/3295 [20:31<13:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1989/3295 [20:31<13:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1990/3295 [20:31<13:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1990/3295 [20:31<13:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1991/3295 [20:33<13:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1991/3295 [20:33<13:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1992/3295 [20:33<13:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1992/3295 [20:33<13:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1993/3295 [20:33<13:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  60%|██████    | 1993/3295 [20:33<13:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1994/3295 [20:34<13:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1994/3295 [20:34<13:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1995/3295 [20:34<13:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1995/3295 [20:34<13:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1996/3295 [20:34<13:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1996/3295 [20:34<13:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1997/3295 [20:36<13:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1997/3295 [20:36<13:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1998/3295 [20:36<13:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1998/3295 [20:36<13:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1999/3295 [20:36<13:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 1999/3295 [20:36<13:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2000/3295 [20:38<13:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2000/3295 [20:38<13:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2001/3295 [20:38<13:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2001/3295 [20:38<13:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2002/3295 [20:38<13:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2002/3295 [20:38<13:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2003/3295 [20:40<13:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2003/3295 [20:40<13:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2004/3295 [20:40<13:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2004/3295 [20:40<13:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2005/3295 [20:40<13:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2005/3295 [20:40<13:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2006/3295 [20:42<13:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2006/3295 [20:42<13:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2007/3295 [20:42<13:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2007/3295 [20:42<13:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2008/3295 [20:42<13:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2008/3295 [20:42<13:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2009/3295 [20:44<13:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2009/3295 [20:44<13:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2010/3295 [20:44<13:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2010/3295 [20:44<13:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2011/3295 [20:44<13:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2011/3295 [20:44<13:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2012/3295 [20:46<13:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2012/3295 [20:46<13:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2013/3295 [20:46<13:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2013/3295 [20:46<13:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2014/3295 [20:46<13:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2014/3295 [20:46<13:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2015/3295 [20:47<13:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2015/3295 [20:47<13:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2016/3295 [20:47<13:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2016/3295 [20:47<13:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2017/3295 [20:47<13:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2017/3295 [20:47<13:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2018/3295 [20:49<13:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████    | 2018/3295 [20:49<13:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2019/3295 [20:49<13:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2019/3295 [20:49<13:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2020/3295 [20:49<13:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2020/3295 [20:49<13:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2021/3295 [20:51<13:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2021/3295 [20:51<13:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2022/3295 [20:51<13:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2022/3295 [20:51<13:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2023/3295 [20:51<13:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2023/3295 [20:51<13:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2024/3295 [20:53<13:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2024/3295 [20:53<13:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2025/3295 [20:53<13:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2025/3295 [20:53<13:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2026/3295 [20:53<13:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  61%|██████▏   | 2026/3295 [20:53<13:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2027/3295 [20:55<13:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2027/3295 [20:55<13:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2028/3295 [20:55<13:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2028/3295 [20:55<13:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2029/3295 [20:55<13:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2029/3295 [20:55<13:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2030/3295 [20:57<13:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2030/3295 [20:57<13:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2031/3295 [20:57<13:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2031/3295 [20:57<13:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2032/3295 [20:57<13:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2032/3295 [20:57<13:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2033/3295 [20:59<13:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2033/3295 [20:59<13:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2034/3295 [20:59<13:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2034/3295 [20:59<13:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2035/3295 [20:59<12:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2035/3295 [20:59<12:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2036/3295 [21:00<12:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2036/3295 [21:00<12:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2037/3295 [21:01<12:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2037/3295 [21:01<12:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2038/3295 [21:01<12:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2038/3295 [21:01<12:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2039/3295 [21:02<12:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2039/3295 [21:02<12:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2040/3295 [21:02<12:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2040/3295 [21:02<12:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2041/3295 [21:02<12:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2041/3295 [21:02<12:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2042/3295 [21:04<12:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2042/3295 [21:04<12:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2043/3295 [21:04<12:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2043/3295 [21:04<12:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2044/3295 [21:04<12:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2044/3295 [21:04<12:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2045/3295 [21:06<12:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2045/3295 [21:06<12:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2046/3295 [21:06<12:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2046/3295 [21:06<12:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2047/3295 [21:06<12:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2047/3295 [21:06<12:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2048/3295 [21:08<12:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2048/3295 [21:08<12:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2049/3295 [21:08<12:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2049/3295 [21:08<12:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2050/3295 [21:08<12:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2050/3295 [21:08<12:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2051/3295 [21:10<12:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2051/3295 [21:10<12:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2052/3295 [21:10<12:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2052/3295 [21:10<12:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2053/3295 [21:10<12:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2053/3295 [21:10<12:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2054/3295 [21:12<12:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2054/3295 [21:12<12:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2055/3295 [21:12<12:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2055/3295 [21:12<12:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2056/3295 [21:12<12:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2056/3295 [21:12<12:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2057/3295 [21:14<12:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2057/3295 [21:14<12:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2058/3295 [21:14<12:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2058/3295 [21:14<12:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2059/3295 [21:14<12:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  62%|██████▏   | 2059/3295 [21:14<12:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2060/3295 [21:16<12:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2060/3295 [21:16<12:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2061/3295 [21:16<12:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2061/3295 [21:16<12:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2062/3295 [21:16<12:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2062/3295 [21:16<12:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2063/3295 [21:18<12:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2063/3295 [21:18<12:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2064/3295 [21:18<12:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2064/3295 [21:18<12:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2065/3295 [21:18<12:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2065/3295 [21:18<12:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2066/3295 [21:19<12:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2066/3295 [21:19<12:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2067/3295 [21:19<12:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2067/3295 [21:19<12:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2068/3295 [21:19<12:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2068/3295 [21:19<12:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2069/3295 [21:21<12:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2069/3295 [21:21<12:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2070/3295 [21:21<12:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2070/3295 [21:21<12:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2071/3295 [21:21<12:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2071/3295 [21:21<12:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2072/3295 [21:23<12:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2072/3295 [21:23<12:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2073/3295 [21:23<12:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2073/3295 [21:23<12:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2074/3295 [21:23<12:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2074/3295 [21:23<12:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2075/3295 [21:25<12:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2075/3295 [21:25<12:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2076/3295 [21:25<12:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2076/3295 [21:25<12:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2077/3295 [21:25<12:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2077/3295 [21:25<12:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2078/3295 [21:27<12:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2078/3295 [21:27<12:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2079/3295 [21:27<12:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2079/3295 [21:27<12:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2080/3295 [21:27<12:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2080/3295 [21:27<12:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2081/3295 [21:29<12:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2081/3295 [21:29<12:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2082/3295 [21:29<12:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2082/3295 [21:29<12:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2083/3295 [21:29<12:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2083/3295 [21:29<12:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2084/3295 [21:31<12:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2084/3295 [21:31<12:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2085/3295 [21:31<12:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2085/3295 [21:31<12:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2086/3295 [21:31<12:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2086/3295 [21:31<12:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2087/3295 [21:33<12:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2087/3295 [21:33<12:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2088/3295 [21:33<12:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2088/3295 [21:33<12:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2089/3295 [21:33<12:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2089/3295 [21:33<12:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2090/3295 [21:35<12:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2090/3295 [21:35<12:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2091/3295 [21:35<12:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2091/3295 [21:35<12:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2092/3295 [21:35<12:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  63%|██████▎   | 2092/3295 [21:35<12:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2093/3295 [21:36<12:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2093/3295 [21:36<12:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2094/3295 [21:37<12:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2094/3295 [21:37<12:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2095/3295 [21:37<12:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2095/3295 [21:37<12:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2096/3295 [21:38<12:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2096/3295 [21:38<12:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2097/3295 [21:38<12:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2097/3295 [21:38<12:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2098/3295 [21:38<12:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2098/3295 [21:38<12:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2099/3295 [21:40<12:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2099/3295 [21:40<12:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2100/3295 [21:40<12:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▎   | 2100/3295 [21:40<12:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2101/3295 [21:40<12:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2101/3295 [21:40<12:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2102/3295 [21:42<12:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2102/3295 [21:42<12:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2103/3295 [21:42<12:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2103/3295 [21:42<12:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2104/3295 [21:42<12:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2104/3295 [21:42<12:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2105/3295 [21:44<12:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2105/3295 [21:44<12:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2106/3295 [21:44<12:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2106/3295 [21:44<12:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2107/3295 [21:44<12:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2107/3295 [21:44<12:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2108/3295 [21:46<12:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2108/3295 [21:46<12:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2109/3295 [21:46<12:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2109/3295 [21:46<12:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2110/3295 [21:46<12:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2110/3295 [21:46<12:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2111/3295 [21:47<12:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2111/3295 [21:47<12:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2112/3295 [21:47<12:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2112/3295 [21:47<12:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2113/3295 [21:47<12:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2113/3295 [21:47<12:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2114/3295 [21:49<12:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2114/3295 [21:49<12:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2115/3295 [21:49<12:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2115/3295 [21:49<12:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2116/3295 [21:49<12:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2116/3295 [21:49<12:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2117/3295 [21:51<12:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2117/3295 [21:51<12:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2118/3295 [21:51<12:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2118/3295 [21:51<12:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2119/3295 [21:51<12:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2119/3295 [21:51<12:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2120/3295 [21:53<12:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2120/3295 [21:53<12:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2121/3295 [21:53<12:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2121/3295 [21:53<12:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2122/3295 [21:53<12:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2122/3295 [21:53<12:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2123/3295 [21:55<12:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2123/3295 [21:55<12:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2124/3295 [21:55<12:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2124/3295 [21:55<12:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2125/3295 [21:55<12:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  64%|██████▍   | 2125/3295 [21:55<12:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2126/3295 [21:57<12:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2126/3295 [21:57<12:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2127/3295 [21:57<12:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2127/3295 [21:57<12:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2128/3295 [21:57<12:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2128/3295 [21:57<12:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2129/3295 [21:58<12:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2129/3295 [21:58<12:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2130/3295 [21:58<12:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2130/3295 [21:58<12:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2131/3295 [21:58<12:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2131/3295 [21:58<12:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2132/3295 [22:00<12:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2132/3295 [22:00<12:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2133/3295 [22:00<11:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2133/3295 [22:00<11:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2134/3295 [22:00<11:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2134/3295 [22:00<11:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2135/3295 [22:02<11:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2135/3295 [22:02<11:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2136/3295 [22:02<11:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2136/3295 [22:02<11:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2137/3295 [22:02<11:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2137/3295 [22:02<11:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2138/3295 [22:04<11:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2138/3295 [22:04<11:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2139/3295 [22:04<11:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2139/3295 [22:04<11:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2140/3295 [22:04<11:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2140/3295 [22:04<11:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2141/3295 [22:06<11:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▍   | 2141/3295 [22:06<11:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2142/3295 [22:06<11:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2142/3295 [22:06<11:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2143/3295 [22:06<11:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2143/3295 [22:06<11:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2144/3295 [22:08<11:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2144/3295 [22:08<11:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2145/3295 [22:08<11:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2145/3295 [22:08<11:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2146/3295 [22:08<11:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2146/3295 [22:08<11:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2147/3295 [22:09<11:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2147/3295 [22:09<11:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2148/3295 [22:09<11:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2148/3295 [22:09<11:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2149/3295 [22:10<11:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2149/3295 [22:10<11:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2150/3295 [22:11<11:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2150/3295 [22:11<11:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2151/3295 [22:11<11:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2151/3295 [22:11<11:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2152/3295 [22:11<11:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2152/3295 [22:11<11:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2153/3295 [22:13<11:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2153/3295 [22:13<11:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2154/3295 [22:13<11:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2154/3295 [22:13<11:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2155/3295 [22:13<11:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2155/3295 [22:13<11:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2156/3295 [22:15<11:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2156/3295 [22:15<11:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2157/3295 [22:15<11:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2157/3295 [22:15<11:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2158/3295 [22:15<11:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  65%|██████▌   | 2158/3295 [22:15<11:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2159/3295 [22:17<11:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2159/3295 [22:17<11:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2160/3295 [22:17<11:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2160/3295 [22:17<11:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2161/3295 [22:17<11:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2161/3295 [22:17<11:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2162/3295 [22:19<11:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2162/3295 [22:19<11:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2163/3295 [22:19<11:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2163/3295 [22:19<11:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2164/3295 [22:19<11:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2164/3295 [22:19<11:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2165/3295 [22:21<11:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2165/3295 [22:21<11:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2166/3295 [22:21<11:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2166/3295 [22:21<11:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2167/3295 [22:21<11:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2167/3295 [22:21<11:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2168/3295 [22:23<11:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2168/3295 [22:23<11:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2169/3295 [22:23<11:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2169/3295 [22:23<11:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2170/3295 [22:23<11:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2170/3295 [22:23<11:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2171/3295 [22:25<11:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2171/3295 [22:25<11:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2172/3295 [22:25<11:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2172/3295 [22:25<11:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2173/3295 [22:25<11:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2173/3295 [22:25<11:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2174/3295 [22:27<11:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2174/3295 [22:27<11:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2175/3295 [22:27<11:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2175/3295 [22:27<11:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2176/3295 [22:27<11:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2176/3295 [22:27<11:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2177/3295 [22:28<11:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2177/3295 [22:28<11:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2178/3295 [22:28<11:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2178/3295 [22:28<11:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2179/3295 [22:28<11:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2179/3295 [22:28<11:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2180/3295 [22:30<11:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2180/3295 [22:30<11:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2181/3295 [22:30<11:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2181/3295 [22:30<11:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2182/3295 [22:30<11:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▌   | 2182/3295 [22:30<11:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2183/3295 [22:32<11:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2183/3295 [22:32<11:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2184/3295 [22:32<11:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2184/3295 [22:32<11:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2185/3295 [22:32<11:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2185/3295 [22:32<11:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2186/3295 [22:34<11:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2186/3295 [22:34<11:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2187/3295 [22:34<11:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2187/3295 [22:34<11:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2188/3295 [22:34<11:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2188/3295 [22:34<11:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2189/3295 [22:36<11:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2189/3295 [22:36<11:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2190/3295 [22:36<11:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2190/3295 [22:36<11:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2191/3295 [22:36<11:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  66%|██████▋   | 2191/3295 [22:36<11:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2192/3295 [22:38<11:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2192/3295 [22:38<11:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2193/3295 [22:38<11:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2193/3295 [22:38<11:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2194/3295 [22:38<11:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2194/3295 [22:38<11:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2195/3295 [22:40<11:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2195/3295 [22:40<11:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2196/3295 [22:40<11:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2196/3295 [22:40<11:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2197/3295 [22:40<11:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2197/3295 [22:40<11:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2198/3295 [22:42<11:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2198/3295 [22:42<11:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2199/3295 [22:42<11:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2199/3295 [22:42<11:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2200/3295 [22:42<11:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2200/3295 [22:42<11:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2201/3295 [22:43<11:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2201/3295 [22:43<11:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2202/3295 [22:43<11:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2202/3295 [22:43<11:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2203/3295 [22:43<11:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2203/3295 [22:43<11:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2204/3295 [22:45<11:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2204/3295 [22:45<11:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2205/3295 [22:45<11:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2205/3295 [22:45<11:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2206/3295 [22:45<11:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2206/3295 [22:45<11:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2207/3295 [22:47<11:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2207/3295 [22:47<11:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2208/3295 [22:47<11:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2208/3295 [22:47<11:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2209/3295 [22:47<11:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2209/3295 [22:47<11:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2210/3295 [22:49<11:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2210/3295 [22:49<11:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2211/3295 [22:49<11:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2211/3295 [22:49<11:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2212/3295 [22:49<11:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2212/3295 [22:49<11:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2213/3295 [22:51<11:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2213/3295 [22:51<11:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2214/3295 [22:51<11:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2214/3295 [22:51<11:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2215/3295 [22:51<11:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2215/3295 [22:51<11:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2216/3295 [22:53<11:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2216/3295 [22:53<11:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2217/3295 [22:53<11:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2217/3295 [22:53<11:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2218/3295 [22:53<11:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2218/3295 [22:53<11:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2219/3295 [22:54<11:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2219/3295 [22:54<11:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2220/3295 [22:54<11:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2220/3295 [22:54<11:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2221/3295 [22:54<11:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2221/3295 [22:54<11:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2222/3295 [22:56<11:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2222/3295 [22:56<11:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2223/3295 [22:56<11:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2223/3295 [22:56<11:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2224/3295 [22:56<11:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  67%|██████▋   | 2224/3295 [22:56<11:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2225/3295 [22:58<11:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2225/3295 [22:58<11:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2226/3295 [22:58<11:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2226/3295 [22:58<11:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2227/3295 [22:58<11:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2227/3295 [22:58<11:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2228/3295 [23:00<11:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2228/3295 [23:00<11:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2229/3295 [23:00<11:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2229/3295 [23:00<11:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2230/3295 [23:00<10:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2230/3295 [23:00<10:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2231/3295 [23:02<10:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2231/3295 [23:02<10:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2232/3295 [23:02<10:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2232/3295 [23:02<10:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2233/3295 [23:02<10:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2233/3295 [23:02<10:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2234/3295 [23:04<10:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2234/3295 [23:04<10:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2235/3295 [23:04<10:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2235/3295 [23:04<10:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2236/3295 [23:04<10:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2236/3295 [23:04<10:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2237/3295 [23:06<10:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2237/3295 [23:06<10:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2238/3295 [23:06<10:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2238/3295 [23:06<10:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2239/3295 [23:06<10:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2239/3295 [23:06<10:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2240/3295 [23:07<10:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2240/3295 [23:07<10:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2241/3295 [23:07<10:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2241/3295 [23:07<10:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2242/3295 [23:07<10:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2242/3295 [23:07<10:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2243/3295 [23:09<10:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2243/3295 [23:09<10:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2244/3295 [23:09<10:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2244/3295 [23:09<10:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2245/3295 [23:09<10:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2245/3295 [23:09<10:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2246/3295 [23:11<10:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2246/3295 [23:11<10:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2247/3295 [23:11<10:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2247/3295 [23:11<10:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2248/3295 [23:11<10:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2248/3295 [23:11<10:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2249/3295 [23:13<10:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2249/3295 [23:13<10:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2250/3295 [23:13<10:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2250/3295 [23:13<10:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2251/3295 [23:13<10:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2251/3295 [23:13<10:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2252/3295 [23:15<10:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2252/3295 [23:15<10:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2253/3295 [23:15<10:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2253/3295 [23:15<10:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2254/3295 [23:15<10:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2254/3295 [23:15<10:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2255/3295 [23:16<10:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2255/3295 [23:16<10:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2256/3295 [23:17<10:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2256/3295 [23:17<10:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2257/3295 [23:17<10:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  68%|██████▊   | 2257/3295 [23:17<10:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2258/3295 [23:18<10:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2258/3295 [23:18<10:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2259/3295 [23:18<10:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2259/3295 [23:18<10:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2260/3295 [23:18<10:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2260/3295 [23:18<10:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2261/3295 [23:20<10:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2261/3295 [23:20<10:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2262/3295 [23:20<10:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2262/3295 [23:20<10:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2263/3295 [23:20<10:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2263/3295 [23:20<10:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2264/3295 [23:22<10:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2264/3295 [23:22<10:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2265/3295 [23:22<10:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▊   | 2265/3295 [23:22<10:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2266/3295 [23:22<10:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2266/3295 [23:22<10:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2267/3295 [23:24<10:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2267/3295 [23:24<10:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2268/3295 [23:24<10:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2268/3295 [23:24<10:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2269/3295 [23:24<10:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2269/3295 [23:24<10:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2270/3295 [23:26<10:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2270/3295 [23:26<10:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2271/3295 [23:26<10:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2271/3295 [23:26<10:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2272/3295 [23:26<10:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2272/3295 [23:26<10:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2273/3295 [23:27<10:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2273/3295 [23:27<10:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2274/3295 [23:27<10:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2274/3295 [23:27<10:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2275/3295 [23:27<10:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2275/3295 [23:27<10:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2276/3295 [23:29<10:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2276/3295 [23:29<10:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2277/3295 [23:29<10:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2277/3295 [23:29<10:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2278/3295 [23:29<10:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2278/3295 [23:29<10:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2279/3295 [23:31<10:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2279/3295 [23:31<10:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2280/3295 [23:31<10:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2280/3295 [23:31<10:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2281/3295 [23:31<10:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2281/3295 [23:31<10:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2282/3295 [23:33<10:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2282/3295 [23:33<10:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2283/3295 [23:33<10:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2283/3295 [23:33<10:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2284/3295 [23:33<10:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2284/3295 [23:33<10:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2285/3295 [23:35<10:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2285/3295 [23:35<10:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2286/3295 [23:35<10:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2286/3295 [23:35<10:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2287/3295 [23:35<10:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2287/3295 [23:35<10:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2288/3295 [23:37<10:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2288/3295 [23:37<10:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2289/3295 [23:37<10:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2289/3295 [23:37<10:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2290/3295 [23:37<10:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  69%|██████▉   | 2290/3295 [23:37<10:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2291/3295 [23:39<10:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2291/3295 [23:39<10:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2292/3295 [23:39<10:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2292/3295 [23:39<10:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2293/3295 [23:39<10:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2293/3295 [23:39<10:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2294/3295 [23:41<10:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2294/3295 [23:41<10:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2295/3295 [23:41<10:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2295/3295 [23:41<10:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2296/3295 [23:41<10:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2296/3295 [23:41<10:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2297/3295 [23:42<10:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2297/3295 [23:42<10:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2298/3295 [23:42<10:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2298/3295 [23:42<10:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2299/3295 [23:42<10:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2299/3295 [23:42<10:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2300/3295 [23:44<10:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2300/3295 [23:44<10:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2301/3295 [23:44<10:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2301/3295 [23:44<10:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2302/3295 [23:44<10:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2302/3295 [23:44<10:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2303/3295 [23:46<10:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2303/3295 [23:46<10:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2304/3295 [23:46<10:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2304/3295 [23:46<10:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2305/3295 [23:46<10:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2305/3295 [23:46<10:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2306/3295 [23:48<10:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|██████▉   | 2306/3295 [23:48<10:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2307/3295 [23:48<10:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2307/3295 [23:48<10:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2308/3295 [23:48<10:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2308/3295 [23:48<10:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2309/3295 [23:50<10:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2309/3295 [23:50<10:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2310/3295 [23:50<10:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2310/3295 [23:50<10:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2311/3295 [23:50<10:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2311/3295 [23:50<10:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2312/3295 [23:52<10:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2312/3295 [23:52<10:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2313/3295 [23:52<10:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2313/3295 [23:52<10:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2314/3295 [23:52<10:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2314/3295 [23:52<10:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2315/3295 [23:54<10:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2315/3295 [23:54<10:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2316/3295 [23:54<10:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2316/3295 [23:54<10:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2317/3295 [23:54<10:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2317/3295 [23:54<10:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2318/3295 [23:55<10:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2318/3295 [23:55<10:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2319/3295 [23:55<10:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2319/3295 [23:55<10:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2320/3295 [23:55<10:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2320/3295 [23:55<10:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2321/3295 [23:57<10:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2321/3295 [23:57<10:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2322/3295 [23:57<10:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  70%|███████   | 2322/3295 [23:57<10:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2323/3295 [23:57<10:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2323/3295 [23:57<10:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2324/3295 [23:59<10:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2324/3295 [23:59<10:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2325/3295 [23:59<10:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2325/3295 [23:59<10:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2326/3295 [23:59<09:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2326/3295 [23:59<09:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2327/3295 [24:01<09:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2327/3295 [24:01<09:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2328/3295 [24:01<09:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2328/3295 [24:01<09:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2329/3295 [24:01<09:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2329/3295 [24:01<09:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2330/3295 [24:03<09:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2330/3295 [24:03<09:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2331/3295 [24:03<09:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2331/3295 [24:03<09:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2332/3295 [24:03<09:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2332/3295 [24:03<09:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2333/3295 [24:05<09:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2333/3295 [24:05<09:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2334/3295 [24:05<09:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2334/3295 [24:05<09:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2335/3295 [24:05<09:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2335/3295 [24:05<09:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2336/3295 [24:07<09:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2336/3295 [24:07<09:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2337/3295 [24:07<09:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2337/3295 [24:07<09:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2338/3295 [24:07<09:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2338/3295 [24:07<09:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2339/3295 [24:08<09:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2339/3295 [24:08<09:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2340/3295 [24:09<09:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2340/3295 [24:09<09:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2341/3295 [24:09<09:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2341/3295 [24:09<09:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2342/3295 [24:10<09:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2342/3295 [24:10<09:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2343/3295 [24:10<09:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2343/3295 [24:10<09:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2344/3295 [24:10<09:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2344/3295 [24:10<09:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2345/3295 [24:12<09:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2345/3295 [24:12<09:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2346/3295 [24:12<09:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2346/3295 [24:12<09:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2347/3295 [24:12<09:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████   | 2347/3295 [24:12<09:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2348/3295 [24:14<09:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2348/3295 [24:14<09:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2349/3295 [24:14<09:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2349/3295 [24:14<09:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2350/3295 [24:14<09:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2350/3295 [24:14<09:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2351/3295 [24:16<09:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2351/3295 [24:16<09:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2352/3295 [24:16<09:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2352/3295 [24:16<09:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2353/3295 [24:16<09:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2353/3295 [24:16<09:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2354/3295 [24:18<09:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2354/3295 [24:18<09:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2355/3295 [24:18<09:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  71%|███████▏  | 2355/3295 [24:18<09:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2356/3295 [24:18<09:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2356/3295 [24:18<09:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2357/3295 [24:20<09:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2357/3295 [24:20<09:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2358/3295 [24:20<09:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2358/3295 [24:20<09:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2359/3295 [24:20<09:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2359/3295 [24:20<09:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2360/3295 [24:22<09:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2360/3295 [24:22<09:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2361/3295 [24:22<09:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2361/3295 [24:22<09:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2362/3295 [24:22<09:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2362/3295 [24:22<09:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2363/3295 [24:23<09:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2363/3295 [24:23<09:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2364/3295 [24:24<09:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2364/3295 [24:24<09:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2365/3295 [24:24<09:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2365/3295 [24:24<09:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2366/3295 [24:25<09:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2366/3295 [24:25<09:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2367/3295 [24:26<09:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2367/3295 [24:26<09:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2368/3295 [24:26<09:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2368/3295 [24:26<09:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2369/3295 [24:27<09:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2369/3295 [24:27<09:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2370/3295 [24:27<09:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2370/3295 [24:27<09:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2371/3295 [24:27<09:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2371/3295 [24:27<09:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2372/3295 [24:29<09:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2372/3295 [24:29<09:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2373/3295 [24:29<09:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2373/3295 [24:29<09:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2374/3295 [24:29<09:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2374/3295 [24:29<09:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2375/3295 [24:31<09:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2375/3295 [24:31<09:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2376/3295 [24:31<09:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2376/3295 [24:31<09:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2377/3295 [24:31<09:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2377/3295 [24:31<09:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2378/3295 [24:33<09:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2378/3295 [24:33<09:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2379/3295 [24:33<09:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2379/3295 [24:33<09:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2380/3295 [24:33<09:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2380/3295 [24:33<09:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2381/3295 [24:35<09:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2381/3295 [24:35<09:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2382/3295 [24:35<09:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2382/3295 [24:35<09:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2383/3295 [24:35<09:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2383/3295 [24:35<09:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2384/3295 [24:37<09:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2384/3295 [24:37<09:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2385/3295 [24:37<09:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2385/3295 [24:37<09:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2386/3295 [24:37<09:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2386/3295 [24:37<09:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2387/3295 [24:38<09:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2387/3295 [24:38<09:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2388/3295 [24:38<09:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  72%|███████▏  | 2388/3295 [24:38<09:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2389/3295 [24:38<09:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2389/3295 [24:38<09:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2390/3295 [24:40<09:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2390/3295 [24:40<09:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2391/3295 [24:40<09:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2391/3295 [24:40<09:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2392/3295 [24:40<09:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2392/3295 [24:40<09:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2393/3295 [24:42<09:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2393/3295 [24:42<09:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2394/3295 [24:42<09:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2394/3295 [24:42<09:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2395/3295 [24:42<09:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2395/3295 [24:42<09:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2396/3295 [24:44<09:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2396/3295 [24:44<09:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2397/3295 [24:44<09:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2397/3295 [24:44<09:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2398/3295 [24:44<09:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2398/3295 [24:44<09:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2399/3295 [24:46<09:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2399/3295 [24:46<09:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2400/3295 [24:46<09:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2400/3295 [24:46<09:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2401/3295 [24:46<09:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2401/3295 [24:46<09:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2402/3295 [24:47<09:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2402/3295 [24:47<09:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2403/3295 [24:47<09:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2403/3295 [24:47<09:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2404/3295 [24:47<09:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2404/3295 [24:47<09:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2405/3295 [24:49<09:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2405/3295 [24:49<09:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2406/3295 [24:49<09:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2406/3295 [24:49<09:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2407/3295 [24:49<09:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2407/3295 [24:49<09:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2408/3295 [24:51<09:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2408/3295 [24:51<09:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2409/3295 [24:51<09:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2409/3295 [24:51<09:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2410/3295 [24:51<09:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2410/3295 [24:51<09:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2411/3295 [24:53<09:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2411/3295 [24:53<09:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2412/3295 [24:53<09:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2412/3295 [24:53<09:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2413/3295 [24:53<09:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2413/3295 [24:53<09:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2414/3295 [24:55<09:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2414/3295 [24:55<09:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2415/3295 [24:55<09:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2415/3295 [24:55<09:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2416/3295 [24:55<09:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2416/3295 [24:55<09:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2417/3295 [24:57<09:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2417/3295 [24:57<09:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2418/3295 [24:57<09:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2418/3295 [24:57<09:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2419/3295 [24:57<09:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2419/3295 [24:57<09:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2420/3295 [24:59<09:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2420/3295 [24:59<09:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2421/3295 [24:59<09:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  73%|███████▎  | 2421/3295 [24:59<09:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2422/3295 [24:59<09:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2422/3295 [24:59<09:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2423/3295 [25:00<09:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2423/3295 [25:00<09:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2424/3295 [25:00<08:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2424/3295 [25:00<08:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2425/3295 [25:00<08:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2425/3295 [25:00<08:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2426/3295 [25:02<08:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2426/3295 [25:02<08:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2427/3295 [25:02<08:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2427/3295 [25:02<08:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2428/3295 [25:02<08:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2428/3295 [25:02<08:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2429/3295 [25:04<08:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2429/3295 [25:04<08:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2430/3295 [25:04<08:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▎  | 2430/3295 [25:04<08:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2431/3295 [25:04<08:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2431/3295 [25:04<08:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2432/3295 [25:06<08:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2432/3295 [25:06<08:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2433/3295 [25:06<08:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2433/3295 [25:06<08:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2434/3295 [25:06<08:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2434/3295 [25:06<08:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2435/3295 [25:08<08:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2435/3295 [25:08<08:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2436/3295 [25:08<08:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2436/3295 [25:08<08:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2437/3295 [25:08<08:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2437/3295 [25:08<08:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2438/3295 [25:10<08:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2438/3295 [25:10<08:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2439/3295 [25:10<08:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2439/3295 [25:10<08:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2440/3295 [25:10<08:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2440/3295 [25:10<08:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2441/3295 [25:12<08:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2441/3295 [25:12<08:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2442/3295 [25:12<08:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2442/3295 [25:12<08:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2443/3295 [25:12<08:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2443/3295 [25:12<08:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2444/3295 [25:14<08:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2444/3295 [25:14<08:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2445/3295 [25:14<08:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2445/3295 [25:14<08:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2446/3295 [25:14<08:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2446/3295 [25:14<08:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2447/3295 [25:15<08:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2447/3295 [25:15<08:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2448/3295 [25:16<08:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2448/3295 [25:16<08:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2449/3295 [25:16<08:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2449/3295 [25:16<08:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2450/3295 [25:17<08:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2450/3295 [25:17<08:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2451/3295 [25:17<08:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2451/3295 [25:17<08:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2452/3295 [25:17<08:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2452/3295 [25:17<08:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2453/3295 [25:19<08:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2453/3295 [25:19<08:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2454/3295 [25:19<08:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  74%|███████▍  | 2454/3295 [25:19<08:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2455/3295 [25:19<08:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2455/3295 [25:19<08:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2456/3295 [25:21<08:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2456/3295 [25:21<08:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2457/3295 [25:21<08:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2457/3295 [25:21<08:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2458/3295 [25:21<08:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2458/3295 [25:21<08:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2459/3295 [25:23<08:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2459/3295 [25:23<08:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2460/3295 [25:23<08:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2460/3295 [25:23<08:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2461/3295 [25:23<08:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2461/3295 [25:23<08:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2462/3295 [25:24<08:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2462/3295 [25:24<08:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2463/3295 [25:24<08:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2463/3295 [25:24<08:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2464/3295 [25:24<08:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2464/3295 [25:24<08:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2465/3295 [25:26<08:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2465/3295 [25:26<08:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2466/3295 [25:26<08:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2466/3295 [25:26<08:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2467/3295 [25:26<08:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2467/3295 [25:26<08:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2468/3295 [25:28<08:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2468/3295 [25:28<08:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2469/3295 [25:28<08:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2469/3295 [25:28<08:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2470/3295 [25:28<08:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2470/3295 [25:28<08:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2471/3295 [25:30<08:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▍  | 2471/3295 [25:30<08:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2472/3295 [25:30<08:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2472/3295 [25:30<08:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2473/3295 [25:30<08:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2473/3295 [25:30<08:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2474/3295 [25:32<08:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2474/3295 [25:32<08:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2475/3295 [25:32<08:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2475/3295 [25:32<08:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2476/3295 [25:32<08:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2476/3295 [25:32<08:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2477/3295 [25:33<08:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2477/3295 [25:33<08:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2478/3295 [25:33<08:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2478/3295 [25:33<08:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2479/3295 [25:34<08:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2479/3295 [25:34<08:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2480/3295 [25:35<08:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2480/3295 [25:35<08:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2481/3295 [25:35<08:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2481/3295 [25:35<08:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2482/3295 [25:35<08:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2482/3295 [25:35<08:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2483/3295 [25:37<08:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2483/3295 [25:37<08:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2484/3295 [25:37<08:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2484/3295 [25:37<08:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2485/3295 [25:37<08:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2485/3295 [25:37<08:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2486/3295 [25:39<08:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2486/3295 [25:39<08:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2487/3295 [25:39<08:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  75%|███████▌  | 2487/3295 [25:39<08:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2488/3295 [25:39<08:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2488/3295 [25:39<08:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2489/3295 [25:41<08:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2489/3295 [25:41<08:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2490/3295 [25:41<08:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2490/3295 [25:41<08:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2491/3295 [25:41<08:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2491/3295 [25:41<08:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2492/3295 [25:43<08:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2492/3295 [25:43<08:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2493/3295 [25:43<08:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2493/3295 [25:43<08:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2494/3295 [25:43<08:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2494/3295 [25:43<08:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2495/3295 [25:44<08:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2495/3295 [25:44<08:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2496/3295 [25:45<08:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2496/3295 [25:45<08:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2497/3295 [25:45<08:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2497/3295 [25:45<08:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2498/3295 [25:46<08:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2498/3295 [25:46<08:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2499/3295 [25:47<08:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2499/3295 [25:47<08:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2500/3295 [25:47<08:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2500/3295 [25:47<08:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2501/3295 [25:48<08:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2501/3295 [25:48<08:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2502/3295 [25:49<08:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2502/3295 [25:49<08:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2503/3295 [25:49<08:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2503/3295 [25:49<08:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2504/3295 [25:50<08:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2504/3295 [25:50<08:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2505/3295 [25:50<08:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2505/3295 [25:50<08:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2506/3295 [25:51<08:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2506/3295 [25:51<08:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2507/3295 [25:52<08:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2507/3295 [25:52<08:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2508/3295 [25:52<08:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2508/3295 [25:52<08:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2509/3295 [25:52<08:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2509/3295 [25:52<08:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2510/3295 [25:54<08:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2510/3295 [25:54<08:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2511/3295 [25:54<08:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2511/3295 [25:54<08:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2512/3295 [25:54<08:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▌  | 2512/3295 [25:54<08:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2513/3295 [25:56<08:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2513/3295 [25:56<08:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2514/3295 [25:56<08:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2514/3295 [25:56<08:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2515/3295 [25:56<08:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2515/3295 [25:56<08:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2516/3295 [25:58<08:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2516/3295 [25:58<08:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2517/3295 [25:58<08:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2517/3295 [25:58<08:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2518/3295 [25:58<08:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2518/3295 [25:58<08:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2519/3295 [25:59<08:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2519/3295 [25:59<08:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2520/3295 [26:00<07:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  76%|███████▋  | 2520/3295 [26:00<07:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2521/3295 [26:00<07:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2521/3295 [26:00<07:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2522/3295 [26:01<07:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2522/3295 [26:01<07:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2523/3295 [26:02<07:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2523/3295 [26:02<07:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2524/3295 [26:02<07:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2524/3295 [26:02<07:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2525/3295 [26:03<07:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2525/3295 [26:03<07:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2526/3295 [26:04<07:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2526/3295 [26:04<07:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2527/3295 [26:04<07:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2527/3295 [26:04<07:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2528/3295 [26:05<07:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2528/3295 [26:05<07:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2529/3295 [26:05<07:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2529/3295 [26:05<07:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2530/3295 [26:05<07:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2530/3295 [26:05<07:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2531/3295 [26:07<07:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2531/3295 [26:07<07:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2532/3295 [26:07<07:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2532/3295 [26:07<07:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2533/3295 [26:07<07:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2533/3295 [26:07<07:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2534/3295 [26:09<07:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2534/3295 [26:09<07:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2535/3295 [26:09<07:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2535/3295 [26:09<07:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2536/3295 [26:09<07:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2536/3295 [26:09<07:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2537/3295 [26:11<07:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2537/3295 [26:11<07:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2538/3295 [26:11<07:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2538/3295 [26:11<07:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2539/3295 [26:11<07:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2539/3295 [26:11<07:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2540/3295 [26:12<07:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2540/3295 [26:12<07:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2541/3295 [26:13<07:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2541/3295 [26:13<07:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2542/3295 [26:13<07:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2542/3295 [26:13<07:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2543/3295 [26:14<07:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2543/3295 [26:14<07:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2544/3295 [26:15<07:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2544/3295 [26:15<07:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2545/3295 [26:15<07:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2545/3295 [26:15<07:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2546/3295 [26:16<07:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2546/3295 [26:16<07:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2547/3295 [26:16<07:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2547/3295 [26:16<07:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2548/3295 [26:16<07:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2548/3295 [26:16<07:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2549/3295 [26:18<07:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2549/3295 [26:18<07:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2550/3295 [26:18<07:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2550/3295 [26:18<07:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2551/3295 [26:18<07:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2551/3295 [26:18<07:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2552/3295 [26:20<07:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2552/3295 [26:20<07:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2553/3295 [26:20<07:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  77%|███████▋  | 2553/3295 [26:20<07:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2554/3295 [26:20<07:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2554/3295 [26:20<07:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2555/3295 [26:21<07:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2555/3295 [26:21<07:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2556/3295 [26:22<07:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2556/3295 [26:22<07:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2557/3295 [26:22<07:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2557/3295 [26:22<07:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2558/3295 [26:23<07:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2558/3295 [26:23<07:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2559/3295 [26:23<07:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2559/3295 [26:23<07:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2560/3295 [26:23<07:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2560/3295 [26:23<07:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2561/3295 [26:25<07:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2561/3295 [26:25<07:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2562/3295 [26:25<07:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2562/3295 [26:25<07:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2563/3295 [26:25<07:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2563/3295 [26:25<07:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2564/3295 [26:27<07:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2564/3295 [26:27<07:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2565/3295 [26:27<07:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2565/3295 [26:27<07:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2566/3295 [26:27<07:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2566/3295 [26:27<07:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2567/3295 [26:29<07:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2567/3295 [26:29<07:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2568/3295 [26:29<07:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2568/3295 [26:29<07:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2569/3295 [26:29<07:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2569/3295 [26:29<07:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2570/3295 [26:30<07:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2570/3295 [26:30<07:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2571/3295 [26:31<07:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2571/3295 [26:31<07:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2572/3295 [26:31<07:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2572/3295 [26:31<07:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2573/3295 [26:33<07:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2573/3295 [26:33<07:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2574/3295 [26:33<07:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2574/3295 [26:33<07:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2575/3295 [26:33<07:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2575/3295 [26:33<07:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2576/3295 [26:34<07:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2576/3295 [26:34<07:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2577/3295 [26:34<07:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2577/3295 [26:34<07:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2578/3295 [26:34<07:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2578/3295 [26:34<07:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2579/3295 [26:36<07:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2579/3295 [26:36<07:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2580/3295 [26:36<07:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2580/3295 [26:36<07:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2581/3295 [26:36<07:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2581/3295 [26:36<07:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2582/3295 [26:38<07:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2582/3295 [26:38<07:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2583/3295 [26:38<07:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2583/3295 [26:38<07:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2584/3295 [26:38<07:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2584/3295 [26:38<07:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2585/3295 [26:40<07:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2585/3295 [26:40<07:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2586/3295 [26:40<07:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  78%|███████▊  | 2586/3295 [26:40<07:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2587/3295 [26:40<07:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2587/3295 [26:40<07:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2588/3295 [26:42<07:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2588/3295 [26:42<07:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2589/3295 [26:42<07:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2589/3295 [26:42<07:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2590/3295 [26:42<07:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2590/3295 [26:42<07:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2591/3295 [26:44<07:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2591/3295 [26:44<07:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2592/3295 [26:44<07:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2592/3295 [26:44<07:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2593/3295 [26:44<07:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2593/3295 [26:44<07:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2594/3295 [26:46<07:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▊  | 2594/3295 [26:46<07:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2595/3295 [26:46<07:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2595/3295 [26:46<07:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2596/3295 [26:46<07:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2596/3295 [26:46<07:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2597/3295 [26:48<07:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2597/3295 [26:48<07:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2598/3295 [26:48<07:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2598/3295 [26:48<07:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2599/3295 [26:48<07:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2599/3295 [26:48<07:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2600/3295 [26:49<07:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2600/3295 [26:49<07:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2601/3295 [26:49<07:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2601/3295 [26:49<07:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2602/3295 [26:49<07:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2602/3295 [26:49<07:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2603/3295 [26:51<07:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2603/3295 [26:51<07:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2604/3295 [26:51<07:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2604/3295 [26:51<07:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2605/3295 [26:51<07:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2605/3295 [26:51<07:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2606/3295 [26:53<07:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2606/3295 [26:53<07:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2607/3295 [26:53<07:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2607/3295 [26:53<07:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2608/3295 [26:53<07:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2608/3295 [26:53<07:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2609/3295 [26:55<07:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2609/3295 [26:55<07:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2610/3295 [26:55<07:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2610/3295 [26:55<07:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2611/3295 [26:55<07:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2611/3295 [26:55<07:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2612/3295 [26:57<07:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2612/3295 [26:57<07:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2613/3295 [26:57<07:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2613/3295 [26:57<07:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2614/3295 [26:57<07:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2614/3295 [26:57<07:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2615/3295 [26:59<07:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2615/3295 [26:59<07:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2616/3295 [26:59<07:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2616/3295 [26:59<07:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2617/3295 [26:59<06:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2617/3295 [26:59<06:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2618/3295 [27:01<06:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2618/3295 [27:01<06:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2619/3295 [27:01<06:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  79%|███████▉  | 2619/3295 [27:01<06:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2620/3295 [27:01<06:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2620/3295 [27:01<06:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2621/3295 [27:03<06:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2621/3295 [27:03<06:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2622/3295 [27:03<06:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2622/3295 [27:03<06:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2623/3295 [27:03<06:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2623/3295 [27:03<06:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2624/3295 [27:04<06:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2624/3295 [27:04<06:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2625/3295 [27:05<06:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2625/3295 [27:05<06:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2626/3295 [27:05<06:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2626/3295 [27:05<06:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2627/3295 [27:06<06:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2627/3295 [27:06<06:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2628/3295 [27:06<06:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2628/3295 [27:06<06:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2629/3295 [27:06<06:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2629/3295 [27:06<06:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2630/3295 [27:08<06:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2630/3295 [27:08<06:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2631/3295 [27:08<06:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2631/3295 [27:08<06:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2632/3295 [27:08<06:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2632/3295 [27:08<06:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2633/3295 [27:10<06:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2633/3295 [27:10<06:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2634/3295 [27:10<06:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2634/3295 [27:10<06:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2635/3295 [27:10<06:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|███████▉  | 2635/3295 [27:10<06:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2636/3295 [27:12<06:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2636/3295 [27:12<06:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2637/3295 [27:12<06:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2637/3295 [27:12<06:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2638/3295 [27:12<06:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2638/3295 [27:12<06:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2639/3295 [27:14<06:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2639/3295 [27:14<06:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2640/3295 [27:14<06:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2640/3295 [27:14<06:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2641/3295 [27:14<06:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2641/3295 [27:14<06:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2642/3295 [27:15<06:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2642/3295 [27:15<06:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2643/3295 [27:16<06:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2643/3295 [27:16<06:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2644/3295 [27:16<06:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2644/3295 [27:16<06:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2645/3295 [27:17<06:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2645/3295 [27:17<06:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2646/3295 [27:17<06:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2646/3295 [27:17<06:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2647/3295 [27:17<06:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2647/3295 [27:17<06:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2648/3295 [27:19<06:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2648/3295 [27:19<06:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2649/3295 [27:19<06:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2649/3295 [27:19<06:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2650/3295 [27:19<06:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2650/3295 [27:19<06:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2651/3295 [27:21<06:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2651/3295 [27:21<06:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2652/3295 [27:21<06:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  80%|████████  | 2652/3295 [27:21<06:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2653/3295 [27:21<06:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2653/3295 [27:21<06:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2654/3295 [27:23<06:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2654/3295 [27:23<06:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2655/3295 [27:23<06:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2655/3295 [27:23<06:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2656/3295 [27:23<06:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2656/3295 [27:23<06:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2657/3295 [27:25<06:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2657/3295 [27:25<06:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2658/3295 [27:25<06:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2658/3295 [27:25<06:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2659/3295 [27:25<06:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2659/3295 [27:25<06:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2660/3295 [27:27<06:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2660/3295 [27:27<06:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2661/3295 [27:27<06:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2661/3295 [27:27<06:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2662/3295 [27:27<06:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2662/3295 [27:27<06:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2663/3295 [27:29<06:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2663/3295 [27:29<06:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2664/3295 [27:29<06:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2664/3295 [27:29<06:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2665/3295 [27:29<06:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2665/3295 [27:29<06:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2666/3295 [27:31<06:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2666/3295 [27:31<06:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2667/3295 [27:31<06:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2667/3295 [27:31<06:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2668/3295 [27:31<06:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2668/3295 [27:31<06:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2669/3295 [27:32<06:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2669/3295 [27:32<06:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2670/3295 [27:32<06:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2670/3295 [27:32<06:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2671/3295 [27:32<06:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2671/3295 [27:32<06:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2672/3295 [27:34<06:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2672/3295 [27:34<06:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2673/3295 [27:34<06:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2673/3295 [27:34<06:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2674/3295 [27:34<06:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2674/3295 [27:34<06:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2675/3295 [27:36<06:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2675/3295 [27:36<06:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2676/3295 [27:36<06:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2676/3295 [27:36<06:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2677/3295 [27:36<06:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████  | 2677/3295 [27:36<06:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2678/3295 [27:38<06:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2678/3295 [27:38<06:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2679/3295 [27:38<06:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2679/3295 [27:38<06:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2680/3295 [27:38<06:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2680/3295 [27:38<06:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2681/3295 [27:40<06:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2681/3295 [27:40<06:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2682/3295 [27:40<06:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2682/3295 [27:40<06:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2683/3295 [27:40<06:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2683/3295 [27:40<06:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2684/3295 [27:41<06:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2684/3295 [27:41<06:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2685/3295 [27:41<06:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  81%|████████▏ | 2685/3295 [27:41<06:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2686/3295 [27:41<06:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2686/3295 [27:41<06:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2687/3295 [27:43<06:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2687/3295 [27:43<06:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2688/3295 [27:43<06:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2688/3295 [27:43<06:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2689/3295 [27:43<06:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2689/3295 [27:43<06:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2690/3295 [27:45<06:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2690/3295 [27:45<06:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2691/3295 [27:45<06:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2691/3295 [27:45<06:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2692/3295 [27:45<06:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2692/3295 [27:45<06:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2693/3295 [27:47<06:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2693/3295 [27:47<06:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2694/3295 [27:47<06:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2694/3295 [27:47<06:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2695/3295 [27:47<06:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2695/3295 [27:47<06:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2696/3295 [27:49<06:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2696/3295 [27:49<06:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2697/3295 [27:49<06:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2697/3295 [27:49<06:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2698/3295 [27:49<06:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2698/3295 [27:49<06:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2699/3295 [27:51<06:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2699/3295 [27:51<06:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2700/3295 [27:51<06:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2700/3295 [27:51<06:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2701/3295 [27:51<06:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2701/3295 [27:51<06:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2702/3295 [27:52<06:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2702/3295 [27:52<06:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2703/3295 [27:52<06:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2703/3295 [27:52<06:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2704/3295 [27:52<06:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2704/3295 [27:52<06:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2705/3295 [27:54<06:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2705/3295 [27:54<06:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2706/3295 [27:54<06:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2706/3295 [27:54<06:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2707/3295 [27:54<06:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2707/3295 [27:54<06:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2708/3295 [27:56<06:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2708/3295 [27:56<06:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2709/3295 [27:56<06:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2709/3295 [27:56<06:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2710/3295 [27:56<06:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2710/3295 [27:56<06:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2711/3295 [27:58<06:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2711/3295 [27:58<06:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2712/3295 [27:58<06:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2712/3295 [27:58<06:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2713/3295 [27:58<06:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2713/3295 [27:58<06:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2714/3295 [28:00<05:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2714/3295 [28:00<05:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2715/3295 [28:00<05:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2715/3295 [28:00<05:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2716/3295 [28:00<05:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2716/3295 [28:00<05:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2717/3295 [28:02<05:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2717/3295 [28:02<05:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2718/3295 [28:02<05:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  82%|████████▏ | 2718/3295 [28:02<05:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2719/3295 [28:02<05:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2719/3295 [28:02<05:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2720/3295 [28:04<05:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2720/3295 [28:04<05:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2721/3295 [28:04<05:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2721/3295 [28:04<05:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2722/3295 [28:04<05:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2722/3295 [28:04<05:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2723/3295 [28:05<05:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2723/3295 [28:05<05:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2724/3295 [28:05<05:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2724/3295 [28:05<05:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2725/3295 [28:05<05:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2725/3295 [28:05<05:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2726/3295 [28:07<05:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2726/3295 [28:07<05:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2727/3295 [28:07<05:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2727/3295 [28:07<05:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2728/3295 [28:07<05:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2728/3295 [28:07<05:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2729/3295 [28:09<05:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2729/3295 [28:09<05:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2730/3295 [28:09<05:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2730/3295 [28:09<05:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2731/3295 [28:09<05:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2731/3295 [28:09<05:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2732/3295 [28:11<05:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2732/3295 [28:11<05:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2733/3295 [28:11<05:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2733/3295 [28:11<05:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2734/3295 [28:11<05:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2734/3295 [28:11<05:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2735/3295 [28:13<05:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2735/3295 [28:13<05:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2736/3295 [28:13<05:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2736/3295 [28:13<05:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2737/3295 [28:13<05:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2737/3295 [28:13<05:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2738/3295 [28:15<05:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2738/3295 [28:15<05:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2739/3295 [28:15<05:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2739/3295 [28:15<05:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2740/3295 [28:15<05:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2740/3295 [28:15<05:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2741/3295 [28:17<05:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2741/3295 [28:17<05:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2742/3295 [28:17<05:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2742/3295 [28:17<05:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2743/3295 [28:17<05:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2743/3295 [28:17<05:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2744/3295 [28:18<05:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2744/3295 [28:18<05:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2745/3295 [28:19<05:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2745/3295 [28:19<05:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2746/3295 [28:19<05:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2746/3295 [28:19<05:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2747/3295 [28:20<05:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2747/3295 [28:20<05:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2748/3295 [28:20<05:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2748/3295 [28:20<05:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2749/3295 [28:20<05:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2749/3295 [28:20<05:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2750/3295 [28:22<05:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2750/3295 [28:22<05:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2751/3295 [28:22<05:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  83%|████████▎ | 2751/3295 [28:22<05:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2752/3295 [28:22<05:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2752/3295 [28:22<05:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2753/3295 [28:24<05:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2753/3295 [28:24<05:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2754/3295 [28:24<05:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2754/3295 [28:24<05:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2755/3295 [28:24<05:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2755/3295 [28:24<05:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2756/3295 [28:26<05:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2756/3295 [28:26<05:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2757/3295 [28:26<05:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2757/3295 [28:26<05:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2758/3295 [28:26<05:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2758/3295 [28:26<05:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2759/3295 [28:28<05:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▎ | 2759/3295 [28:28<05:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2760/3295 [28:28<05:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2760/3295 [28:28<05:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2761/3295 [28:28<05:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2761/3295 [28:28<05:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2762/3295 [28:30<05:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2762/3295 [28:30<05:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2763/3295 [28:30<05:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2763/3295 [28:30<05:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2764/3295 [28:30<05:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2764/3295 [28:30<05:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2765/3295 [28:31<05:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2765/3295 [28:31<05:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2766/3295 [28:32<05:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2766/3295 [28:32<05:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2767/3295 [28:32<05:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2767/3295 [28:32<05:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2768/3295 [28:33<05:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2768/3295 [28:33<05:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2769/3295 [28:33<05:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2769/3295 [28:33<05:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2770/3295 [28:33<05:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2770/3295 [28:33<05:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2771/3295 [28:35<05:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2771/3295 [28:35<05:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2772/3295 [28:35<05:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2772/3295 [28:35<05:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2773/3295 [28:35<05:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2773/3295 [28:35<05:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2774/3295 [28:37<05:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2774/3295 [28:37<05:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2775/3295 [28:37<05:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2775/3295 [28:37<05:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2776/3295 [28:37<05:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2776/3295 [28:37<05:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2777/3295 [28:39<05:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2777/3295 [28:39<05:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2778/3295 [28:39<05:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2778/3295 [28:39<05:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2779/3295 [28:39<05:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2779/3295 [28:39<05:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2780/3295 [28:41<05:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2780/3295 [28:41<05:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2781/3295 [28:41<05:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2781/3295 [28:41<05:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2782/3295 [28:41<05:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2782/3295 [28:41<05:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2783/3295 [28:42<05:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2783/3295 [28:42<05:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2784/3295 [28:42<05:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  84%|████████▍ | 2784/3295 [28:42<05:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2785/3295 [28:43<05:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2785/3295 [28:43<05:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2786/3295 [28:44<05:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2786/3295 [28:44<05:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2787/3295 [28:44<05:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2787/3295 [28:44<05:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2788/3295 [28:44<05:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2788/3295 [28:44<05:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2789/3295 [28:46<05:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2789/3295 [28:46<05:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2790/3295 [28:46<05:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2790/3295 [28:46<05:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2791/3295 [28:46<05:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2791/3295 [28:46<05:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2792/3295 [28:48<05:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2792/3295 [28:48<05:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2793/3295 [28:48<05:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2793/3295 [28:48<05:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2794/3295 [28:48<05:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2794/3295 [28:48<05:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2795/3295 [28:50<05:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2795/3295 [28:50<05:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2796/3295 [28:50<05:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2796/3295 [28:50<05:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2797/3295 [28:50<05:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2797/3295 [28:50<05:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2798/3295 [28:52<05:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2798/3295 [28:52<05:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2799/3295 [28:52<05:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2799/3295 [28:52<05:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2800/3295 [28:52<05:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▍ | 2800/3295 [28:52<05:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2801/3295 [28:54<05:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2801/3295 [28:54<05:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2802/3295 [28:54<05:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2802/3295 [28:54<05:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2803/3295 [28:54<05:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2803/3295 [28:54<05:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2804/3295 [28:55<05:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2804/3295 [28:55<05:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2805/3295 [28:55<05:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2805/3295 [28:55<05:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2806/3295 [28:55<05:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2806/3295 [28:55<05:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2807/3295 [28:57<05:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2807/3295 [28:57<05:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2808/3295 [28:57<05:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2808/3295 [28:57<05:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2809/3295 [28:57<05:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2809/3295 [28:57<05:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2810/3295 [28:59<05:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2810/3295 [28:59<05:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2811/3295 [28:59<04:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2811/3295 [28:59<04:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2812/3295 [28:59<04:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2812/3295 [28:59<04:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2813/3295 [29:01<04:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2813/3295 [29:01<04:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2814/3295 [29:01<04:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2814/3295 [29:01<04:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2815/3295 [29:01<04:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2815/3295 [29:01<04:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2816/3295 [29:03<04:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2816/3295 [29:03<04:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2817/3295 [29:03<04:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  85%|████████▌ | 2817/3295 [29:03<04:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2818/3295 [29:03<04:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2818/3295 [29:03<04:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2819/3295 [29:05<04:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2819/3295 [29:05<04:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2820/3295 [29:05<04:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2820/3295 [29:05<04:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2821/3295 [29:05<04:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2821/3295 [29:05<04:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2822/3295 [29:07<04:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2822/3295 [29:07<04:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2823/3295 [29:07<04:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2823/3295 [29:07<04:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2824/3295 [29:07<04:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2824/3295 [29:07<04:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2825/3295 [29:09<04:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2825/3295 [29:09<04:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2826/3295 [29:09<04:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2826/3295 [29:09<04:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2827/3295 [29:09<04:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2827/3295 [29:09<04:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2828/3295 [29:10<04:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2828/3295 [29:10<04:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2829/3295 [29:10<04:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2829/3295 [29:10<04:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2830/3295 [29:10<04:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2830/3295 [29:10<04:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2831/3295 [29:12<04:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2831/3295 [29:12<04:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2832/3295 [29:12<04:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2832/3295 [29:12<04:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2833/3295 [29:12<04:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2833/3295 [29:12<04:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2834/3295 [29:14<04:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2834/3295 [29:14<04:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2835/3295 [29:14<04:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2835/3295 [29:14<04:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2836/3295 [29:14<04:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2836/3295 [29:14<04:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2837/3295 [29:16<04:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2837/3295 [29:16<04:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2838/3295 [29:16<04:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2838/3295 [29:16<04:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2839/3295 [29:16<04:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2839/3295 [29:16<04:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2840/3295 [29:18<04:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2840/3295 [29:18<04:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2841/3295 [29:18<04:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▌ | 2841/3295 [29:18<04:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2842/3295 [29:18<04:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2842/3295 [29:18<04:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2843/3295 [29:19<04:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2843/3295 [29:19<04:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2844/3295 [29:19<04:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2844/3295 [29:19<04:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2845/3295 [29:19<04:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2845/3295 [29:19<04:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2846/3295 [29:21<04:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2846/3295 [29:21<04:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2847/3295 [29:21<04:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2847/3295 [29:21<04:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2848/3295 [29:21<04:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2848/3295 [29:21<04:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2849/3295 [29:23<04:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2849/3295 [29:23<04:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2850/3295 [29:23<04:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  86%|████████▋ | 2850/3295 [29:23<04:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2851/3295 [29:23<04:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2851/3295 [29:23<04:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2852/3295 [29:25<04:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2852/3295 [29:25<04:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2853/3295 [29:25<04:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2853/3295 [29:25<04:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2854/3295 [29:25<04:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2854/3295 [29:25<04:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2855/3295 [29:27<04:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2855/3295 [29:27<04:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2856/3295 [29:27<04:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2856/3295 [29:27<04:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2857/3295 [29:27<04:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2857/3295 [29:27<04:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2858/3295 [29:29<04:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2858/3295 [29:29<04:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2859/3295 [29:29<04:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2859/3295 [29:29<04:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2860/3295 [29:29<04:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2860/3295 [29:29<04:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2861/3295 [29:31<04:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2861/3295 [29:31<04:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2862/3295 [29:31<04:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2862/3295 [29:31<04:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2863/3295 [29:31<04:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2863/3295 [29:31<04:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2864/3295 [29:32<04:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2864/3295 [29:32<04:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2865/3295 [29:32<04:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2865/3295 [29:32<04:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2866/3295 [29:33<04:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2866/3295 [29:33<04:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2867/3295 [29:34<04:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2867/3295 [29:34<04:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2868/3295 [29:34<04:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2868/3295 [29:34<04:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2869/3295 [29:34<04:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2869/3295 [29:34<04:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2870/3295 [29:36<04:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2870/3295 [29:36<04:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2871/3295 [29:36<04:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2871/3295 [29:36<04:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2872/3295 [29:36<04:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2872/3295 [29:36<04:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2873/3295 [29:38<04:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2873/3295 [29:38<04:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2874/3295 [29:38<04:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2874/3295 [29:38<04:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2875/3295 [29:38<04:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2875/3295 [29:38<04:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2876/3295 [29:40<04:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2876/3295 [29:40<04:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2877/3295 [29:40<04:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2877/3295 [29:40<04:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2878/3295 [29:40<04:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2878/3295 [29:40<04:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2879/3295 [29:42<04:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2879/3295 [29:42<04:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2880/3295 [29:42<04:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2880/3295 [29:42<04:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2881/3295 [29:42<04:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2881/3295 [29:42<04:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2882/3295 [29:44<04:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2882/3295 [29:44<04:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2883/3295 [29:44<04:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  87%|████████▋ | 2883/3295 [29:44<04:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2884/3295 [29:44<04:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2884/3295 [29:44<04:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2885/3295 [29:46<04:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2885/3295 [29:46<04:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2886/3295 [29:46<04:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2886/3295 [29:46<04:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2887/3295 [29:46<04:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2887/3295 [29:46<04:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2888/3295 [29:47<04:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2888/3295 [29:47<04:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2889/3295 [29:47<04:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2889/3295 [29:47<04:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2890/3295 [29:47<04:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2890/3295 [29:47<04:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2891/3295 [29:49<04:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2891/3295 [29:49<04:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2892/3295 [29:49<04:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2892/3295 [29:49<04:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2893/3295 [29:49<04:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2893/3295 [29:49<04:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2894/3295 [29:51<04:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2894/3295 [29:51<04:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2895/3295 [29:51<04:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2895/3295 [29:51<04:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2896/3295 [29:51<04:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2896/3295 [29:51<04:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2897/3295 [29:53<04:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2897/3295 [29:53<04:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2898/3295 [29:53<04:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2898/3295 [29:53<04:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2899/3295 [29:53<04:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2899/3295 [29:53<04:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2900/3295 [29:55<04:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2900/3295 [29:55<04:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2901/3295 [29:55<04:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2901/3295 [29:55<04:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2902/3295 [29:55<04:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2902/3295 [29:55<04:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2903/3295 [29:57<04:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2903/3295 [29:57<04:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2904/3295 [29:57<04:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2904/3295 [29:57<04:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2905/3295 [29:57<04:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2905/3295 [29:57<04:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2906/3295 [29:58<04:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2906/3295 [29:58<04:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2907/3295 [29:58<04:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2907/3295 [29:58<04:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2908/3295 [29:59<03:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2908/3295 [29:59<03:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2909/3295 [30:00<03:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2909/3295 [30:00<03:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2910/3295 [30:00<03:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2910/3295 [30:00<03:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2911/3295 [30:00<03:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2911/3295 [30:00<03:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2912/3295 [30:02<03:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2912/3295 [30:02<03:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2913/3295 [30:02<03:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2913/3295 [30:02<03:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2914/3295 [30:02<03:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2914/3295 [30:02<03:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2915/3295 [30:04<03:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2915/3295 [30:04<03:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2916/3295 [30:04<03:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  88%|████████▊ | 2916/3295 [30:04<03:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2917/3295 [30:04<03:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2917/3295 [30:04<03:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2918/3295 [30:06<03:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2918/3295 [30:06<03:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2919/3295 [30:06<03:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2919/3295 [30:06<03:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2920/3295 [30:06<03:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2920/3295 [30:06<03:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2921/3295 [30:08<03:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2921/3295 [30:08<03:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2922/3295 [30:08<03:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2922/3295 [30:08<03:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2923/3295 [30:08<03:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2923/3295 [30:08<03:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2924/3295 [30:10<03:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▊ | 2924/3295 [30:10<03:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2925/3295 [30:10<03:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2925/3295 [30:10<03:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2926/3295 [30:10<03:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2926/3295 [30:10<03:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2927/3295 [30:12<03:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2927/3295 [30:12<03:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2928/3295 [30:12<03:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2928/3295 [30:12<03:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2929/3295 [30:12<03:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2929/3295 [30:12<03:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2930/3295 [30:13<03:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2930/3295 [30:13<03:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2931/3295 [30:13<03:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2931/3295 [30:13<03:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2932/3295 [30:13<03:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2932/3295 [30:13<03:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2933/3295 [30:15<03:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2933/3295 [30:15<03:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2934/3295 [30:15<03:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2934/3295 [30:15<03:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2935/3295 [30:15<03:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2935/3295 [30:15<03:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2936/3295 [30:17<03:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2936/3295 [30:17<03:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2937/3295 [30:17<03:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2937/3295 [30:17<03:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2938/3295 [30:17<03:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2938/3295 [30:17<03:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2939/3295 [30:19<03:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2939/3295 [30:19<03:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2940/3295 [30:19<03:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2940/3295 [30:19<03:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2941/3295 [30:19<03:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2941/3295 [30:19<03:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2942/3295 [30:21<03:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2942/3295 [30:21<03:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2943/3295 [30:21<03:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2943/3295 [30:21<03:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2944/3295 [30:21<03:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2944/3295 [30:21<03:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2945/3295 [30:23<03:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2945/3295 [30:23<03:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2946/3295 [30:23<03:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2946/3295 [30:23<03:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2947/3295 [30:23<03:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2947/3295 [30:23<03:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2948/3295 [30:25<03:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2948/3295 [30:25<03:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2949/3295 [30:25<03:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  89%|████████▉ | 2949/3295 [30:25<03:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2950/3295 [30:25<03:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2950/3295 [30:25<03:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2951/3295 [30:27<03:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2951/3295 [30:27<03:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2952/3295 [30:27<03:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2952/3295 [30:27<03:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2953/3295 [30:27<03:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2953/3295 [30:27<03:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2954/3295 [30:28<03:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2954/3295 [30:28<03:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2955/3295 [30:28<03:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2955/3295 [30:28<03:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2956/3295 [30:28<03:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2956/3295 [30:28<03:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2957/3295 [30:30<03:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2957/3295 [30:30<03:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2958/3295 [30:30<03:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2958/3295 [30:30<03:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2959/3295 [30:30<03:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2959/3295 [30:30<03:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2960/3295 [30:32<03:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2960/3295 [30:32<03:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2961/3295 [30:32<03:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2961/3295 [30:32<03:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2962/3295 [30:32<03:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2962/3295 [30:32<03:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2963/3295 [30:34<03:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2963/3295 [30:34<03:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2964/3295 [30:34<03:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2964/3295 [30:34<03:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2965/3295 [30:34<03:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|████████▉ | 2965/3295 [30:34<03:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2966/3295 [30:36<03:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2966/3295 [30:36<03:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2967/3295 [30:36<03:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2967/3295 [30:36<03:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2968/3295 [30:36<03:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2968/3295 [30:36<03:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2969/3295 [30:38<03:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2969/3295 [30:38<03:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2970/3295 [30:38<03:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2970/3295 [30:38<03:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2971/3295 [30:38<03:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2971/3295 [30:38<03:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2972/3295 [30:40<03:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2972/3295 [30:40<03:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2973/3295 [30:40<03:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2973/3295 [30:40<03:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2974/3295 [30:40<03:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2974/3295 [30:40<03:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2975/3295 [30:42<03:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2975/3295 [30:42<03:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2976/3295 [30:42<03:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2976/3295 [30:42<03:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2977/3295 [30:42<03:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2977/3295 [30:42<03:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2978/3295 [30:44<03:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2978/3295 [30:44<03:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2979/3295 [30:44<03:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2979/3295 [30:44<03:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2980/3295 [30:44<03:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2980/3295 [30:44<03:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2981/3295 [30:45<03:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  90%|█████████ | 2981/3295 [30:45<03:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2982/3295 [30:45<03:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2982/3295 [30:45<03:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2983/3295 [30:45<03:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2983/3295 [30:45<03:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2984/3295 [30:47<03:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2984/3295 [30:47<03:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2985/3295 [30:47<03:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2985/3295 [30:47<03:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2986/3295 [30:47<03:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2986/3295 [30:47<03:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2987/3295 [30:49<03:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2987/3295 [30:49<03:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2988/3295 [30:49<03:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2988/3295 [30:49<03:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2989/3295 [30:49<03:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2989/3295 [30:49<03:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2990/3295 [30:51<03:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2990/3295 [30:51<03:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2991/3295 [30:51<03:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2991/3295 [30:51<03:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2992/3295 [30:51<03:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2992/3295 [30:51<03:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2993/3295 [30:53<03:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2993/3295 [30:53<03:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2994/3295 [30:53<03:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2994/3295 [30:53<03:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2995/3295 [30:53<03:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2995/3295 [30:53<03:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2996/3295 [30:55<03:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2996/3295 [30:55<03:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2997/3295 [30:55<03:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2997/3295 [30:55<03:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2998/3295 [30:55<03:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2998/3295 [30:55<03:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2999/3295 [30:57<03:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 2999/3295 [30:57<03:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3000/3295 [30:57<03:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3000/3295 [30:57<03:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3001/3295 [30:57<03:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3001/3295 [30:57<03:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3002/3295 [30:58<03:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3002/3295 [30:58<03:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3003/3295 [30:58<03:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3003/3295 [30:58<03:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3004/3295 [30:58<03:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3004/3295 [30:58<03:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3005/3295 [31:00<02:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3005/3295 [31:00<02:59,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3006/3295 [31:00<02:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████ | 3006/3295 [31:00<02:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3007/3295 [31:00<02:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3007/3295 [31:00<02:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3008/3295 [31:02<02:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3008/3295 [31:02<02:57,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3009/3295 [31:02<02:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3009/3295 [31:02<02:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3010/3295 [31:02<02:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3010/3295 [31:02<02:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3011/3295 [31:04<02:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3011/3295 [31:04<02:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3012/3295 [31:04<02:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3012/3295 [31:04<02:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3013/3295 [31:04<02:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3013/3295 [31:04<02:54,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3014/3295 [31:06<02:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  91%|█████████▏| 3014/3295 [31:06<02:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3015/3295 [31:06<02:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3015/3295 [31:06<02:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3016/3295 [31:06<02:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3016/3295 [31:06<02:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3017/3295 [31:08<02:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3017/3295 [31:08<02:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3018/3295 [31:08<02:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3018/3295 [31:08<02:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3019/3295 [31:08<02:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3019/3295 [31:08<02:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3020/3295 [31:10<02:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3020/3295 [31:10<02:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3021/3295 [31:10<02:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3021/3295 [31:10<02:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3022/3295 [31:10<02:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3022/3295 [31:10<02:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3023/3295 [31:12<02:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3023/3295 [31:12<02:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3024/3295 [31:12<02:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3024/3295 [31:12<02:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3025/3295 [31:12<02:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3025/3295 [31:12<02:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3026/3295 [31:13<02:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3026/3295 [31:13<02:46,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3027/3295 [31:13<02:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3027/3295 [31:13<02:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3028/3295 [31:13<02:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3028/3295 [31:13<02:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3029/3295 [31:15<02:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3029/3295 [31:15<02:44,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3030/3295 [31:15<02:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3030/3295 [31:15<02:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3031/3295 [31:15<02:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3031/3295 [31:15<02:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3032/3295 [31:17<02:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3032/3295 [31:17<02:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3033/3295 [31:17<02:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3033/3295 [31:17<02:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3034/3295 [31:17<02:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3034/3295 [31:17<02:41,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3035/3295 [31:19<02:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3035/3295 [31:19<02:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3036/3295 [31:19<02:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3036/3295 [31:19<02:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3037/3295 [31:19<02:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3037/3295 [31:19<02:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3038/3295 [31:21<02:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3038/3295 [31:21<02:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3039/3295 [31:21<02:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3039/3295 [31:21<02:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3040/3295 [31:21<02:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3040/3295 [31:21<02:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3041/3295 [31:23<02:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3041/3295 [31:23<02:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3042/3295 [31:23<02:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3042/3295 [31:23<02:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3043/3295 [31:23<02:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3043/3295 [31:23<02:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3044/3295 [31:25<02:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3044/3295 [31:25<02:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3045/3295 [31:25<02:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3045/3295 [31:25<02:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3046/3295 [31:25<02:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3046/3295 [31:25<02:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3047/3295 [31:26<02:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  92%|█████████▏| 3047/3295 [31:26<02:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3048/3295 [31:26<02:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3048/3295 [31:26<02:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3049/3295 [31:27<02:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3049/3295 [31:27<02:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3050/3295 [31:28<02:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3050/3295 [31:28<02:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3051/3295 [31:28<02:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3051/3295 [31:28<02:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3052/3295 [31:28<02:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3052/3295 [31:28<02:30,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3053/3295 [31:30<02:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3053/3295 [31:30<02:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3054/3295 [31:30<02:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3054/3295 [31:30<02:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3055/3295 [31:30<02:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3055/3295 [31:30<02:28,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3056/3295 [31:32<02:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3056/3295 [31:32<02:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3057/3295 [31:32<02:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3057/3295 [31:32<02:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3058/3295 [31:32<02:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3058/3295 [31:32<02:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3059/3295 [31:34<02:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3059/3295 [31:34<02:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3060/3295 [31:34<02:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3060/3295 [31:34<02:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3061/3295 [31:34<02:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3061/3295 [31:34<02:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3062/3295 [31:36<02:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3062/3295 [31:36<02:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3063/3295 [31:36<02:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3063/3295 [31:36<02:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3064/3295 [31:36<02:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3064/3295 [31:36<02:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3065/3295 [31:38<02:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3065/3295 [31:38<02:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3066/3295 [31:38<02:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3066/3295 [31:38<02:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3067/3295 [31:38<02:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3067/3295 [31:38<02:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3068/3295 [31:40<02:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3068/3295 [31:40<02:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3069/3295 [31:40<02:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3069/3295 [31:40<02:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3070/3295 [31:40<02:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3070/3295 [31:40<02:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3071/3295 [31:42<02:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3071/3295 [31:42<02:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3072/3295 [31:42<02:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3072/3295 [31:42<02:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3073/3295 [31:42<02:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3073/3295 [31:42<02:17,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3074/3295 [31:43<02:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3074/3295 [31:43<02:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3075/3295 [31:44<02:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3075/3295 [31:44<02:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3076/3295 [31:44<02:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3076/3295 [31:44<02:15,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3077/3295 [31:45<02:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3077/3295 [31:45<02:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3078/3295 [31:45<02:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3078/3295 [31:45<02:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3079/3295 [31:45<02:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3079/3295 [31:45<02:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3080/3295 [31:47<02:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  93%|█████████▎| 3080/3295 [31:47<02:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3081/3295 [31:47<02:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3081/3295 [31:47<02:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3082/3295 [31:47<02:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3082/3295 [31:47<02:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3083/3295 [31:49<02:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3083/3295 [31:49<02:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3084/3295 [31:49<02:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3084/3295 [31:49<02:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3085/3295 [31:49<02:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3085/3295 [31:49<02:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3086/3295 [31:51<02:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3086/3295 [31:51<02:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3087/3295 [31:51<02:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3087/3295 [31:51<02:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3088/3295 [31:51<02:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3088/3295 [31:51<02:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3089/3295 [31:53<02:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▎| 3089/3295 [31:53<02:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3090/3295 [31:53<02:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3090/3295 [31:53<02:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3091/3295 [31:53<02:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3091/3295 [31:53<02:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3092/3295 [31:55<02:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3092/3295 [31:55<02:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3093/3295 [31:55<02:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3093/3295 [31:55<02:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3094/3295 [31:55<02:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3094/3295 [31:55<02:04,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3095/3295 [31:56<02:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3095/3295 [31:56<02:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3096/3295 [31:56<02:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3096/3295 [31:56<02:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3097/3295 [31:57<02:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3097/3295 [31:57<02:02,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3098/3295 [31:58<02:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3098/3295 [31:58<02:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3099/3295 [31:58<02:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3099/3295 [31:58<02:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3100/3295 [31:58<02:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3100/3295 [31:58<02:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3101/3295 [32:00<02:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3101/3295 [32:00<02:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3102/3295 [32:00<01:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3102/3295 [32:00<01:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3103/3295 [32:00<01:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3103/3295 [32:00<01:58,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3104/3295 [32:02<01:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3104/3295 [32:02<01:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3105/3295 [32:02<01:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3105/3295 [32:02<01:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3106/3295 [32:02<01:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3106/3295 [32:02<01:56,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3107/3295 [32:04<01:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3107/3295 [32:04<01:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3108/3295 [32:04<01:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3108/3295 [32:04<01:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3109/3295 [32:04<01:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3109/3295 [32:04<01:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3110/3295 [32:05<01:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3110/3295 [32:05<01:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3111/3295 [32:05<01:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3111/3295 [32:05<01:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3112/3295 [32:05<01:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3112/3295 [32:05<01:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3113/3295 [32:07<01:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  94%|█████████▍| 3113/3295 [32:07<01:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3114/3295 [32:07<01:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3114/3295 [32:07<01:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3115/3295 [32:07<01:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3115/3295 [32:07<01:51,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3116/3295 [32:09<01:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3116/3295 [32:09<01:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3117/3295 [32:09<01:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3117/3295 [32:09<01:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3118/3295 [32:09<01:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3118/3295 [32:09<01:49,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3119/3295 [32:11<01:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3119/3295 [32:11<01:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3120/3295 [32:11<01:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3120/3295 [32:11<01:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3121/3295 [32:11<01:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3121/3295 [32:11<01:47,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3122/3295 [32:13<01:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3122/3295 [32:13<01:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3123/3295 [32:13<01:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3123/3295 [32:13<01:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3124/3295 [32:13<01:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3124/3295 [32:13<01:45,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3125/3295 [32:15<01:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3125/3295 [32:15<01:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3126/3295 [32:15<01:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3126/3295 [32:15<01:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3127/3295 [32:15<01:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3127/3295 [32:15<01:43,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3128/3295 [32:17<01:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3128/3295 [32:17<01:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3129/3295 [32:17<01:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3129/3295 [32:17<01:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3130/3295 [32:17<01:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▍| 3130/3295 [32:17<01:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3131/3295 [32:19<01:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3131/3295 [32:19<01:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3132/3295 [32:19<01:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3132/3295 [32:19<01:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3133/3295 [32:19<01:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3133/3295 [32:19<01:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3134/3295 [32:21<01:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3134/3295 [32:21<01:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3135/3295 [32:21<01:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3135/3295 [32:21<01:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3136/3295 [32:21<01:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3136/3295 [32:21<01:38,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3137/3295 [32:22<01:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3137/3295 [32:22<01:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3138/3295 [32:22<01:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3138/3295 [32:22<01:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3139/3295 [32:23<01:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3139/3295 [32:23<01:36,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3140/3295 [32:24<01:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3140/3295 [32:24<01:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3141/3295 [32:24<01:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3141/3295 [32:24<01:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3142/3295 [32:24<01:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3142/3295 [32:24<01:34,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3143/3295 [32:26<01:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3143/3295 [32:26<01:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3144/3295 [32:26<01:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3144/3295 [32:26<01:33,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3145/3295 [32:26<01:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3145/3295 [32:26<01:32,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3146/3295 [32:28<01:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  95%|█████████▌| 3146/3295 [32:28<01:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3147/3295 [32:28<01:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3147/3295 [32:28<01:31,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3148/3295 [32:28<01:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3148/3295 [32:28<01:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3149/3295 [32:30<01:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3149/3295 [32:30<01:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3150/3295 [32:30<01:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3150/3295 [32:30<01:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3151/3295 [32:30<01:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3151/3295 [32:30<01:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3152/3295 [32:32<01:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3152/3295 [32:32<01:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3153/3295 [32:32<01:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3153/3295 [32:32<01:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3154/3295 [32:32<01:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3154/3295 [32:32<01:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3155/3295 [32:34<01:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3155/3295 [32:34<01:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3156/3295 [32:34<01:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3156/3295 [32:34<01:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3157/3295 [32:34<01:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3157/3295 [32:34<01:25,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3158/3295 [32:36<01:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3158/3295 [32:36<01:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3159/3295 [32:36<01:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3159/3295 [32:36<01:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3160/3295 [32:36<01:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3160/3295 [32:36<01:23,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3161/3295 [32:38<01:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3161/3295 [32:38<01:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3162/3295 [32:38<01:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3162/3295 [32:38<01:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3163/3295 [32:38<01:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3163/3295 [32:38<01:21,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3164/3295 [32:40<01:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3164/3295 [32:40<01:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3165/3295 [32:40<01:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3165/3295 [32:40<01:20,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3166/3295 [32:40<01:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3166/3295 [32:40<01:19,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3167/3295 [32:41<01:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3167/3295 [32:41<01:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3168/3295 [32:41<01:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3168/3295 [32:41<01:18,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3169/3295 [32:41<01:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3169/3295 [32:41<01:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3170/3295 [32:43<01:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3170/3295 [32:43<01:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3171/3295 [32:43<01:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▌| 3171/3295 [32:43<01:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3172/3295 [32:43<01:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3172/3295 [32:43<01:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3173/3295 [32:45<01:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3173/3295 [32:45<01:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3174/3295 [32:45<01:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3174/3295 [32:45<01:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3175/3295 [32:45<01:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3175/3295 [32:45<01:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3176/3295 [32:47<01:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3176/3295 [32:47<01:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3177/3295 [32:47<01:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3177/3295 [32:47<01:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3178/3295 [32:47<01:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3178/3295 [32:47<01:12,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3179/3295 [32:49<01:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  96%|█████████▋| 3179/3295 [32:49<01:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3180/3295 [32:49<01:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3180/3295 [32:49<01:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3181/3295 [32:49<01:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3181/3295 [32:49<01:10,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3182/3295 [32:51<01:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3182/3295 [32:51<01:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3183/3295 [32:51<01:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3183/3295 [32:51<01:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3184/3295 [32:51<01:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3184/3295 [32:51<01:08,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3185/3295 [32:53<01:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3185/3295 [32:53<01:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3186/3295 [32:53<01:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3186/3295 [32:53<01:07,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3187/3295 [32:53<01:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3187/3295 [32:53<01:06,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3188/3295 [32:54<01:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3188/3295 [32:54<01:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3189/3295 [32:54<01:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3189/3295 [32:54<01:05,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3190/3295 [32:54<01:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3190/3295 [32:54<01:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3191/3295 [32:56<01:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3191/3295 [32:56<01:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3192/3295 [32:56<01:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3192/3295 [32:56<01:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3193/3295 [32:56<01:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3193/3295 [32:56<01:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3194/3295 [32:58<01:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3194/3295 [32:58<01:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3195/3295 [32:58<01:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3195/3295 [32:58<01:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3196/3295 [32:58<01:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3196/3295 [32:58<01:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3197/3295 [33:00<01:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3197/3295 [33:00<01:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3198/3295 [33:00<01:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3198/3295 [33:00<01:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3199/3295 [33:00<00:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3199/3295 [33:00<00:59,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3200/3295 [33:02<00:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3200/3295 [33:02<00:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3201/3295 [33:02<00:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3201/3295 [33:02<00:58,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3202/3295 [33:02<00:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3202/3295 [33:02<00:57,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3203/3295 [33:04<00:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3203/3295 [33:04<00:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3204/3295 [33:04<00:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3204/3295 [33:04<00:56,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3205/3295 [33:04<00:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3205/3295 [33:04<00:55,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3206/3295 [33:06<00:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3206/3295 [33:06<00:55,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3207/3295 [33:06<00:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3207/3295 [33:06<00:54,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3208/3295 [33:06<00:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3208/3295 [33:06<00:53,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3209/3295 [33:08<00:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3209/3295 [33:08<00:53,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3210/3295 [33:08<00:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3210/3295 [33:08<00:52,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3211/3295 [33:08<00:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3211/3295 [33:08<00:52,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3212/3295 [33:09<00:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  97%|█████████▋| 3212/3295 [33:09<00:51,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3213/3295 [33:09<00:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3213/3295 [33:09<00:50,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3214/3295 [33:09<00:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3214/3295 [33:09<00:50,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3215/3295 [33:11<00:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3215/3295 [33:11<00:49,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3216/3295 [33:11<00:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3216/3295 [33:11<00:48,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3217/3295 [33:11<00:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3217/3295 [33:11<00:48,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3218/3295 [33:13<00:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3218/3295 [33:13<00:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3219/3295 [33:13<00:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3219/3295 [33:13<00:47,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3220/3295 [33:13<00:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3220/3295 [33:13<00:46,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3221/3295 [33:15<00:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3221/3295 [33:15<00:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3222/3295 [33:15<00:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3222/3295 [33:15<00:45,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3223/3295 [33:15<00:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3223/3295 [33:15<00:44,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3224/3295 [33:17<00:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3224/3295 [33:17<00:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3225/3295 [33:17<00:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3225/3295 [33:17<00:43,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3226/3295 [33:17<00:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3226/3295 [33:17<00:42,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3227/3295 [33:18<00:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3227/3295 [33:18<00:42,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3228/3295 [33:18<00:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3228/3295 [33:18<00:41,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3229/3295 [33:19<00:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3229/3295 [33:19<00:40,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3230/3295 [33:20<00:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3230/3295 [33:20<00:40,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3231/3295 [33:20<00:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3231/3295 [33:20<00:39,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3232/3295 [33:20<00:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3232/3295 [33:20<00:39,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3233/3295 [33:22<00:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3233/3295 [33:22<00:38,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3234/3295 [33:22<00:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3234/3295 [33:22<00:37,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3235/3295 [33:22<00:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3235/3295 [33:22<00:37,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3236/3295 [33:24<00:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3236/3295 [33:24<00:36,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3237/3295 [33:24<00:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3237/3295 [33:24<00:35,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3238/3295 [33:24<00:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3238/3295 [33:24<00:35,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3239/3295 [33:26<00:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3239/3295 [33:26<00:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3240/3295 [33:26<00:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3240/3295 [33:26<00:34,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3241/3295 [33:26<00:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3241/3295 [33:26<00:33,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3242/3295 [33:28<00:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3242/3295 [33:28<00:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3243/3295 [33:28<00:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3243/3295 [33:28<00:32,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3244/3295 [33:28<00:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3244/3295 [33:28<00:31,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3245/3295 [33:30<00:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  98%|█████████▊| 3245/3295 [33:30<00:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3246/3295 [33:30<00:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3246/3295 [33:30<00:30,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3247/3295 [33:30<00:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3247/3295 [33:30<00:29,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3248/3295 [33:32<00:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3248/3295 [33:32<00:29,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3249/3295 [33:32<00:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3249/3295 [33:32<00:28,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3250/3295 [33:32<00:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3250/3295 [33:32<00:27,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3251/3295 [33:33<00:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3251/3295 [33:33<00:27,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3252/3295 [33:33<00:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3252/3295 [33:33<00:26,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3253/3295 [33:33<00:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▊| 3253/3295 [33:33<00:26,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3254/3295 [33:35<00:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3254/3295 [33:35<00:25,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3255/3295 [33:35<00:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3255/3295 [33:35<00:24,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3256/3295 [33:35<00:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3256/3295 [33:35<00:24,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3257/3295 [33:37<00:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3257/3295 [33:37<00:23,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3258/3295 [33:37<00:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3258/3295 [33:37<00:22,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3259/3295 [33:37<00:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3259/3295 [33:37<00:22,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3260/3295 [33:39<00:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3260/3295 [33:39<00:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3261/3295 [33:39<00:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3261/3295 [33:39<00:21,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3262/3295 [33:39<00:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3262/3295 [33:39<00:20,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3263/3295 [33:41<00:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3263/3295 [33:41<00:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3264/3295 [33:41<00:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3264/3295 [33:41<00:19,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3265/3295 [33:41<00:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3265/3295 [33:41<00:18,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3266/3295 [33:43<00:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3266/3295 [33:43<00:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3267/3295 [33:43<00:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3267/3295 [33:43<00:17,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3268/3295 [33:43<00:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3268/3295 [33:43<00:16,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3269/3295 [33:44<00:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3269/3295 [33:44<00:16,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3270/3295 [33:44<00:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3270/3295 [33:44<00:15,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3271/3295 [33:45<00:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3271/3295 [33:45<00:14,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3272/3295 [33:46<00:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3272/3295 [33:46<00:14,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3273/3295 [33:46<00:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3273/3295 [33:46<00:13,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3274/3295 [33:46<00:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3274/3295 [33:46<00:13,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3275/3295 [33:48<00:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3275/3295 [33:48<00:12,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3276/3295 [33:48<00:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3276/3295 [33:48<00:11,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3277/3295 [33:48<00:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3277/3295 [33:48<00:11,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3278/3295 [33:50<00:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1:  99%|█████████▉| 3278/3295 [33:50<00:10,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3279/3295 [33:50<00:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3279/3295 [33:50<00:09,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3280/3295 [33:50<00:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3280/3295 [33:50<00:09,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3281/3295 [33:52<00:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3281/3295 [33:52<00:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3282/3295 [33:52<00:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3282/3295 [33:52<00:08,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3283/3295 [33:52<00:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3283/3295 [33:52<00:07,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3284/3295 [33:54<00:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3284/3295 [33:54<00:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3285/3295 [33:54<00:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3285/3295 [33:54<00:06,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3286/3295 [33:54<00:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3286/3295 [33:54<00:05,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3287/3295 [33:56<00:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3287/3295 [33:56<00:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3288/3295 [33:56<00:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3288/3295 [33:56<00:04,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3289/3295 [33:56<00:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3289/3295 [33:56<00:03,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3290/3295 [33:57<00:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3290/3295 [33:57<00:03,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3291/3295 [33:57<00:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3291/3295 [33:57<00:02,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3292/3295 [33:58<00:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3292/3295 [33:58<00:01,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3293/3295 [33:59<00:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3293/3295 [33:59<00:01,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3294/3295 [33:59<00:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|█████████▉| 3294/3295 [33:59<00:00,  1.61it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|██████████| 3295/3295 [33:59<00:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|██████████| 3295/3295 [33:59<00:00,  1.62it/s, R@1 (val)=0.363, R@5 (val)=0.577, R@20 (val)=0.772, loss (train)=1.270, R@1 (train)=0.668]

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 0/602 [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 1/602 [00:00<00:08, 68.12it/s][A

Validation DataLoader 0:   0%|          | 2/602 [00:00<00:09, 61.80it/s][A

Validation DataLoader 0:   0%|          | 3/602 [00:00<00:09, 60.78it/s][A

Validation DataLoader 0:   1%|          | 4/602 [00:00<00:09, 62.81it/s][A

Validation DataLoader 0:   1%|          | 5/602 [00:00<00:09, 64.24it/s][A

Validation DataLoader 0:   1%|          | 6/602 [00:00<00:09, 65.43it/s][A

Validation DataLoader 0:   1%|          | 7/602 [00:00<00:30, 19.21it/s][A

Validation DataLoader 0:   1%|▏         | 8/602 [00:00<00:36, 16.08it/s][A

Validation DataLoader 0:   1%|▏         | 9/602 [00:00<00:37, 16.01it/s][A

Validation DataLoader 0:   2%|▏         | 10/602 [00:00<00:42, 13.83it/s][A

Validation DataLoader 0:   2%|▏         | 11/602 [00:00<00:49, 12.02it/s][A

Validation DataLoader 0:   2%|▏         | 12/602 [00:00<00:48, 12.08it/s][A

Validation DataLoader 0:   2%|▏         | 13/602 [00:01<00:52, 11.29it/s][A

Validation DataLoader 0:   2%|▏         | 14/602 [00:01<00:56, 10.37it/s][A

Validation DataLoader 0:   2%|▏         | 15/602 [00:01<00:55, 10.52it/s][A

Validation DataLoader 0:   3%|▎         | 16/602 [00:01<00:56, 10.33it/s][A

Validation DataLoader 0:   3%|▎         | 17/602 [00:01<01:04,  9.07it/s][A

Validation DataLoader 0:   3%|▎         | 18/602 [00:01<01:01,  9.45it/s][A

Validation DataLoader 0:   3%|▎         | 19/602 [00:01<00:59,  9.83it/s][A

Validation DataLoader 0:   3%|▎         | 20/602 [00:02<01:07,  8.65it/s][A

Validation DataLoader 0:   3%|▎         | 21/602 [00:02<01:05,  8.84it/s][A

Validation DataLoader 0:   4%|▎         | 22/602 [00:02<01:02,  9.21it/s][A

Validation DataLoader 0:   4%|▍         | 23/602 [00:02<01:09,  8.39it/s][A

Validation DataLoader 0:   4%|▍         | 24/602 [00:02<01:08,  8.50it/s][A

Validation DataLoader 0:   4%|▍         | 25/602 [00:02<01:05,  8.81it/s][A

Validation DataLoader 0:   4%|▍         | 26/602 [00:03<01:10,  8.17it/s][A

Validation DataLoader 0:   4%|▍         | 27/602 [00:03<01:08,  8.41it/s][A

Validation DataLoader 0:   5%|▍         | 28/602 [00:03<01:07,  8.52it/s][A

Validation DataLoader 0:   5%|▍         | 29/602 [00:03<01:11,  7.96it/s][A

Validation DataLoader 0:   5%|▍         | 30/602 [00:03<01:09,  8.20it/s][A

Validation DataLoader 0:   5%|▌         | 31/602 [00:03<01:07,  8.44it/s][A

Validation DataLoader 0:   5%|▌         | 32/602 [00:04<01:12,  7.81it/s][A

Validation DataLoader 0:   5%|▌         | 33/602 [00:04<01:10,  8.03it/s][A

Validation DataLoader 0:   6%|▌         | 34/602 [00:04<01:10,  8.10it/s][A

Validation DataLoader 0:   6%|▌         | 35/602 [00:04<01:13,  7.68it/s][A

Validation DataLoader 0:   6%|▌         | 36/602 [00:04<01:11,  7.86it/s][A

Validation DataLoader 0:   6%|▌         | 37/602 [00:04<01:10,  8.01it/s][A

Validation DataLoader 0:   6%|▋         | 38/602 [00:04<01:13,  7.63it/s][A

Validation DataLoader 0:   6%|▋         | 39/602 [00:05<01:12,  7.72it/s][A

Validation DataLoader 0:   7%|▋         | 40/602 [00:05<01:11,  7.89it/s][A

Validation DataLoader 0:   7%|▋         | 41/602 [00:05<01:14,  7.57it/s][A

Validation DataLoader 0:   7%|▋         | 42/602 [00:05<01:13,  7.66it/s][A

Validation DataLoader 0:   7%|▋         | 43/602 [00:05<01:11,  7.82it/s][A

Validation DataLoader 0:   7%|▋         | 44/602 [00:05<01:13,  7.56it/s][A

Validation DataLoader 0:   7%|▋         | 45/602 [00:06<01:14,  7.49it/s][A

Validation DataLoader 0:   8%|▊         | 46/602 [00:06<01:12,  7.64it/s][A

Validation DataLoader 0:   8%|▊         | 47/602 [00:06<01:14,  7.45it/s][A

Validation DataLoader 0:   8%|▊         | 48/602 [00:06<01:14,  7.39it/s][A

Validation DataLoader 0:   8%|▊         | 49/602 [00:06<01:13,  7.53it/s][A

Validation DataLoader 0:   8%|▊         | 50/602 [00:06<01:14,  7.44it/s][A

Validation DataLoader 0:   8%|▊         | 51/602 [00:06<01:14,  7.40it/s][A

Validation DataLoader 0:   9%|▊         | 52/602 [00:06<01:13,  7.53it/s][A

Validation DataLoader 0:   9%|▉         | 53/602 [00:07<01:14,  7.38it/s][A

Validation DataLoader 0:   9%|▉         | 54/602 [00:07<01:13,  7.45it/s][A

Validation DataLoader 0:   9%|▉         | 55/602 [00:07<01:12,  7.56it/s][A

Validation DataLoader 0:   9%|▉         | 56/602 [00:07<01:14,  7.31it/s][A

Validation DataLoader 0:   9%|▉         | 57/602 [00:07<01:13,  7.41it/s][A

Validation DataLoader 0:  10%|▉         | 58/602 [00:07<01:12,  7.52it/s][A

Validation DataLoader 0:  10%|▉         | 59/602 [00:08<01:14,  7.28it/s][A

Validation DataLoader 0:  10%|▉         | 60/602 [00:08<01:13,  7.36it/s][A

Validation DataLoader 0:  10%|█         | 61/602 [00:08<01:12,  7.47it/s][A

Validation DataLoader 0:  10%|█         | 62/602 [00:08<01:13,  7.30it/s][A

Validation DataLoader 0:  10%|█         | 63/602 [00:08<01:13,  7.30it/s][A

Validation DataLoader 0:  11%|█         | 64/602 [00:08<01:12,  7.38it/s][A

Validation DataLoader 0:  11%|█         | 65/602 [00:08<01:14,  7.23it/s][A

Validation DataLoader 0:  11%|█         | 66/602 [00:09<01:13,  7.27it/s][A

Validation DataLoader 0:  11%|█         | 67/602 [00:09<01:12,  7.35it/s][A

Validation DataLoader 0:  11%|█▏        | 68/602 [00:09<01:13,  7.23it/s][A

Validation DataLoader 0:  11%|█▏        | 69/602 [00:09<01:12,  7.30it/s][A

Validation DataLoader 0:  12%|█▏        | 70/602 [00:09<01:12,  7.32it/s][A

Validation DataLoader 0:  12%|█▏        | 71/602 [00:09<01:13,  7.24it/s][A

Validation DataLoader 0:  12%|█▏        | 72/602 [00:09<01:12,  7.28it/s][A

Validation DataLoader 0:  12%|█▏        | 73/602 [00:10<01:12,  7.30it/s][A

Validation DataLoader 0:  12%|█▏        | 74/602 [00:10<01:13,  7.20it/s][A

Validation DataLoader 0:  12%|█▏        | 75/602 [00:10<01:12,  7.29it/s][A

Validation DataLoader 0:  13%|█▎        | 76/602 [00:10<01:12,  7.28it/s][A

Validation DataLoader 0:  13%|█▎        | 77/602 [00:10<01:12,  7.20it/s][A

Validation DataLoader 0:  13%|█▎        | 78/602 [00:10<01:11,  7.28it/s][A

Validation DataLoader 0:  13%|█▎        | 79/602 [00:10<01:11,  7.29it/s][A

Validation DataLoader 0:  13%|█▎        | 80/602 [00:11<01:12,  7.19it/s][A

Validation DataLoader 0:  13%|█▎        | 81/602 [00:11<01:11,  7.24it/s][A

Validation DataLoader 0:  14%|█▎        | 82/602 [00:11<01:11,  7.29it/s][A

Validation DataLoader 0:  14%|█▍        | 83/602 [00:11<01:12,  7.17it/s][A

Validation DataLoader 0:  14%|█▍        | 84/602 [00:11<01:11,  7.20it/s][A

Validation DataLoader 0:  14%|█▍        | 85/602 [00:11<01:11,  7.27it/s][A

Validation DataLoader 0:  14%|█▍        | 86/602 [00:11<01:11,  7.22it/s][A

Validation DataLoader 0:  14%|█▍        | 87/602 [00:12<01:11,  7.18it/s][A

Validation DataLoader 0:  15%|█▍        | 88/602 [00:12<01:10,  7.26it/s][A

Validation DataLoader 0:  15%|█▍        | 89/602 [00:12<01:11,  7.18it/s][A

Validation DataLoader 0:  15%|█▍        | 90/602 [00:12<01:11,  7.17it/s][A

Validation DataLoader 0:  15%|█▌        | 91/602 [00:12<01:10,  7.22it/s][A

Validation DataLoader 0:  15%|█▌        | 92/602 [00:12<01:11,  7.13it/s][A

Validation DataLoader 0:  15%|█▌        | 93/602 [00:13<01:11,  7.14it/s][A

Validation DataLoader 0:  16%|█▌        | 94/602 [00:13<01:10,  7.18it/s][A

Validation DataLoader 0:  16%|█▌        | 95/602 [00:13<01:11,  7.11it/s][A

Validation DataLoader 0:  16%|█▌        | 96/602 [00:13<01:11,  7.12it/s][A

Validation DataLoader 0:  16%|█▌        | 97/602 [00:13<01:10,  7.16it/s][A

Validation DataLoader 0:  16%|█▋        | 98/602 [00:13<01:10,  7.13it/s][A

Validation DataLoader 0:  16%|█▋        | 99/602 [00:13<01:10,  7.14it/s][A

Validation DataLoader 0:  17%|█▋        | 100/602 [00:14<01:10,  7.11it/s][A

Validation DataLoader 0:  17%|█▋        | 101/602 [00:14<01:10,  7.13it/s][A

Validation DataLoader 0:  17%|█▋        | 102/602 [00:14<01:10,  7.10it/s][A

Validation DataLoader 0:  17%|█▋        | 103/602 [00:14<01:10,  7.13it/s][A

Validation DataLoader 0:  17%|█▋        | 104/602 [00:14<01:09,  7.12it/s][A

Validation DataLoader 0:  17%|█▋        | 105/602 [00:14<01:09,  7.10it/s][A

Validation DataLoader 0:  18%|█▊        | 106/602 [00:14<01:09,  7.10it/s][A

Validation DataLoader 0:  18%|█▊        | 107/602 [00:15<01:09,  7.11it/s][A

Validation DataLoader 0:  18%|█▊        | 108/602 [00:15<01:09,  7.11it/s][A

Validation DataLoader 0:  18%|█▊        | 109/602 [00:15<01:09,  7.06it/s][A

Validation DataLoader 0:  18%|█▊        | 110/602 [00:15<01:09,  7.06it/s][A

Validation DataLoader 0:  18%|█▊        | 111/602 [00:15<01:09,  7.10it/s][A

Validation DataLoader 0:  19%|█▊        | 112/602 [00:15<01:09,  7.03it/s][A

Validation DataLoader 0:  19%|█▉        | 113/602 [00:15<01:09,  7.09it/s][A

Validation DataLoader 0:  19%|█▉        | 114/602 [00:16<01:08,  7.10it/s][A

Validation DataLoader 0:  19%|█▉        | 115/602 [00:16<01:09,  7.03it/s][A

Validation DataLoader 0:  19%|█▉        | 116/602 [00:16<01:08,  7.05it/s][A

Validation DataLoader 0:  19%|█▉        | 117/602 [00:16<01:08,  7.09it/s][A

Validation DataLoader 0:  20%|█▉        | 118/602 [00:16<01:08,  7.01it/s][A

Validation DataLoader 0:  20%|█▉        | 119/602 [00:16<01:08,  7.06it/s][A

Validation DataLoader 0:  20%|█▉        | 120/602 [00:16<01:07,  7.11it/s][A

Validation DataLoader 0:  20%|██        | 121/602 [00:17<01:08,  7.02it/s][A

Validation DataLoader 0:  20%|██        | 122/602 [00:17<01:08,  7.05it/s][A

Validation DataLoader 0:  20%|██        | 123/602 [00:17<01:07,  7.11it/s][A

Validation DataLoader 0:  21%|██        | 124/602 [00:17<01:08,  6.98it/s][A

Validation DataLoader 0:  21%|██        | 125/602 [00:17<01:07,  7.03it/s][A

Validation DataLoader 0:  21%|██        | 126/602 [00:17<01:07,  7.08it/s][A

Validation DataLoader 0:  21%|██        | 127/602 [00:18<01:08,  6.96it/s][A

Validation DataLoader 0:  21%|██▏       | 128/602 [00:18<01:07,  7.01it/s][A

Validation DataLoader 0:  21%|██▏       | 129/602 [00:18<01:07,  7.05it/s][A

Validation DataLoader 0:  22%|██▏       | 130/602 [00:18<01:07,  6.94it/s][A

Validation DataLoader 0:  22%|██▏       | 131/602 [00:18<01:07,  6.99it/s][A

Validation DataLoader 0:  22%|██▏       | 132/602 [00:18<01:06,  7.04it/s][A

Validation DataLoader 0:  22%|██▏       | 133/602 [00:19<01:07,  6.93it/s][A

Validation DataLoader 0:  22%|██▏       | 134/602 [00:19<01:07,  6.98it/s][A

Validation DataLoader 0:  22%|██▏       | 135/602 [00:19<01:06,  7.02it/s][A

Validation DataLoader 0:  23%|██▎       | 136/602 [00:19<01:07,  6.92it/s][A

Validation DataLoader 0:  23%|██▎       | 137/602 [00:19<01:06,  6.96it/s][A

Validation DataLoader 0:  23%|██▎       | 138/602 [00:19<01:06,  7.01it/s][A

Validation DataLoader 0:  23%|██▎       | 139/602 [00:20<01:06,  6.91it/s][A

Validation DataLoader 0:  23%|██▎       | 140/602 [00:20<01:06,  6.96it/s][A

Validation DataLoader 0:  23%|██▎       | 141/602 [00:20<01:05,  7.00it/s][A

Validation DataLoader 0:  24%|██▎       | 142/602 [00:20<01:06,  6.89it/s][A

Validation DataLoader 0:  24%|██▍       | 143/602 [00:20<01:06,  6.93it/s][A

Validation DataLoader 0:  24%|██▍       | 144/602 [00:20<01:05,  6.98it/s][A

Validation DataLoader 0:  24%|██▍       | 145/602 [00:21<01:06,  6.90it/s][A

Validation DataLoader 0:  24%|██▍       | 146/602 [00:21<01:05,  6.95it/s][A

Validation DataLoader 0:  24%|██▍       | 147/602 [00:21<01:05,  6.99it/s][A

Validation DataLoader 0:  25%|██▍       | 148/602 [00:21<01:05,  6.88it/s][A

Validation DataLoader 0:  25%|██▍       | 149/602 [00:21<01:05,  6.92it/s][A

Validation DataLoader 0:  25%|██▍       | 150/602 [00:21<01:04,  6.96it/s][A

Validation DataLoader 0:  25%|██▌       | 151/602 [00:21<01:05,  6.87it/s][A

Validation DataLoader 0:  25%|██▌       | 152/602 [00:22<01:05,  6.91it/s][A

Validation DataLoader 0:  25%|██▌       | 153/602 [00:22<01:04,  6.95it/s][A

Validation DataLoader 0:  26%|██▌       | 154/602 [00:22<01:05,  6.88it/s][A

Validation DataLoader 0:  26%|██▌       | 155/602 [00:22<01:04,  6.92it/s][A

Validation DataLoader 0:  26%|██▌       | 156/602 [00:22<01:04,  6.96it/s][A

Validation DataLoader 0:  26%|██▌       | 157/602 [00:22<01:04,  6.90it/s][A

Validation DataLoader 0:  26%|██▌       | 158/602 [00:22<01:04,  6.94it/s][A

Validation DataLoader 0:  26%|██▋       | 159/602 [00:22<01:03,  6.98it/s][A

Validation DataLoader 0:  27%|██▋       | 160/602 [00:23<01:04,  6.88it/s][A

Validation DataLoader 0:  27%|██▋       | 161/602 [00:23<01:03,  6.92it/s][A

Validation DataLoader 0:  27%|██▋       | 162/602 [00:23<01:03,  6.96it/s][A

Validation DataLoader 0:  27%|██▋       | 163/602 [00:23<01:03,  6.88it/s][A

Validation DataLoader 0:  27%|██▋       | 164/602 [00:23<01:03,  6.92it/s][A

Validation DataLoader 0:  27%|██▋       | 165/602 [00:23<01:02,  6.95it/s][A

Validation DataLoader 0:  28%|██▊       | 166/602 [00:24<01:03,  6.86it/s][A

Validation DataLoader 0:  28%|██▊       | 167/602 [00:24<01:03,  6.90it/s][A

Validation DataLoader 0:  28%|██▊       | 168/602 [00:24<01:02,  6.94it/s][A

Validation DataLoader 0:  28%|██▊       | 169/602 [00:24<01:03,  6.85it/s][A

Validation DataLoader 0:  28%|██▊       | 170/602 [00:24<01:02,  6.88it/s][A

Validation DataLoader 0:  28%|██▊       | 171/602 [00:24<01:02,  6.92it/s][A

Validation DataLoader 0:  29%|██▊       | 172/602 [00:25<01:02,  6.85it/s][A

Validation DataLoader 0:  29%|██▊       | 173/602 [00:25<01:02,  6.89it/s][A

Validation DataLoader 0:  29%|██▉       | 174/602 [00:25<01:01,  6.93it/s][A

Validation DataLoader 0:  29%|██▉       | 175/602 [00:25<01:02,  6.85it/s][A

Validation DataLoader 0:  29%|██▉       | 176/602 [00:25<01:01,  6.88it/s][A

Validation DataLoader 0:  29%|██▉       | 177/602 [00:25<01:01,  6.92it/s][A

Validation DataLoader 0:  30%|██▉       | 178/602 [00:25<01:01,  6.85it/s][A

Validation DataLoader 0:  30%|██▉       | 179/602 [00:26<01:01,  6.88it/s][A

Validation DataLoader 0:  30%|██▉       | 180/602 [00:26<01:01,  6.92it/s][A

Validation DataLoader 0:  30%|███       | 181/602 [00:26<01:01,  6.85it/s][A

Validation DataLoader 0:  30%|███       | 182/602 [00:26<01:00,  6.89it/s][A

Validation DataLoader 0:  30%|███       | 183/602 [00:26<01:00,  6.92it/s][A

Validation DataLoader 0:  31%|███       | 184/602 [00:26<01:01,  6.85it/s][A

Validation DataLoader 0:  31%|███       | 185/602 [00:26<01:00,  6.88it/s][A

Validation DataLoader 0:  31%|███       | 186/602 [00:26<01:00,  6.91it/s][A

Validation DataLoader 0:  31%|███       | 187/602 [00:27<01:00,  6.84it/s][A

Validation DataLoader 0:  31%|███       | 188/602 [00:27<01:00,  6.87it/s][A

Validation DataLoader 0:  31%|███▏      | 189/602 [00:27<00:59,  6.91it/s][A

Validation DataLoader 0:  32%|███▏      | 190/602 [00:27<01:00,  6.83it/s][A

Validation DataLoader 0:  32%|███▏      | 191/602 [00:27<00:59,  6.86it/s][A

Validation DataLoader 0:  32%|███▏      | 192/602 [00:27<00:59,  6.89it/s][A

Validation DataLoader 0:  32%|███▏      | 193/602 [00:28<00:59,  6.83it/s][A

Validation DataLoader 0:  32%|███▏      | 194/602 [00:28<00:59,  6.86it/s][A

Validation DataLoader 0:  32%|███▏      | 195/602 [00:28<00:59,  6.89it/s][A

Validation DataLoader 0:  33%|███▎      | 196/602 [00:28<00:59,  6.83it/s][A

Validation DataLoader 0:  33%|███▎      | 197/602 [00:28<00:59,  6.86it/s][A

Validation DataLoader 0:  33%|███▎      | 198/602 [00:28<00:58,  6.89it/s][A

Validation DataLoader 0:  33%|███▎      | 199/602 [00:29<00:58,  6.84it/s][A

Validation DataLoader 0:  33%|███▎      | 200/602 [00:29<00:58,  6.87it/s][A

Validation DataLoader 0:  33%|███▎      | 201/602 [00:29<00:58,  6.90it/s][A

Validation DataLoader 0:  34%|███▎      | 202/602 [00:29<00:58,  6.83it/s][A

Validation DataLoader 0:  34%|███▎      | 203/602 [00:29<00:58,  6.86it/s][A

Validation DataLoader 0:  34%|███▍      | 204/602 [00:29<00:57,  6.89it/s][A

Validation DataLoader 0:  34%|███▍      | 205/602 [00:29<00:58,  6.84it/s][A

Validation DataLoader 0:  34%|███▍      | 206/602 [00:29<00:57,  6.87it/s][A

Validation DataLoader 0:  34%|███▍      | 207/602 [00:29<00:57,  6.90it/s][A

Validation DataLoader 0:  35%|███▍      | 208/602 [00:30<00:57,  6.85it/s][A

Validation DataLoader 0:  35%|███▍      | 209/602 [00:30<00:57,  6.88it/s][A

Validation DataLoader 0:  35%|███▍      | 210/602 [00:30<00:56,  6.91it/s][A

Validation DataLoader 0:  35%|███▌      | 211/602 [00:30<00:57,  6.85it/s][A

Validation DataLoader 0:  35%|███▌      | 212/602 [00:30<00:56,  6.88it/s][A

Validation DataLoader 0:  35%|███▌      | 213/602 [00:30<00:56,  6.91it/s][A

Validation DataLoader 0:  36%|███▌      | 214/602 [00:31<00:56,  6.85it/s][A

Validation DataLoader 0:  36%|███▌      | 215/602 [00:31<00:56,  6.88it/s][A

Validation DataLoader 0:  36%|███▌      | 216/602 [00:31<00:55,  6.91it/s][A

Validation DataLoader 0:  36%|███▌      | 217/602 [00:31<00:56,  6.85it/s][A

Validation DataLoader 0:  36%|███▌      | 218/602 [00:31<00:55,  6.87it/s][A

Validation DataLoader 0:  36%|███▋      | 219/602 [00:31<00:55,  6.90it/s][A

Validation DataLoader 0:  37%|███▋      | 220/602 [00:32<00:55,  6.85it/s][A

Validation DataLoader 0:  37%|███▋      | 221/602 [00:32<00:55,  6.88it/s][A

Validation DataLoader 0:  37%|███▋      | 222/602 [00:32<00:55,  6.91it/s][A

Validation DataLoader 0:  37%|███▋      | 223/602 [00:32<00:55,  6.84it/s][A

Validation DataLoader 0:  37%|███▋      | 224/602 [00:32<00:55,  6.87it/s][A

Validation DataLoader 0:  37%|███▋      | 225/602 [00:32<00:54,  6.89it/s][A

Validation DataLoader 0:  38%|███▊      | 226/602 [00:33<00:55,  6.83it/s][A

Validation DataLoader 0:  38%|███▊      | 227/602 [00:33<00:54,  6.86it/s][A

Validation DataLoader 0:  38%|███▊      | 228/602 [00:33<00:54,  6.89it/s][A

Validation DataLoader 0:  38%|███▊      | 229/602 [00:33<00:54,  6.85it/s][A

Validation DataLoader 0:  38%|███▊      | 230/602 [00:33<00:54,  6.88it/s][A

Validation DataLoader 0:  38%|███▊      | 231/602 [00:33<00:53,  6.90it/s][A

Validation DataLoader 0:  39%|███▊      | 232/602 [00:33<00:54,  6.85it/s][A

Validation DataLoader 0:  39%|███▊      | 233/602 [00:33<00:53,  6.88it/s][A

Validation DataLoader 0:  39%|███▉      | 234/602 [00:33<00:53,  6.90it/s][A

Validation DataLoader 0:  39%|███▉      | 235/602 [00:34<00:53,  6.83it/s][A

Validation DataLoader 0:  39%|███▉      | 236/602 [00:34<00:53,  6.86it/s][A

Validation DataLoader 0:  39%|███▉      | 237/602 [00:34<00:53,  6.89it/s][A

Validation DataLoader 0:  40%|███▉      | 238/602 [00:34<00:53,  6.83it/s][A

Validation DataLoader 0:  40%|███▉      | 239/602 [00:34<00:52,  6.86it/s][A

Validation DataLoader 0:  40%|███▉      | 240/602 [00:34<00:52,  6.88it/s][A

Validation DataLoader 0:  40%|████      | 241/602 [00:35<00:52,  6.83it/s][A

Validation DataLoader 0:  40%|████      | 242/602 [00:35<00:52,  6.85it/s][A

Validation DataLoader 0:  40%|████      | 243/602 [00:35<00:52,  6.88it/s][A

Validation DataLoader 0:  41%|████      | 244/602 [00:35<00:52,  6.81it/s][A

Validation DataLoader 0:  41%|████      | 245/602 [00:35<00:52,  6.84it/s][A

Validation DataLoader 0:  41%|████      | 246/602 [00:35<00:51,  6.86it/s][A

Validation DataLoader 0:  41%|████      | 247/602 [00:36<00:52,  6.81it/s][A

Validation DataLoader 0:  41%|████      | 248/602 [00:36<00:51,  6.84it/s][A

Validation DataLoader 0:  41%|████▏     | 249/602 [00:36<00:51,  6.86it/s][A

Validation DataLoader 0:  42%|████▏     | 250/602 [00:36<00:51,  6.81it/s][A

Validation DataLoader 0:  42%|████▏     | 251/602 [00:36<00:51,  6.84it/s][A

Validation DataLoader 0:  42%|████▏     | 252/602 [00:36<00:51,  6.86it/s][A

Validation DataLoader 0:  42%|████▏     | 253/602 [00:37<00:51,  6.81it/s][A

Validation DataLoader 0:  42%|████▏     | 254/602 [00:37<00:50,  6.83it/s][A

Validation DataLoader 0:  42%|████▏     | 255/602 [00:37<00:50,  6.85it/s][A

Validation DataLoader 0:  43%|████▎     | 256/602 [00:37<00:50,  6.81it/s][A

Validation DataLoader 0:  43%|████▎     | 257/602 [00:37<00:50,  6.83it/s][A

Validation DataLoader 0:  43%|████▎     | 258/602 [00:37<00:50,  6.86it/s][A

Validation DataLoader 0:  43%|████▎     | 259/602 [00:38<00:50,  6.81it/s][A

Validation DataLoader 0:  43%|████▎     | 260/602 [00:38<00:50,  6.83it/s][A

Validation DataLoader 0:  43%|████▎     | 261/602 [00:38<00:49,  6.85it/s][A

Validation DataLoader 0:  44%|████▎     | 262/602 [00:38<00:49,  6.82it/s][A

Validation DataLoader 0:  44%|████▎     | 263/602 [00:38<00:49,  6.84it/s][A

Validation DataLoader 0:  44%|████▍     | 264/602 [00:38<00:49,  6.86it/s][A

Validation DataLoader 0:  44%|████▍     | 265/602 [00:38<00:49,  6.81it/s][A

Validation DataLoader 0:  44%|████▍     | 266/602 [00:38<00:49,  6.84it/s][A

Validation DataLoader 0:  44%|████▍     | 267/602 [00:38<00:48,  6.86it/s][A

Validation DataLoader 0:  45%|████▍     | 268/602 [00:39<00:48,  6.82it/s][A

Validation DataLoader 0:  45%|████▍     | 269/602 [00:39<00:48,  6.84it/s][A

Validation DataLoader 0:  45%|████▍     | 270/602 [00:39<00:48,  6.86it/s][A

Validation DataLoader 0:  45%|████▌     | 271/602 [00:39<00:48,  6.83it/s][A

Validation DataLoader 0:  45%|████▌     | 272/602 [00:39<00:48,  6.85it/s][A

Validation DataLoader 0:  45%|████▌     | 273/602 [00:39<00:47,  6.87it/s][A

Validation DataLoader 0:  46%|████▌     | 274/602 [00:40<00:48,  6.83it/s][A

Validation DataLoader 0:  46%|████▌     | 275/602 [00:40<00:47,  6.85it/s][A

Validation DataLoader 0:  46%|████▌     | 276/602 [00:40<00:47,  6.87it/s][A

Validation DataLoader 0:  46%|████▌     | 277/602 [00:40<00:47,  6.82it/s][A

Validation DataLoader 0:  46%|████▌     | 278/602 [00:40<00:47,  6.84it/s][A

Validation DataLoader 0:  46%|████▋     | 279/602 [00:40<00:47,  6.86it/s][A

Validation DataLoader 0:  47%|████▋     | 280/602 [00:41<00:47,  6.82it/s][A

Validation DataLoader 0:  47%|████▋     | 281/602 [00:41<00:46,  6.84it/s][A

Validation DataLoader 0:  47%|████▋     | 282/602 [00:41<00:46,  6.86it/s][A

Validation DataLoader 0:  47%|████▋     | 283/602 [00:41<00:46,  6.81it/s][A

Validation DataLoader 0:  47%|████▋     | 284/602 [00:41<00:46,  6.83it/s][A

Validation DataLoader 0:  47%|████▋     | 285/602 [00:41<00:46,  6.85it/s][A

Validation DataLoader 0:  48%|████▊     | 286/602 [00:41<00:46,  6.81it/s][A

Validation DataLoader 0:  48%|████▊     | 287/602 [00:42<00:46,  6.83it/s][A

Validation DataLoader 0:  48%|████▊     | 288/602 [00:42<00:45,  6.85it/s][A

Validation DataLoader 0:  48%|████▊     | 289/602 [00:42<00:45,  6.82it/s][A

Validation DataLoader 0:  48%|████▊     | 290/602 [00:42<00:45,  6.84it/s][A

Validation DataLoader 0:  48%|████▊     | 291/602 [00:42<00:45,  6.86it/s][A

Validation DataLoader 0:  49%|████▊     | 292/602 [00:42<00:45,  6.80it/s][A

Validation DataLoader 0:  49%|████▊     | 293/602 [00:42<00:45,  6.82it/s][A

Validation DataLoader 0:  49%|████▉     | 294/602 [00:42<00:44,  6.85it/s][A

Validation DataLoader 0:  49%|████▉     | 295/602 [00:43<00:45,  6.81it/s][A

Validation DataLoader 0:  49%|████▉     | 296/602 [00:43<00:44,  6.83it/s][A

Validation DataLoader 0:  49%|████▉     | 297/602 [00:43<00:44,  6.85it/s][A

Validation DataLoader 0:  50%|████▉     | 298/602 [00:43<00:44,  6.79it/s][A

Validation DataLoader 0:  50%|████▉     | 299/602 [00:43<00:44,  6.81it/s][A

Validation DataLoader 0:  50%|████▉     | 300/602 [00:43<00:44,  6.83it/s][A

Validation DataLoader 0:  50%|█████     | 301/602 [00:44<00:44,  6.79it/s][A

Validation DataLoader 0:  50%|█████     | 302/602 [00:44<00:44,  6.81it/s][A

Validation DataLoader 0:  50%|█████     | 303/602 [00:44<00:43,  6.83it/s][A

Validation DataLoader 0:  50%|█████     | 304/602 [00:44<00:43,  6.78it/s][A

Validation DataLoader 0:  51%|█████     | 305/602 [00:44<00:43,  6.80it/s][A

Validation DataLoader 0:  51%|█████     | 306/602 [00:44<00:43,  6.82it/s][A

Validation DataLoader 0:  51%|█████     | 307/602 [00:45<00:43,  6.79it/s][A

Validation DataLoader 0:  51%|█████     | 308/602 [00:45<00:43,  6.81it/s][A

Validation DataLoader 0:  51%|█████▏    | 309/602 [00:45<00:42,  6.83it/s][A

Validation DataLoader 0:  51%|█████▏    | 310/602 [00:45<00:43,  6.79it/s][A

Validation DataLoader 0:  52%|█████▏    | 311/602 [00:45<00:42,  6.81it/s][A

Validation DataLoader 0:  52%|█████▏    | 312/602 [00:45<00:42,  6.83it/s][A

Validation DataLoader 0:  52%|█████▏    | 313/602 [00:46<00:42,  6.80it/s][A

Validation DataLoader 0:  52%|█████▏    | 314/602 [00:46<00:42,  6.82it/s][A

Validation DataLoader 0:  52%|█████▏    | 315/602 [00:46<00:41,  6.84it/s][A

Validation DataLoader 0:  52%|█████▏    | 316/602 [00:46<00:42,  6.79it/s][A

Validation DataLoader 0:  53%|█████▎    | 317/602 [00:46<00:41,  6.81it/s][A

Validation DataLoader 0:  53%|█████▎    | 318/602 [00:46<00:41,  6.83it/s][A

Validation DataLoader 0:  53%|█████▎    | 319/602 [00:46<00:41,  6.80it/s][A

Validation DataLoader 0:  53%|█████▎    | 320/602 [00:46<00:41,  6.81it/s][A

Validation DataLoader 0:  53%|█████▎    | 321/602 [00:46<00:41,  6.83it/s][A

Validation DataLoader 0:  53%|█████▎    | 322/602 [00:47<00:41,  6.80it/s][A

Validation DataLoader 0:  54%|█████▎    | 323/602 [00:47<00:40,  6.82it/s][A

Validation DataLoader 0:  54%|█████▍    | 324/602 [00:47<00:40,  6.84it/s][A

Validation DataLoader 0:  54%|█████▍    | 325/602 [00:47<00:40,  6.80it/s][A

Validation DataLoader 0:  54%|█████▍    | 326/602 [00:47<00:40,  6.82it/s][A

Validation DataLoader 0:  54%|█████▍    | 327/602 [00:47<00:40,  6.84it/s][A

Validation DataLoader 0:  54%|█████▍    | 328/602 [00:48<00:40,  6.79it/s][A

Validation DataLoader 0:  55%|█████▍    | 329/602 [00:48<00:40,  6.81it/s][A

Validation DataLoader 0:  55%|█████▍    | 330/602 [00:48<00:39,  6.83it/s][A

Validation DataLoader 0:  55%|█████▍    | 331/602 [00:48<00:39,  6.78it/s][A

Validation DataLoader 0:  55%|█████▌    | 332/602 [00:48<00:39,  6.80it/s][A

Validation DataLoader 0:  55%|█████▌    | 333/602 [00:48<00:39,  6.82it/s][A

Validation DataLoader 0:  55%|█████▌    | 334/602 [00:49<00:39,  6.78it/s][A

Validation DataLoader 0:  56%|█████▌    | 335/602 [00:49<00:39,  6.79it/s][A

Validation DataLoader 0:  56%|█████▌    | 336/602 [00:49<00:39,  6.81it/s][A

Validation DataLoader 0:  56%|█████▌    | 337/602 [00:49<00:39,  6.78it/s][A

Validation DataLoader 0:  56%|█████▌    | 338/602 [00:49<00:38,  6.80it/s][A

Validation DataLoader 0:  56%|█████▋    | 339/602 [00:49<00:38,  6.82it/s][A

Validation DataLoader 0:  56%|█████▋    | 340/602 [00:50<00:38,  6.78it/s][A

Validation DataLoader 0:  57%|█████▋    | 341/602 [00:50<00:38,  6.80it/s][A

Validation DataLoader 0:  57%|█████▋    | 342/602 [00:50<00:38,  6.82it/s][A

Validation DataLoader 0:  57%|█████▋    | 343/602 [00:50<00:38,  6.78it/s][A

Validation DataLoader 0:  57%|█████▋    | 344/602 [00:50<00:37,  6.80it/s][A

Validation DataLoader 0:  57%|█████▋    | 345/602 [00:50<00:37,  6.82it/s][A

Validation DataLoader 0:  57%|█████▋    | 346/602 [00:50<00:37,  6.79it/s][A

Validation DataLoader 0:  58%|█████▊    | 347/602 [00:50<00:37,  6.81it/s][A

Validation DataLoader 0:  58%|█████▊    | 348/602 [00:50<00:37,  6.83it/s][A

Validation DataLoader 0:  58%|█████▊    | 349/602 [00:51<00:37,  6.80it/s][A

Validation DataLoader 0:  58%|█████▊    | 350/602 [00:51<00:36,  6.81it/s][A

Validation DataLoader 0:  58%|█████▊    | 351/602 [00:51<00:36,  6.83it/s][A

Validation DataLoader 0:  58%|█████▊    | 352/602 [00:51<00:36,  6.80it/s][A

Validation DataLoader 0:  59%|█████▊    | 353/602 [00:51<00:36,  6.82it/s][A

Validation DataLoader 0:  59%|█████▉    | 354/602 [00:51<00:36,  6.84it/s][A

Validation DataLoader 0:  59%|█████▉    | 355/602 [00:52<00:36,  6.81it/s][A

Validation DataLoader 0:  59%|█████▉    | 356/602 [00:52<00:36,  6.82it/s][A

Validation DataLoader 0:  59%|█████▉    | 357/602 [00:52<00:35,  6.84it/s][A

Validation DataLoader 0:  59%|█████▉    | 358/602 [00:52<00:35,  6.80it/s][A

Validation DataLoader 0:  60%|█████▉    | 359/602 [00:52<00:35,  6.82it/s][A

Validation DataLoader 0:  60%|█████▉    | 360/602 [00:52<00:35,  6.83it/s][A

Validation DataLoader 0:  60%|█████▉    | 361/602 [00:53<00:35,  6.80it/s][A

Validation DataLoader 0:  60%|██████    | 362/602 [00:53<00:35,  6.82it/s][A

Validation DataLoader 0:  60%|██████    | 363/602 [00:53<00:35,  6.83it/s][A

Validation DataLoader 0:  60%|██████    | 364/602 [00:53<00:34,  6.80it/s][A

Validation DataLoader 0:  61%|██████    | 365/602 [00:53<00:34,  6.82it/s][A

Validation DataLoader 0:  61%|██████    | 366/602 [00:53<00:34,  6.84it/s][A

Validation DataLoader 0:  61%|██████    | 367/602 [00:53<00:34,  6.80it/s][A

Validation DataLoader 0:  61%|██████    | 368/602 [00:53<00:34,  6.82it/s][A

Validation DataLoader 0:  61%|██████▏   | 369/602 [00:53<00:34,  6.84it/s][A

Validation DataLoader 0:  61%|██████▏   | 370/602 [00:54<00:34,  6.80it/s][A

Validation DataLoader 0:  62%|██████▏   | 371/602 [00:54<00:33,  6.82it/s][A

Validation DataLoader 0:  62%|██████▏   | 372/602 [00:54<00:33,  6.83it/s][A

Validation DataLoader 0:  62%|██████▏   | 373/602 [00:54<00:33,  6.80it/s][A

Validation DataLoader 0:  62%|██████▏   | 374/602 [00:54<00:33,  6.82it/s][A

Validation DataLoader 0:  62%|██████▏   | 375/602 [00:54<00:33,  6.83it/s][A

Validation DataLoader 0:  62%|██████▏   | 376/602 [00:55<00:33,  6.80it/s][A

Validation DataLoader 0:  63%|██████▎   | 377/602 [00:55<00:33,  6.82it/s][A

Validation DataLoader 0:  63%|██████▎   | 378/602 [00:55<00:32,  6.83it/s][A

Validation DataLoader 0:  63%|██████▎   | 379/602 [00:55<00:32,  6.80it/s][A

Validation DataLoader 0:  63%|██████▎   | 380/602 [00:55<00:32,  6.81it/s][A

Validation DataLoader 0:  63%|██████▎   | 381/602 [00:55<00:32,  6.83it/s][A

Validation DataLoader 0:  63%|██████▎   | 382/602 [00:56<00:32,  6.79it/s][A

Validation DataLoader 0:  64%|██████▎   | 383/602 [00:56<00:32,  6.80it/s][A

Validation DataLoader 0:  64%|██████▍   | 384/602 [00:56<00:31,  6.82it/s][A

Validation DataLoader 0:  64%|██████▍   | 385/602 [00:56<00:31,  6.79it/s][A

Validation DataLoader 0:  64%|██████▍   | 386/602 [00:56<00:31,  6.80it/s][A

Validation DataLoader 0:  64%|██████▍   | 387/602 [00:56<00:31,  6.82it/s][A

Validation DataLoader 0:  64%|██████▍   | 388/602 [00:57<00:31,  6.79it/s][A

Validation DataLoader 0:  65%|██████▍   | 389/602 [00:57<00:31,  6.81it/s][A

Validation DataLoader 0:  65%|██████▍   | 390/602 [00:57<00:31,  6.81it/s][A

Validation DataLoader 0:  65%|██████▍   | 391/602 [00:57<00:31,  6.78it/s][A

Validation DataLoader 0:  65%|██████▌   | 392/602 [00:57<00:30,  6.80it/s][A

Validation DataLoader 0:  65%|██████▌   | 393/602 [00:57<00:30,  6.81it/s][A

Validation DataLoader 0:  65%|██████▌   | 394/602 [00:58<00:30,  6.79it/s][A

Validation DataLoader 0:  66%|██████▌   | 395/602 [00:58<00:30,  6.80it/s][A

Validation DataLoader 0:  66%|██████▌   | 396/602 [00:58<00:30,  6.81it/s][A

Validation DataLoader 0:  66%|██████▌   | 397/602 [00:58<00:30,  6.78it/s][A

Validation DataLoader 0:  66%|██████▌   | 398/602 [00:58<00:30,  6.80it/s][A

Validation DataLoader 0:  66%|██████▋   | 399/602 [00:58<00:29,  6.81it/s][A

Validation DataLoader 0:  66%|██████▋   | 400/602 [00:58<00:29,  6.79it/s][A

Validation DataLoader 0:  67%|██████▋   | 401/602 [00:58<00:29,  6.80it/s][A

Validation DataLoader 0:  67%|██████▋   | 402/602 [00:59<00:29,  6.80it/s][A

Validation DataLoader 0:  67%|██████▋   | 403/602 [00:59<00:29,  6.78it/s][A

Validation DataLoader 0:  67%|██████▋   | 404/602 [00:59<00:29,  6.79it/s][A

Validation DataLoader 0:  67%|██████▋   | 405/602 [00:59<00:28,  6.80it/s][A

Validation DataLoader 0:  67%|██████▋   | 406/602 [00:59<00:28,  6.77it/s][A

Validation DataLoader 0:  68%|██████▊   | 407/602 [00:59<00:28,  6.79it/s][A

Validation DataLoader 0:  68%|██████▊   | 408/602 [00:59<00:28,  6.80it/s][A

Validation DataLoader 0:  68%|██████▊   | 409/602 [01:00<00:28,  6.78it/s][A

Validation DataLoader 0:  68%|██████▊   | 410/602 [01:00<00:28,  6.79it/s][A

Validation DataLoader 0:  68%|██████▊   | 411/602 [01:00<00:28,  6.80it/s][A

Validation DataLoader 0:  68%|██████▊   | 412/602 [01:00<00:28,  6.76it/s][A

Validation DataLoader 0:  69%|██████▊   | 413/602 [01:00<00:27,  6.78it/s][A

Validation DataLoader 0:  69%|██████▉   | 414/602 [01:00<00:27,  6.79it/s][A

Validation DataLoader 0:  69%|██████▉   | 415/602 [01:01<00:27,  6.76it/s][A

Validation DataLoader 0:  69%|██████▉   | 416/602 [01:01<00:27,  6.78it/s][A

Validation DataLoader 0:  69%|██████▉   | 417/602 [01:01<00:27,  6.79it/s][A

Validation DataLoader 0:  69%|██████▉   | 418/602 [01:01<00:27,  6.76it/s][A

Validation DataLoader 0:  70%|██████▉   | 419/602 [01:01<00:27,  6.77it/s][A

Validation DataLoader 0:  70%|██████▉   | 420/602 [01:01<00:26,  6.79it/s][A

Validation DataLoader 0:  70%|██████▉   | 421/602 [01:02<00:26,  6.75it/s][A

Validation DataLoader 0:  70%|███████   | 422/602 [01:02<00:26,  6.76it/s][A

Validation DataLoader 0:  70%|███████   | 423/602 [01:02<00:26,  6.78it/s][A

Validation DataLoader 0:  70%|███████   | 424/602 [01:02<00:26,  6.76it/s][A

Validation DataLoader 0:  71%|███████   | 425/602 [01:02<00:26,  6.77it/s][A

Validation DataLoader 0:  71%|███████   | 426/602 [01:02<00:25,  6.79it/s][A

Validation DataLoader 0:  71%|███████   | 427/602 [01:03<00:25,  6.76it/s][A

Validation DataLoader 0:  71%|███████   | 428/602 [01:03<00:25,  6.77it/s][A

Validation DataLoader 0:  71%|███████▏  | 429/602 [01:03<00:25,  6.79it/s][A

Validation DataLoader 0:  71%|███████▏  | 430/602 [01:03<00:25,  6.76it/s][A

Validation DataLoader 0:  72%|███████▏  | 431/602 [01:03<00:25,  6.77it/s][A

Validation DataLoader 0:  72%|███████▏  | 432/602 [01:03<00:25,  6.78it/s][A

Validation DataLoader 0:  72%|███████▏  | 433/602 [01:04<00:25,  6.76it/s][A

Validation DataLoader 0:  72%|███████▏  | 434/602 [01:04<00:24,  6.77it/s][A

Validation DataLoader 0:  72%|███████▏  | 435/602 [01:04<00:24,  6.79it/s][A

Validation DataLoader 0:  72%|███████▏  | 436/602 [01:04<00:24,  6.75it/s][A

Validation DataLoader 0:  73%|███████▎  | 437/602 [01:04<00:24,  6.76it/s][A

Validation DataLoader 0:  73%|███████▎  | 438/602 [01:04<00:24,  6.77it/s][A

Validation DataLoader 0:  73%|███████▎  | 439/602 [01:05<00:24,  6.75it/s][A

Validation DataLoader 0:  73%|███████▎  | 440/602 [01:05<00:23,  6.76it/s][A

Validation DataLoader 0:  73%|███████▎  | 441/602 [01:05<00:23,  6.77it/s][A

Validation DataLoader 0:  73%|███████▎  | 442/602 [01:05<00:23,  6.75it/s][A

Validation DataLoader 0:  74%|███████▎  | 443/602 [01:05<00:23,  6.76it/s][A

Validation DataLoader 0:  74%|███████▍  | 444/602 [01:05<00:23,  6.78it/s][A

Validation DataLoader 0:  74%|███████▍  | 445/602 [01:05<00:23,  6.76it/s][A

Validation DataLoader 0:  74%|███████▍  | 446/602 [01:05<00:23,  6.77it/s][A

Validation DataLoader 0:  74%|███████▍  | 447/602 [01:05<00:22,  6.79it/s][A

Validation DataLoader 0:  74%|███████▍  | 448/602 [01:06<00:22,  6.76it/s][A

Validation DataLoader 0:  75%|███████▍  | 449/602 [01:06<00:22,  6.77it/s][A

Validation DataLoader 0:  75%|███████▍  | 450/602 [01:06<00:22,  6.78it/s][A

Validation DataLoader 0:  75%|███████▍  | 451/602 [01:06<00:22,  6.76it/s][A

Validation DataLoader 0:  75%|███████▌  | 452/602 [01:06<00:22,  6.77it/s][A

Validation DataLoader 0:  75%|███████▌  | 453/602 [01:06<00:21,  6.79it/s][A

Validation DataLoader 0:  75%|███████▌  | 454/602 [01:07<00:21,  6.77it/s][A

Validation DataLoader 0:  76%|███████▌  | 455/602 [01:07<00:21,  6.78it/s][A

Validation DataLoader 0:  76%|███████▌  | 456/602 [01:07<00:21,  6.79it/s][A

Validation DataLoader 0:  76%|███████▌  | 457/602 [01:07<00:21,  6.76it/s][A

Validation DataLoader 0:  76%|███████▌  | 458/602 [01:07<00:21,  6.77it/s][A

Validation DataLoader 0:  76%|███████▌  | 459/602 [01:07<00:21,  6.79it/s][A

Validation DataLoader 0:  76%|███████▋  | 460/602 [01:08<00:21,  6.75it/s][A

Validation DataLoader 0:  77%|███████▋  | 461/602 [01:08<00:20,  6.77it/s][A

Validation DataLoader 0:  77%|███████▋  | 462/602 [01:08<00:20,  6.78it/s][A

Validation DataLoader 0:  77%|███████▋  | 463/602 [01:08<00:20,  6.76it/s][A

Validation DataLoader 0:  77%|███████▋  | 464/602 [01:08<00:20,  6.77it/s][A

Validation DataLoader 0:  77%|███████▋  | 465/602 [01:08<00:20,  6.79it/s][A

Validation DataLoader 0:  77%|███████▋  | 466/602 [01:08<00:20,  6.76it/s][A

Validation DataLoader 0:  78%|███████▊  | 467/602 [01:08<00:19,  6.77it/s][A

Validation DataLoader 0:  78%|███████▊  | 468/602 [01:08<00:19,  6.78it/s][A

Validation DataLoader 0:  78%|███████▊  | 469/602 [01:09<00:19,  6.76it/s][A

Validation DataLoader 0:  78%|███████▊  | 470/602 [01:09<00:19,  6.77it/s][A

Validation DataLoader 0:  78%|███████▊  | 471/602 [01:09<00:19,  6.78it/s][A

Validation DataLoader 0:  78%|███████▊  | 472/602 [01:09<00:19,  6.76it/s][A

Validation DataLoader 0:  79%|███████▊  | 473/602 [01:09<00:19,  6.77it/s][A

Validation DataLoader 0:  79%|███████▊  | 474/602 [01:09<00:18,  6.79it/s][A

Validation DataLoader 0:  79%|███████▉  | 475/602 [01:10<00:18,  6.77it/s][A

Validation DataLoader 0:  79%|███████▉  | 476/602 [01:10<00:18,  6.78it/s][A

Validation DataLoader 0:  79%|███████▉  | 477/602 [01:10<00:18,  6.79it/s][A

Validation DataLoader 0:  79%|███████▉  | 478/602 [01:10<00:18,  6.77it/s][A

Validation DataLoader 0:  80%|███████▉  | 479/602 [01:10<00:18,  6.78it/s][A

Validation DataLoader 0:  80%|███████▉  | 480/602 [01:10<00:17,  6.79it/s][A

Validation DataLoader 0:  80%|███████▉  | 481/602 [01:11<00:17,  6.77it/s][A

Validation DataLoader 0:  80%|████████  | 482/602 [01:11<00:17,  6.78it/s][A

Validation DataLoader 0:  80%|████████  | 483/602 [01:11<00:17,  6.79it/s][A

Validation DataLoader 0:  80%|████████  | 484/602 [01:11<00:17,  6.76it/s][A

Validation DataLoader 0:  81%|████████  | 485/602 [01:11<00:17,  6.77it/s][A

Validation DataLoader 0:  81%|████████  | 486/602 [01:11<00:17,  6.78it/s][A

Validation DataLoader 0:  81%|████████  | 487/602 [01:11<00:16,  6.77it/s][A

Validation DataLoader 0:  81%|████████  | 488/602 [01:11<00:16,  6.78it/s][A

Validation DataLoader 0:  81%|████████  | 489/602 [01:12<00:16,  6.79it/s][A

Validation DataLoader 0:  81%|████████▏ | 490/602 [01:12<00:16,  6.76it/s][A

Validation DataLoader 0:  82%|████████▏ | 491/602 [01:12<00:16,  6.78it/s][A

Validation DataLoader 0:  82%|████████▏ | 492/602 [01:12<00:16,  6.79it/s][A

Validation DataLoader 0:  82%|████████▏ | 493/602 [01:12<00:16,  6.76it/s][A

Validation DataLoader 0:  82%|████████▏ | 494/602 [01:12<00:15,  6.78it/s][A

Validation DataLoader 0:  82%|████████▏ | 495/602 [01:12<00:15,  6.79it/s][A

Validation DataLoader 0:  82%|████████▏ | 496/602 [01:13<00:15,  6.76it/s][A

Validation DataLoader 0:  83%|████████▎ | 497/602 [01:13<00:15,  6.78it/s][A

Validation DataLoader 0:  83%|████████▎ | 498/602 [01:13<00:15,  6.79it/s][A

Validation DataLoader 0:  83%|████████▎ | 499/602 [01:13<00:15,  6.77it/s][A

Validation DataLoader 0:  83%|████████▎ | 500/602 [01:13<00:15,  6.79it/s][A

Validation DataLoader 0:  83%|████████▎ | 501/602 [01:13<00:14,  6.80it/s][A

Validation DataLoader 0:  83%|████████▎ | 502/602 [01:14<00:14,  6.78it/s][A

Validation DataLoader 0:  84%|████████▎ | 503/602 [01:14<00:14,  6.79it/s][A

Validation DataLoader 0:  84%|████████▎ | 504/602 [01:14<00:14,  6.80it/s][A

Validation DataLoader 0:  84%|████████▍ | 505/602 [01:14<00:14,  6.78it/s][A

Validation DataLoader 0:  84%|████████▍ | 506/602 [01:14<00:14,  6.79it/s][A

Validation DataLoader 0:  84%|████████▍ | 507/602 [01:14<00:13,  6.80it/s][A

Validation DataLoader 0:  84%|████████▍ | 508/602 [01:15<00:13,  6.77it/s][A

Validation DataLoader 0:  85%|████████▍ | 509/602 [01:15<00:13,  6.78it/s][A

Validation DataLoader 0:  85%|████████▍ | 510/602 [01:15<00:13,  6.79it/s][A

Validation DataLoader 0:  85%|████████▍ | 511/602 [01:15<00:13,  6.77it/s][A

Validation DataLoader 0:  85%|████████▌ | 512/602 [01:15<00:13,  6.78it/s][A

Validation DataLoader 0:  85%|████████▌ | 513/602 [01:15<00:13,  6.79it/s][A

Validation DataLoader 0:  85%|████████▌ | 514/602 [01:15<00:13,  6.76it/s][A

Validation DataLoader 0:  86%|████████▌ | 515/602 [01:16<00:12,  6.78it/s][A

Validation DataLoader 0:  86%|████████▌ | 516/602 [01:16<00:12,  6.79it/s][A

Validation DataLoader 0:  86%|████████▌ | 517/602 [01:16<00:12,  6.76it/s][A

Validation DataLoader 0:  86%|████████▌ | 518/602 [01:16<00:12,  6.77it/s][A

Validation DataLoader 0:  86%|████████▌ | 519/602 [01:16<00:12,  6.79it/s][A

Validation DataLoader 0:  86%|████████▋ | 520/602 [01:16<00:12,  6.77it/s][A

Validation DataLoader 0:  87%|████████▋ | 521/602 [01:16<00:11,  6.78it/s][A

Validation DataLoader 0:  87%|████████▋ | 522/602 [01:16<00:11,  6.79it/s][A

Validation DataLoader 0:  87%|████████▋ | 523/602 [01:17<00:11,  6.77it/s][A

Validation DataLoader 0:  87%|████████▋ | 524/602 [01:17<00:11,  6.78it/s][A

Validation DataLoader 0:  87%|████████▋ | 525/602 [01:17<00:11,  6.79it/s][A

Validation DataLoader 0:  87%|████████▋ | 526/602 [01:17<00:11,  6.77it/s][A

Validation DataLoader 0:  88%|████████▊ | 527/602 [01:17<00:11,  6.78it/s][A

Validation DataLoader 0:  88%|████████▊ | 528/602 [01:17<00:10,  6.79it/s][A

Validation DataLoader 0:  88%|████████▊ | 529/602 [01:18<00:10,  6.76it/s][A

Validation DataLoader 0:  88%|████████▊ | 530/602 [01:18<00:10,  6.78it/s][A

Validation DataLoader 0:  88%|████████▊ | 531/602 [01:18<00:10,  6.79it/s][A

Validation DataLoader 0:  88%|████████▊ | 532/602 [01:18<00:10,  6.76it/s][A

Validation DataLoader 0:  89%|████████▊ | 533/602 [01:18<00:10,  6.77it/s][A

Validation DataLoader 0:  89%|████████▊ | 534/602 [01:18<00:10,  6.78it/s][A

Validation DataLoader 0:  89%|████████▉ | 535/602 [01:19<00:09,  6.76it/s][A

Validation DataLoader 0:  89%|████████▉ | 536/602 [01:19<00:09,  6.77it/s][A

Validation DataLoader 0:  89%|████████▉ | 537/602 [01:19<00:09,  6.78it/s][A

Validation DataLoader 0:  89%|████████▉ | 538/602 [01:19<00:09,  6.76it/s][A

Validation DataLoader 0:  90%|████████▉ | 539/602 [01:19<00:09,  6.77it/s][A

Validation DataLoader 0:  90%|████████▉ | 540/602 [01:19<00:09,  6.78it/s][A

Validation DataLoader 0:  90%|████████▉ | 541/602 [01:20<00:09,  6.76it/s][A

Validation DataLoader 0:  90%|█████████ | 542/602 [01:20<00:08,  6.77it/s][A

Validation DataLoader 0:  90%|█████████ | 543/602 [01:20<00:08,  6.78it/s][A

Validation DataLoader 0:  90%|█████████ | 544/602 [01:20<00:08,  6.76it/s][A

Validation DataLoader 0:  91%|█████████ | 545/602 [01:20<00:08,  6.77it/s][A

Validation DataLoader 0:  91%|█████████ | 546/602 [01:20<00:08,  6.78it/s][A

Validation DataLoader 0:  91%|█████████ | 547/602 [01:20<00:08,  6.76it/s][A

Validation DataLoader 0:  91%|█████████ | 548/602 [01:20<00:07,  6.77it/s][A

Validation DataLoader 0:  91%|█████████ | 549/602 [01:20<00:07,  6.78it/s][A

Validation DataLoader 0:  91%|█████████▏| 550/602 [01:21<00:07,  6.76it/s][A

Validation DataLoader 0:  92%|█████████▏| 551/602 [01:21<00:07,  6.77it/s][A

Validation DataLoader 0:  92%|█████████▏| 552/602 [01:21<00:07,  6.78it/s][A

Validation DataLoader 0:  92%|█████████▏| 553/602 [01:21<00:07,  6.76it/s][A

Validation DataLoader 0:  92%|█████████▏| 554/602 [01:21<00:07,  6.77it/s][A

Validation DataLoader 0:  92%|█████████▏| 555/602 [01:21<00:06,  6.77it/s][A

Validation DataLoader 0:  92%|█████████▏| 556/602 [01:22<00:06,  6.76it/s][A

Validation DataLoader 0:  93%|█████████▎| 557/602 [01:22<00:06,  6.77it/s][A

Validation DataLoader 0:  93%|█████████▎| 558/602 [01:22<00:06,  6.78it/s][A

Validation DataLoader 0:  93%|█████████▎| 559/602 [01:22<00:06,  6.76it/s][A

Validation DataLoader 0:  93%|█████████▎| 560/602 [01:22<00:06,  6.77it/s][A

Validation DataLoader 0:  93%|█████████▎| 561/602 [01:22<00:06,  6.77it/s][A

Validation DataLoader 0:  93%|█████████▎| 562/602 [01:23<00:05,  6.76it/s][A

Validation DataLoader 0:  94%|█████████▎| 563/602 [01:23<00:05,  6.77it/s][A

Validation DataLoader 0:  94%|█████████▎| 564/602 [01:23<00:05,  6.78it/s][A

Validation DataLoader 0:  94%|█████████▍| 565/602 [01:23<00:05,  6.76it/s][A

Validation DataLoader 0:  94%|█████████▍| 566/602 [01:23<00:05,  6.77it/s][A

Validation DataLoader 0:  94%|█████████▍| 567/602 [01:23<00:05,  6.78it/s][A

Validation DataLoader 0:  94%|█████████▍| 568/602 [01:23<00:05,  6.76it/s][A

Validation DataLoader 0:  95%|█████████▍| 569/602 [01:23<00:04,  6.78it/s][A

Validation DataLoader 0:  95%|█████████▍| 570/602 [01:24<00:04,  6.78it/s][A

Validation DataLoader 0:  95%|█████████▍| 571/602 [01:24<00:04,  6.77it/s][A

Validation DataLoader 0:  95%|█████████▌| 572/602 [01:24<00:04,  6.78it/s][A

Validation DataLoader 0:  95%|█████████▌| 573/602 [01:24<00:04,  6.78it/s][A

Validation DataLoader 0:  95%|█████████▌| 574/602 [01:24<00:04,  6.77it/s][A

Validation DataLoader 0:  96%|█████████▌| 575/602 [01:24<00:03,  6.78it/s][A

Validation DataLoader 0:  96%|█████████▌| 576/602 [01:25<00:03,  6.77it/s][A

Validation DataLoader 0:  96%|█████████▌| 577/602 [01:25<00:03,  6.77it/s][A

Validation DataLoader 0:  96%|█████████▌| 578/602 [01:25<00:03,  6.78it/s][A

Validation DataLoader 0:  96%|█████████▌| 579/602 [01:25<00:03,  6.77it/s][A

Validation DataLoader 0:  96%|█████████▋| 580/602 [01:25<00:03,  6.77it/s][A

Validation DataLoader 0:  97%|█████████▋| 581/602 [01:25<00:03,  6.78it/s][A

Validation DataLoader 0:  97%|█████████▋| 582/602 [01:25<00:02,  6.77it/s][A

Validation DataLoader 0:  97%|█████████▋| 583/602 [01:26<00:02,  6.76it/s][A

Validation DataLoader 0:  97%|█████████▋| 584/602 [01:26<00:02,  6.77it/s][A

Validation DataLoader 0:  97%|█████████▋| 585/602 [01:26<00:02,  6.76it/s][A

Validation DataLoader 0:  97%|█████████▋| 586/602 [01:26<00:02,  6.77it/s][A

Validation DataLoader 0:  98%|█████████▊| 587/602 [01:26<00:02,  6.78it/s][A

Validation DataLoader 0:  98%|█████████▊| 588/602 [01:26<00:02,  6.77it/s][A

Validation DataLoader 0:  98%|█████████▊| 589/602 [01:27<00:01,  6.77it/s][A

Validation DataLoader 0:  98%|█████████▊| 590/602 [01:27<00:01,  6.78it/s][A

Validation DataLoader 0:  98%|█████████▊| 591/602 [01:27<00:01,  6.77it/s][A

Validation DataLoader 0:  98%|█████████▊| 592/602 [01:27<00:01,  6.76it/s][A

Validation DataLoader 0:  99%|█████████▊| 593/602 [01:27<00:01,  6.77it/s][A

Validation DataLoader 0:  99%|█████████▊| 594/602 [01:27<00:01,  6.77it/s][A

Validation DataLoader 0:  99%|█████████▉| 595/602 [01:27<00:01,  6.77it/s][A

Validation DataLoader 0:  99%|█████████▉| 596/602 [01:27<00:00,  6.78it/s][A

Validation DataLoader 0:  99%|█████████▉| 597/602 [01:28<00:00,  6.77it/s][A

Validation DataLoader 0:  99%|█████████▉| 598/602 [01:28<00:00,  6.77it/s][A

Validation DataLoader 0: 100%|█████████▉| 599/602 [01:28<00:00,  6.78it/s][A

Validation DataLoader 0: 100%|█████████▉| 600/602 [01:28<00:00,  6.76it/s][A

Validation DataLoader 0: 100%|█████████▉| 601/602 [01:28<00:00,  6.77it/s][A

Validation DataLoader 0: 100%|██████████| 602/602 [01:28<00:00,  6.78it/s][A

                                                                          [A
Epoch 1: 100%|██████████| 3295/3295 [35:52<00:00,  1.53it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=1.270, R@1 (train)=0.668]
Epoch 1: 100%|██████████| 3295/3295 [35:52<00:00,  1.53it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]Epoch 1, global step 6590: 'R@1 (val)' reached 0.39456 (best 0.39456), saving model to '/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=01-R@1 (val)=0.395.ckpt' as top 1

Epoch 1:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]           
Epoch 2:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 1/3295 [00:01<1:47:59,  0.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 1/3295 [00:01<1:48:00,  0.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 2/3295 [00:01<54:40,  1.00it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]  
Epoch 2:   0%|          | 2/3295 [00:01<54:41,  1.00it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 3/3295 [00:02<36:58,  1.48it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 3/3295 [00:02<36:58,  1.48it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 4/3295 [00:03<50:53,  1.08it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 4/3295 [00:03<50:53,  1.08it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 5/3295 [00:03<41:01,  1.34it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 5/3295 [00:03<41:01,  1.34it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 6/3295 [00:03<34:25,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 6/3295 [00:03<34:25,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 7/3295 [00:05<44:26,  1.23it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 7/3295 [00:05<44:26,  1.23it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 8/3295 [00:05<39:04,  1.40it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 8/3295 [00:05<39:04,  1.40it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 9/3295 [00:05<34:53,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 9/3295 [00:05<34:53,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 10/3295 [00:07<40:26,  1.35it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 10/3295 [00:07<40:26,  1.35it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 11/3295 [00:07<37:13,  1.47it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 11/3295 [00:07<37:13,  1.47it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 12/3295 [00:07<34:13,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 12/3295 [00:07<34:13,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 13/3295 [00:09<38:48,  1.41it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 13/3295 [00:09<38:48,  1.41it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 14/3295 [00:09<36:16,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 14/3295 [00:09<36:16,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 15/3295 [00:09<33:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 15/3295 [00:09<33:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 16/3295 [00:11<37:51,  1.44it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   0%|          | 16/3295 [00:11<37:51,  1.44it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 17/3295 [00:11<35:43,  1.53it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 17/3295 [00:11<35:43,  1.53it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 18/3295 [00:11<33:48,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 18/3295 [00:11<33:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 19/3295 [00:13<37:37,  1.45it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 19/3295 [00:13<37:37,  1.45it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 20/3295 [00:13<35:49,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 20/3295 [00:13<35:49,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 21/3295 [00:13<34:10,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 21/3295 [00:13<34:10,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 22/3295 [00:15<37:35,  1.45it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 22/3295 [00:15<37:35,  1.45it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 23/3295 [00:15<36:01,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 23/3295 [00:15<36:01,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 24/3295 [00:15<34:34,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 24/3295 [00:15<34:34,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 25/3295 [00:17<37:14,  1.46it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 25/3295 [00:17<37:14,  1.46it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 26/3295 [00:17<35:51,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 26/3295 [00:17<35:51,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 27/3295 [00:17<34:34,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 27/3295 [00:17<34:34,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 28/3295 [00:18<36:43,  1.48it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 28/3295 [00:18<36:43,  1.48it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 29/3295 [00:18<35:29,  1.53it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 29/3295 [00:18<35:29,  1.53it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 30/3295 [00:18<34:21,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 30/3295 [00:18<34:21,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 31/3295 [00:20<36:28,  1.49it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 31/3295 [00:20<36:28,  1.49it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 32/3295 [00:20<35:22,  1.54it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 32/3295 [00:20<35:22,  1.54it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 33/3295 [00:20<34:20,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 33/3295 [00:20<34:20,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 34/3295 [00:22<35:59,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 34/3295 [00:22<35:59,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 35/3295 [00:22<34:59,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 35/3295 [00:22<34:59,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 36/3295 [00:22<34:03,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 36/3295 [00:22<34:03,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 37/3295 [00:24<35:53,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 37/3295 [00:24<35:53,  1.51it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 38/3295 [00:24<34:58,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 38/3295 [00:24<34:58,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 39/3295 [00:24<34:06,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 39/3295 [00:24<34:06,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 40/3295 [00:26<35:46,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 40/3295 [00:26<35:46,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 41/3295 [00:26<34:55,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|          | 41/3295 [00:26<34:55,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 42/3295 [00:26<34:07,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 42/3295 [00:26<34:07,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 43/3295 [00:28<35:33,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 43/3295 [00:28<35:33,  1.52it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 44/3295 [00:28<34:46,  1.56it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 44/3295 [00:28<34:46,  1.56it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 45/3295 [00:28<34:02,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 45/3295 [00:28<34:02,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 46/3295 [00:29<35:11,  1.54it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 46/3295 [00:29<35:11,  1.54it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 47/3295 [00:29<34:28,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 47/3295 [00:29<34:28,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 48/3295 [00:29<33:46,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 48/3295 [00:29<33:46,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 49/3295 [00:31<34:59,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   1%|▏         | 49/3295 [00:31<34:59,  1.55it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 50/3295 [00:31<34:18,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 50/3295 [00:31<34:18,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 51/3295 [00:31<33:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 51/3295 [00:31<33:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 52/3295 [00:33<34:42,  1.56it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 52/3295 [00:33<34:42,  1.56it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 53/3295 [00:33<34:04,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 53/3295 [00:33<34:04,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 54/3295 [00:33<33:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 54/3295 [00:33<33:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 55/3295 [00:35<34:29,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 55/3295 [00:35<34:29,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 56/3295 [00:35<33:53,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 56/3295 [00:35<33:53,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 57/3295 [00:35<33:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 57/3295 [00:35<33:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 58/3295 [00:36<34:22,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 58/3295 [00:36<34:22,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 59/3295 [00:36<33:48,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 59/3295 [00:36<33:48,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 60/3295 [00:37<33:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 60/3295 [00:37<33:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 61/3295 [00:38<34:17,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 61/3295 [00:38<34:17,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 62/3295 [00:38<33:53,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 62/3295 [00:38<33:53,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 63/3295 [00:39<33:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 63/3295 [00:39<33:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 64/3295 [00:40<34:14,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 64/3295 [00:40<34:14,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 65/3295 [00:40<33:44,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 65/3295 [00:40<33:44,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 66/3295 [00:40<33:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 66/3295 [00:40<33:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 67/3295 [00:42<34:10,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 67/3295 [00:42<34:10,  1.57it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 68/3295 [00:42<33:40,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 68/3295 [00:42<33:40,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 69/3295 [00:42<33:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 69/3295 [00:42<33:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 70/3295 [00:44<34:02,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 70/3295 [00:44<34:02,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 71/3295 [00:44<33:35,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 71/3295 [00:44<33:35,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 72/3295 [00:44<33:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 72/3295 [00:44<33:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 73/3295 [00:46<33:59,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 73/3295 [00:46<33:59,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 74/3295 [00:46<33:32,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 74/3295 [00:46<33:32,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 75/3295 [00:46<33:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 75/3295 [00:46<33:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 76/3295 [00:48<33:54,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 76/3295 [00:48<33:54,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 77/3295 [00:48<33:28,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 77/3295 [00:48<33:28,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 78/3295 [00:48<33:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 78/3295 [00:48<33:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 79/3295 [00:49<33:52,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 79/3295 [00:49<33:52,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 80/3295 [00:49<33:27,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 80/3295 [00:49<33:27,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 81/3295 [00:49<33:03,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 81/3295 [00:49<33:03,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 82/3295 [00:51<33:47,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   2%|▏         | 82/3295 [00:51<33:47,  1.58it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 83/3295 [00:51<33:24,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 83/3295 [00:51<33:24,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 84/3295 [00:51<33:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 84/3295 [00:51<33:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 85/3295 [00:53<33:42,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 85/3295 [00:53<33:42,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 86/3295 [00:53<33:19,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 86/3295 [00:53<33:19,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 87/3295 [00:53<32:57,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 87/3295 [00:53<32:57,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 88/3295 [00:55<33:39,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 88/3295 [00:55<33:39,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 89/3295 [00:55<33:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 89/3295 [00:55<33:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 90/3295 [00:55<32:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 90/3295 [00:55<32:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 91/3295 [00:57<33:39,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 91/3295 [00:57<33:39,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 92/3295 [00:57<33:17,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 92/3295 [00:57<33:17,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 93/3295 [00:57<32:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 93/3295 [00:57<32:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 94/3295 [00:59<33:34,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 94/3295 [00:59<33:34,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 95/3295 [00:59<33:13,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 95/3295 [00:59<33:13,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 96/3295 [00:59<32:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 96/3295 [00:59<32:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 97/3295 [01:00<33:30,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 97/3295 [01:00<33:30,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 98/3295 [01:01<33:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 98/3295 [01:01<33:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 99/3295 [01:01<32:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 99/3295 [01:01<32:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 100/3295 [01:02<33:25,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 100/3295 [01:02<33:25,  1.59it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 101/3295 [01:02<33:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 101/3295 [01:02<33:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 102/3295 [01:02<32:50,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 102/3295 [01:02<32:50,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 103/3295 [01:04<33:20,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 103/3295 [01:04<33:20,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 104/3295 [01:04<33:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 104/3295 [01:04<33:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 105/3295 [01:04<32:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 105/3295 [01:04<32:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 106/3295 [01:06<33:17,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 106/3295 [01:06<33:17,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 107/3295 [01:06<33:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 107/3295 [01:06<33:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 108/3295 [01:06<32:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 108/3295 [01:06<32:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 109/3295 [01:08<33:10,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 109/3295 [01:08<33:10,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 110/3295 [01:08<33:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 110/3295 [01:08<33:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 111/3295 [01:08<32:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 111/3295 [01:08<32:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 112/3295 [01:09<33:04,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 112/3295 [01:09<33:04,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 113/3295 [01:10<33:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 113/3295 [01:10<33:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 114/3295 [01:10<32:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 114/3295 [01:10<32:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 115/3295 [01:11<33:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   3%|▎         | 115/3295 [01:11<33:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 116/3295 [01:12<32:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 116/3295 [01:12<32:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 117/3295 [01:12<32:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 117/3295 [01:12<32:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 118/3295 [01:13<32:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 118/3295 [01:13<32:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 119/3295 [01:14<32:59,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 119/3295 [01:14<32:59,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 120/3295 [01:14<32:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 120/3295 [01:14<32:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 121/3295 [01:15<32:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 121/3295 [01:15<32:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 122/3295 [01:15<32:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 122/3295 [01:15<32:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 123/3295 [01:15<32:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▎         | 123/3295 [01:15<32:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 124/3295 [01:16<32:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 124/3295 [01:16<32:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 125/3295 [01:17<32:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 125/3295 [01:17<32:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 126/3295 [01:17<32:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 126/3295 [01:17<32:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 127/3295 [01:18<32:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 127/3295 [01:18<32:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 128/3295 [01:19<32:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 128/3295 [01:19<32:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 129/3295 [01:19<32:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 129/3295 [01:19<32:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 130/3295 [01:20<32:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 130/3295 [01:20<32:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 131/3295 [01:21<32:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 131/3295 [01:21<32:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 132/3295 [01:21<32:34,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 132/3295 [01:21<32:34,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 133/3295 [01:22<32:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 133/3295 [01:22<32:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 134/3295 [01:23<32:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 134/3295 [01:23<32:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 135/3295 [01:23<32:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 135/3295 [01:23<32:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 136/3295 [01:24<32:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 136/3295 [01:24<32:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 137/3295 [01:25<32:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 137/3295 [01:25<32:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 138/3295 [01:25<32:29,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 138/3295 [01:25<32:29,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 139/3295 [01:26<32:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 139/3295 [01:26<32:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 140/3295 [01:26<32:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 140/3295 [01:26<32:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 141/3295 [01:26<32:24,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 141/3295 [01:26<32:24,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 142/3295 [01:27<32:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 142/3295 [01:27<32:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 143/3295 [01:28<32:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 143/3295 [01:28<32:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 144/3295 [01:28<32:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 144/3295 [01:28<32:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 145/3295 [01:29<32:30,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 145/3295 [01:29<32:30,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 146/3295 [01:30<32:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 146/3295 [01:30<32:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 147/3295 [01:30<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 147/3295 [01:30<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 148/3295 [01:31<32:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   4%|▍         | 148/3295 [01:31<32:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 149/3295 [01:32<32:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 149/3295 [01:32<32:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 150/3295 [01:32<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 150/3295 [01:32<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 151/3295 [01:33<32:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 151/3295 [01:33<32:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 152/3295 [01:34<32:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 152/3295 [01:34<32:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 153/3295 [01:34<32:19,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 153/3295 [01:34<32:19,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 154/3295 [01:35<32:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 154/3295 [01:35<32:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 155/3295 [01:36<32:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 155/3295 [01:36<32:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 156/3295 [01:36<32:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 156/3295 [01:36<32:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 157/3295 [01:37<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 157/3295 [01:37<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 158/3295 [01:37<32:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 158/3295 [01:37<32:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 159/3295 [01:38<32:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 159/3295 [01:38<32:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 160/3295 [01:39<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 160/3295 [01:39<32:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 161/3295 [01:39<32:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 161/3295 [01:39<32:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 162/3295 [01:39<32:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 162/3295 [01:39<32:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 163/3295 [01:40<32:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 163/3295 [01:40<32:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 164/3295 [01:41<32:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▍         | 164/3295 [01:41<32:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 165/3295 [01:41<32:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 165/3295 [01:41<32:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 166/3295 [01:42<32:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 166/3295 [01:42<32:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 167/3295 [01:43<32:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 167/3295 [01:43<32:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 168/3295 [01:43<32:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 168/3295 [01:43<32:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 169/3295 [01:44<32:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 169/3295 [01:44<32:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 170/3295 [01:45<32:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 170/3295 [01:45<32:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 171/3295 [01:45<32:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 171/3295 [01:45<32:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 172/3295 [01:46<32:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 172/3295 [01:46<32:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 173/3295 [01:47<32:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 173/3295 [01:47<32:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 174/3295 [01:47<32:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 174/3295 [01:47<32:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 175/3295 [01:48<32:07,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 175/3295 [01:48<32:07,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 176/3295 [01:49<32:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 176/3295 [01:49<32:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 177/3295 [01:49<32:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 177/3295 [01:49<32:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 178/3295 [01:49<32:03,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 178/3295 [01:49<32:03,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 179/3295 [01:50<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 179/3295 [01:50<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 180/3295 [01:50<31:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 180/3295 [01:50<32:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 181/3295 [01:51<32:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   5%|▌         | 181/3295 [01:51<32:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 182/3295 [01:52<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 182/3295 [01:52<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 183/3295 [01:52<31:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 183/3295 [01:52<31:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 184/3295 [01:53<31:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 184/3295 [01:53<31:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 185/3295 [01:54<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 185/3295 [01:54<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 186/3295 [01:54<32:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 186/3295 [01:54<32:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 187/3295 [01:55<31:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 187/3295 [01:55<31:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 188/3295 [01:56<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 188/3295 [01:56<32:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 189/3295 [01:56<31:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 189/3295 [01:56<31:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 190/3295 [01:57<31:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 190/3295 [01:57<31:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 191/3295 [01:58<32:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 191/3295 [01:58<32:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 192/3295 [01:58<31:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 192/3295 [01:58<31:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 193/3295 [01:59<31:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 193/3295 [01:59<31:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 194/3295 [02:00<32:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 194/3295 [02:00<32:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 195/3295 [02:00<31:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 195/3295 [02:00<31:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 196/3295 [02:00<31:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 196/3295 [02:00<31:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 197/3295 [02:02<32:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 197/3295 [02:02<32:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 198/3295 [02:02<31:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 198/3295 [02:02<31:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 199/3295 [02:02<31:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 199/3295 [02:02<31:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 200/3295 [02:04<32:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 200/3295 [02:04<32:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 201/3295 [02:04<31:51,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 201/3295 [02:04<31:51,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 202/3295 [02:04<31:45,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 202/3295 [02:04<31:45,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 203/3295 [02:06<31:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 203/3295 [02:06<31:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 204/3295 [02:06<31:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 204/3295 [02:06<31:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 205/3295 [02:06<31:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▌         | 205/3295 [02:06<31:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 206/3295 [02:07<31:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 206/3295 [02:07<31:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 207/3295 [02:07<31:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 207/3295 [02:07<31:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 208/3295 [02:08<31:40,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 208/3295 [02:08<31:40,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 209/3295 [02:09<31:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 209/3295 [02:09<31:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 210/3295 [02:09<31:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 210/3295 [02:09<31:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 211/3295 [02:09<31:38,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 211/3295 [02:09<31:38,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 212/3295 [02:11<31:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 212/3295 [02:11<31:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 213/3295 [02:11<31:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 213/3295 [02:11<31:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 214/3295 [02:11<31:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   6%|▋         | 214/3295 [02:11<31:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 215/3295 [02:13<31:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 215/3295 [02:13<31:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 216/3295 [02:13<31:44,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 216/3295 [02:13<31:44,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 217/3295 [02:13<31:36,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 217/3295 [02:13<31:36,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 218/3295 [02:15<31:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 218/3295 [02:15<31:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 219/3295 [02:15<31:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 219/3295 [02:15<31:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 220/3295 [02:15<31:36,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 220/3295 [02:15<31:36,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 221/3295 [02:17<31:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 221/3295 [02:17<31:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 222/3295 [02:17<31:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 222/3295 [02:17<31:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 223/3295 [02:17<31:34,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 223/3295 [02:17<31:34,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 224/3295 [02:19<31:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 224/3295 [02:19<31:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 225/3295 [02:19<31:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 225/3295 [02:19<31:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 226/3295 [02:19<31:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 226/3295 [02:19<31:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 227/3295 [02:21<31:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 227/3295 [02:21<31:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 228/3295 [02:21<31:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 228/3295 [02:21<31:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 229/3295 [02:21<31:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 229/3295 [02:21<31:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 230/3295 [02:23<31:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 230/3295 [02:23<31:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 231/3295 [02:23<31:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 231/3295 [02:23<31:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 232/3295 [02:23<31:30,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 232/3295 [02:23<31:30,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 233/3295 [02:24<31:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 233/3295 [02:24<31:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 234/3295 [02:24<31:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 234/3295 [02:24<31:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 235/3295 [02:24<31:27,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 235/3295 [02:24<31:27,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 236/3295 [02:26<31:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 236/3295 [02:26<31:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 237/3295 [02:26<31:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 237/3295 [02:26<31:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 238/3295 [02:26<31:24,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 238/3295 [02:26<31:24,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 239/3295 [02:28<31:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 239/3295 [02:28<31:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 240/3295 [02:28<31:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 240/3295 [02:28<31:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 241/3295 [02:28<31:24,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 241/3295 [02:28<31:24,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 242/3295 [02:30<31:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 242/3295 [02:30<31:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 243/3295 [02:30<31:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 243/3295 [02:30<31:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 244/3295 [02:30<31:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 244/3295 [02:30<31:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 245/3295 [02:32<31:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 245/3295 [02:32<31:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 246/3295 [02:32<31:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 246/3295 [02:32<31:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 247/3295 [02:32<31:21,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   7%|▋         | 247/3295 [02:32<31:21,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 248/3295 [02:34<31:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 248/3295 [02:34<31:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 249/3295 [02:34<31:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 249/3295 [02:34<31:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 250/3295 [02:34<31:19,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 250/3295 [02:34<31:19,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 251/3295 [02:36<31:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 251/3295 [02:36<31:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 252/3295 [02:36<31:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 252/3295 [02:36<31:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 253/3295 [02:36<31:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 253/3295 [02:36<31:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 254/3295 [02:37<31:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 254/3295 [02:37<31:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 255/3295 [02:37<31:21,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 255/3295 [02:37<31:21,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 256/3295 [02:37<31:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 256/3295 [02:37<31:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 257/3295 [02:39<31:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 257/3295 [02:39<31:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 258/3295 [02:39<31:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 258/3295 [02:39<31:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 259/3295 [02:39<31:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 259/3295 [02:39<31:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 260/3295 [02:41<31:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 260/3295 [02:41<31:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 261/3295 [02:41<31:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 261/3295 [02:41<31:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 262/3295 [02:41<31:10,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 262/3295 [02:41<31:10,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 263/3295 [02:43<31:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 263/3295 [02:43<31:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 264/3295 [02:43<31:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 264/3295 [02:43<31:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 265/3295 [02:43<31:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 265/3295 [02:43<31:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 266/3295 [02:45<31:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 266/3295 [02:45<31:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 267/3295 [02:45<31:13,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 267/3295 [02:45<31:13,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 268/3295 [02:45<31:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 268/3295 [02:45<31:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 269/3295 [02:47<31:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 269/3295 [02:47<31:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 270/3295 [02:47<31:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 270/3295 [02:47<31:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 271/3295 [02:47<31:04,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 271/3295 [02:47<31:04,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 272/3295 [02:48<31:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 272/3295 [02:48<31:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 273/3295 [02:48<31:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 273/3295 [02:48<31:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 274/3295 [02:48<31:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 274/3295 [02:48<31:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 275/3295 [02:50<31:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 275/3295 [02:50<31:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 276/3295 [02:50<31:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 276/3295 [02:50<31:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 277/3295 [02:50<31:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 277/3295 [02:50<31:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 278/3295 [02:52<31:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 278/3295 [02:52<31:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 279/3295 [02:52<31:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 279/3295 [02:52<31:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 280/3295 [02:52<31:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   8%|▊         | 280/3295 [02:52<31:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 281/3295 [02:54<31:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 281/3295 [02:54<31:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 282/3295 [02:54<31:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 282/3295 [02:54<31:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 283/3295 [02:54<30:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 283/3295 [02:54<30:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 284/3295 [02:56<31:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 284/3295 [02:56<31:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 285/3295 [02:56<31:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 285/3295 [02:56<31:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 286/3295 [02:56<30:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 286/3295 [02:56<30:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 287/3295 [02:58<31:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 287/3295 [02:58<31:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 288/3295 [02:58<31:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▊         | 288/3295 [02:58<31:00,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 289/3295 [02:58<30:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 289/3295 [02:58<30:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 290/3295 [03:00<31:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 290/3295 [03:00<31:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 291/3295 [03:00<30:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 291/3295 [03:00<30:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 292/3295 [03:00<30:52,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 292/3295 [03:00<30:52,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 293/3295 [03:01<31:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 293/3295 [03:01<31:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 294/3295 [03:01<30:57,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 294/3295 [03:01<30:57,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 295/3295 [03:01<30:50,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 295/3295 [03:01<30:50,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 296/3295 [03:03<31:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 296/3295 [03:03<31:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 297/3295 [03:03<30:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 297/3295 [03:03<30:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 298/3295 [03:03<30:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 298/3295 [03:03<30:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 299/3295 [03:05<30:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 299/3295 [03:05<30:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 300/3295 [03:05<30:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 300/3295 [03:05<30:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 301/3295 [03:05<30:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 301/3295 [03:05<30:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 302/3295 [03:07<30:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 302/3295 [03:07<30:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 303/3295 [03:07<30:51,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 303/3295 [03:07<30:51,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 304/3295 [03:07<30:45,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 304/3295 [03:07<30:45,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 305/3295 [03:09<30:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 305/3295 [03:09<30:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 306/3295 [03:09<30:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 306/3295 [03:09<30:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 307/3295 [03:09<30:44,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 307/3295 [03:09<30:44,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 308/3295 [03:11<30:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 308/3295 [03:11<30:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 309/3295 [03:11<30:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 309/3295 [03:11<30:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 310/3295 [03:11<30:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 310/3295 [03:11<30:43,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 311/3295 [03:13<30:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 311/3295 [03:13<30:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 312/3295 [03:13<30:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 312/3295 [03:13<30:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 313/3295 [03:13<30:41,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:   9%|▉         | 313/3295 [03:13<30:41,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 314/3295 [03:15<30:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 314/3295 [03:15<30:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 315/3295 [03:15<30:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 315/3295 [03:15<30:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 316/3295 [03:15<30:40,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 316/3295 [03:15<30:40,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 317/3295 [03:17<30:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 317/3295 [03:17<30:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 318/3295 [03:17<30:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 318/3295 [03:17<30:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 319/3295 [03:17<30:38,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 319/3295 [03:17<30:38,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 320/3295 [03:18<30:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 320/3295 [03:18<30:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 321/3295 [03:18<30:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 321/3295 [03:18<30:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 322/3295 [03:18<30:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 322/3295 [03:18<30:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 323/3295 [03:20<30:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 323/3295 [03:20<30:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 324/3295 [03:20<30:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 324/3295 [03:20<30:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 325/3295 [03:20<30:36,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 325/3295 [03:20<30:36,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 326/3295 [03:22<30:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 326/3295 [03:22<30:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 327/3295 [03:22<30:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 327/3295 [03:22<30:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 328/3295 [03:22<30:35,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 328/3295 [03:22<30:35,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 329/3295 [03:24<30:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|▉         | 329/3295 [03:24<30:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 330/3295 [03:24<30:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 330/3295 [03:24<30:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 331/3295 [03:24<30:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 331/3295 [03:24<30:33,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 332/3295 [03:26<30:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 332/3295 [03:26<30:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 333/3295 [03:26<30:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 333/3295 [03:26<30:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 334/3295 [03:26<30:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 334/3295 [03:26<30:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 335/3295 [03:28<30:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 335/3295 [03:28<30:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 336/3295 [03:28<30:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 336/3295 [03:28<30:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 337/3295 [03:28<30:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 337/3295 [03:28<30:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 338/3295 [03:30<30:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 338/3295 [03:30<30:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 339/3295 [03:30<30:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 339/3295 [03:30<30:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 340/3295 [03:30<30:29,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 340/3295 [03:30<30:29,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 341/3295 [03:32<30:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 341/3295 [03:32<30:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 342/3295 [03:32<30:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 342/3295 [03:32<30:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 343/3295 [03:32<30:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 343/3295 [03:32<30:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 344/3295 [03:34<30:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 344/3295 [03:34<30:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 345/3295 [03:34<30:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  10%|█         | 345/3295 [03:34<30:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 346/3295 [03:34<30:25,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 346/3295 [03:34<30:25,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 347/3295 [03:35<30:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 347/3295 [03:35<30:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 348/3295 [03:35<30:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 348/3295 [03:35<30:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 349/3295 [03:36<30:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 349/3295 [03:36<30:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 350/3295 [03:37<30:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 350/3295 [03:37<30:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 351/3295 [03:37<30:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 351/3295 [03:37<30:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 352/3295 [03:37<30:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 352/3295 [03:37<30:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 353/3295 [03:39<30:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 353/3295 [03:39<30:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 354/3295 [03:39<30:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 354/3295 [03:39<30:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 355/3295 [03:39<30:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 355/3295 [03:39<30:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 356/3295 [03:41<30:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 356/3295 [03:41<30:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 357/3295 [03:41<30:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 357/3295 [03:41<30:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 358/3295 [03:41<30:17,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 358/3295 [03:41<30:17,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 359/3295 [03:43<30:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 359/3295 [03:43<30:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 360/3295 [03:43<30:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 360/3295 [03:43<30:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 361/3295 [03:43<30:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 361/3295 [03:43<30:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 362/3295 [03:45<30:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 362/3295 [03:45<30:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 363/3295 [03:45<30:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 363/3295 [03:45<30:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 364/3295 [03:45<30:13,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 364/3295 [03:45<30:13,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 365/3295 [03:46<30:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 365/3295 [03:46<30:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 366/3295 [03:46<30:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 366/3295 [03:46<30:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 367/3295 [03:47<30:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 367/3295 [03:47<30:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 368/3295 [03:48<30:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 368/3295 [03:48<30:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 369/3295 [03:48<30:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 369/3295 [03:48<30:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 370/3295 [03:48<30:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█         | 370/3295 [03:48<30:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 371/3295 [03:50<30:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 371/3295 [03:50<30:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 372/3295 [03:50<30:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 372/3295 [03:50<30:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 373/3295 [03:50<30:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 373/3295 [03:50<30:08,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 374/3295 [03:52<30:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 374/3295 [03:52<30:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 375/3295 [03:52<30:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 375/3295 [03:52<30:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 376/3295 [03:52<30:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 376/3295 [03:52<30:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 377/3295 [03:54<30:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 377/3295 [03:54<30:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 378/3295 [03:54<30:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  11%|█▏        | 378/3295 [03:54<30:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 379/3295 [03:54<30:04,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 379/3295 [03:54<30:04,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 380/3295 [03:56<30:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 380/3295 [03:56<30:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 381/3295 [03:56<30:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 381/3295 [03:56<30:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 382/3295 [03:56<30:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 382/3295 [03:56<30:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 383/3295 [03:58<30:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 383/3295 [03:58<30:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 384/3295 [03:58<30:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 384/3295 [03:58<30:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 385/3295 [03:58<30:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 385/3295 [03:58<30:01,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 386/3295 [04:00<30:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 386/3295 [04:00<30:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 387/3295 [04:00<30:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 387/3295 [04:00<30:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 388/3295 [04:00<29:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 388/3295 [04:00<29:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 389/3295 [04:01<30:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 389/3295 [04:01<30:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 390/3295 [04:01<30:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 390/3295 [04:01<30:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 391/3295 [04:01<29:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 391/3295 [04:01<29:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 392/3295 [04:03<30:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 392/3295 [04:03<30:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 393/3295 [04:03<29:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 393/3295 [04:03<29:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 394/3295 [04:03<29:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 394/3295 [04:03<29:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 395/3295 [04:05<30:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 395/3295 [04:05<30:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 396/3295 [04:05<29:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 396/3295 [04:05<29:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 397/3295 [04:05<29:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 397/3295 [04:05<29:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 398/3295 [04:07<30:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 398/3295 [04:07<30:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 399/3295 [04:07<29:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 399/3295 [04:07<29:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 400/3295 [04:07<29:52,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 400/3295 [04:07<29:52,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 401/3295 [04:09<30:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 401/3295 [04:09<30:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 402/3295 [04:09<29:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 402/3295 [04:09<29:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 403/3295 [04:09<29:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 403/3295 [04:09<29:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 404/3295 [04:11<30:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 404/3295 [04:11<30:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 405/3295 [04:11<29:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 405/3295 [04:11<29:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 406/3295 [04:11<29:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 406/3295 [04:11<29:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 407/3295 [04:13<29:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 407/3295 [04:13<29:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 408/3295 [04:13<29:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 408/3295 [04:13<29:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 409/3295 [04:13<29:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 409/3295 [04:13<29:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 410/3295 [04:15<29:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 410/3295 [04:15<29:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 411/3295 [04:15<29:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  12%|█▏        | 411/3295 [04:15<29:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 412/3295 [04:15<29:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 412/3295 [04:15<29:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 413/3295 [04:17<29:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 413/3295 [04:17<29:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 414/3295 [04:17<29:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 414/3295 [04:17<29:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 415/3295 [04:17<29:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 415/3295 [04:17<29:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 416/3295 [04:19<29:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 416/3295 [04:19<29:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 417/3295 [04:19<29:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 417/3295 [04:19<29:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 418/3295 [04:19<29:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 418/3295 [04:19<29:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 419/3295 [04:20<29:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 419/3295 [04:20<29:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 420/3295 [04:20<29:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 420/3295 [04:20<29:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 421/3295 [04:20<29:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 421/3295 [04:20<29:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 422/3295 [04:22<29:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 422/3295 [04:22<29:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 423/3295 [04:22<29:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 423/3295 [04:22<29:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 424/3295 [04:22<29:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 424/3295 [04:22<29:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 425/3295 [04:24<29:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 425/3295 [04:24<29:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 426/3295 [04:24<29:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 426/3295 [04:24<29:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 427/3295 [04:24<29:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 427/3295 [04:24<29:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 428/3295 [04:26<29:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 428/3295 [04:26<29:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 429/3295 [04:26<29:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 429/3295 [04:26<29:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 430/3295 [04:26<29:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 430/3295 [04:26<29:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 431/3295 [04:28<29:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 431/3295 [04:28<29:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 432/3295 [04:28<29:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 432/3295 [04:28<29:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 433/3295 [04:28<29:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 433/3295 [04:28<29:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 434/3295 [04:30<29:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 434/3295 [04:30<29:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 435/3295 [04:30<29:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 435/3295 [04:30<29:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 436/3295 [04:30<29:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 436/3295 [04:30<29:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 437/3295 [04:31<29:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 437/3295 [04:31<29:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 438/3295 [04:31<29:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 438/3295 [04:31<29:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 439/3295 [04:31<29:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 439/3295 [04:31<29:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 440/3295 [04:33<29:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 440/3295 [04:33<29:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 441/3295 [04:33<29:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 441/3295 [04:33<29:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 442/3295 [04:33<29:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 442/3295 [04:33<29:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 443/3295 [04:35<29:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 443/3295 [04:35<29:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 444/3295 [04:35<29:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  13%|█▎        | 444/3295 [04:35<29:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 445/3295 [04:35<29:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 445/3295 [04:35<29:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 446/3295 [04:37<29:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 446/3295 [04:37<29:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 447/3295 [04:37<29:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 447/3295 [04:37<29:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 448/3295 [04:37<29:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 448/3295 [04:37<29:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 449/3295 [04:39<29:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 449/3295 [04:39<29:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 450/3295 [04:39<29:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 450/3295 [04:39<29:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 451/3295 [04:39<29:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 451/3295 [04:39<29:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 452/3295 [04:41<29:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 452/3295 [04:41<29:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 453/3295 [04:41<29:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▎        | 453/3295 [04:41<29:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 454/3295 [04:41<29:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 454/3295 [04:41<29:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 455/3295 [04:43<29:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 455/3295 [04:43<29:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 456/3295 [04:43<29:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 456/3295 [04:43<29:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 457/3295 [04:43<29:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 457/3295 [04:43<29:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 458/3295 [04:44<29:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 458/3295 [04:44<29:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 459/3295 [04:45<29:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 459/3295 [04:45<29:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 460/3295 [04:45<29:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 460/3295 [04:45<29:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 461/3295 [04:46<29:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 461/3295 [04:46<29:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 462/3295 [04:46<29:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 462/3295 [04:46<29:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 463/3295 [04:46<29:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 463/3295 [04:46<29:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 464/3295 [04:48<29:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 464/3295 [04:48<29:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 465/3295 [04:48<29:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 465/3295 [04:48<29:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 466/3295 [04:48<29:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 466/3295 [04:48<29:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 467/3295 [04:50<29:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 467/3295 [04:50<29:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 468/3295 [04:50<29:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 468/3295 [04:50<29:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 469/3295 [04:50<29:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 469/3295 [04:50<29:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 470/3295 [04:52<29:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 470/3295 [04:52<29:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 471/3295 [04:52<29:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 471/3295 [04:52<29:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 472/3295 [04:52<29:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 472/3295 [04:52<29:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 473/3295 [04:54<29:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 473/3295 [04:54<29:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 474/3295 [04:54<29:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 474/3295 [04:54<29:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 475/3295 [04:54<29:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 475/3295 [04:54<29:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 476/3295 [04:56<29:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 476/3295 [04:56<29:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 477/3295 [04:56<29:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  14%|█▍        | 477/3295 [04:56<29:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 478/3295 [04:56<29:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 478/3295 [04:56<29:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 479/3295 [04:58<29:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 479/3295 [04:58<29:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 480/3295 [04:58<29:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 480/3295 [04:58<29:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 481/3295 [04:58<29:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 481/3295 [04:58<29:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 482/3295 [04:59<29:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 482/3295 [04:59<29:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 483/3295 [04:59<29:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 483/3295 [04:59<29:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 484/3295 [04:59<29:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 484/3295 [04:59<29:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 485/3295 [05:01<29:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 485/3295 [05:01<29:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 486/3295 [05:01<29:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 486/3295 [05:01<29:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 487/3295 [05:01<28:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 487/3295 [05:01<28:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 488/3295 [05:03<29:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 488/3295 [05:03<29:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 489/3295 [05:03<29:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 489/3295 [05:03<29:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 490/3295 [05:03<28:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 490/3295 [05:03<28:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 491/3295 [05:05<29:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 491/3295 [05:05<29:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 492/3295 [05:05<28:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 492/3295 [05:05<28:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 493/3295 [05:05<28:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 493/3295 [05:05<28:54,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 494/3295 [05:06<29:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▍        | 494/3295 [05:06<29:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 495/3295 [05:07<28:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 495/3295 [05:07<28:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 496/3295 [05:07<28:52,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 496/3295 [05:07<28:52,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 497/3295 [05:08<28:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 497/3295 [05:08<28:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 498/3295 [05:08<28:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 498/3295 [05:08<28:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 499/3295 [05:08<28:50,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 499/3295 [05:08<28:50,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 500/3295 [05:10<28:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 500/3295 [05:10<28:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 501/3295 [05:10<28:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 501/3295 [05:10<28:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 502/3295 [05:10<28:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 502/3295 [05:10<28:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 503/3295 [05:12<28:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 503/3295 [05:12<28:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 504/3295 [05:12<28:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 504/3295 [05:12<28:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 505/3295 [05:12<28:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 505/3295 [05:12<28:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 506/3295 [05:14<28:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 506/3295 [05:14<28:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 507/3295 [05:14<28:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 507/3295 [05:14<28:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 508/3295 [05:14<28:45,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 508/3295 [05:14<28:45,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 509/3295 [05:16<28:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 509/3295 [05:16<28:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 510/3295 [05:16<28:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  15%|█▌        | 510/3295 [05:16<28:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 511/3295 [05:16<28:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 511/3295 [05:16<28:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 512/3295 [05:18<28:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 512/3295 [05:18<28:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 513/3295 [05:18<28:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 513/3295 [05:18<28:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 514/3295 [05:18<28:41,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 514/3295 [05:18<28:41,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 515/3295 [05:19<28:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 515/3295 [05:19<28:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 516/3295 [05:19<28:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 516/3295 [05:19<28:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 517/3295 [05:20<28:39,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 517/3295 [05:20<28:39,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 518/3295 [05:21<28:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 518/3295 [05:21<28:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 519/3295 [05:21<28:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 519/3295 [05:21<28:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 520/3295 [05:21<28:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 520/3295 [05:21<28:37,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 521/3295 [05:23<28:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 521/3295 [05:23<28:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 522/3295 [05:23<28:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 522/3295 [05:23<28:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 523/3295 [05:23<28:35,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 523/3295 [05:23<28:35,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 524/3295 [05:25<28:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 524/3295 [05:25<28:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 525/3295 [05:25<28:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 525/3295 [05:25<28:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 526/3295 [05:25<28:34,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 526/3295 [05:25<28:34,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 527/3295 [05:27<28:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 527/3295 [05:27<28:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 528/3295 [05:27<28:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 528/3295 [05:27<28:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 529/3295 [05:27<28:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 529/3295 [05:27<28:32,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 530/3295 [05:29<28:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 530/3295 [05:29<28:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 531/3295 [05:29<28:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 531/3295 [05:29<28:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 532/3295 [05:29<28:30,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 532/3295 [05:29<28:30,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 533/3295 [05:31<28:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 533/3295 [05:31<28:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 534/3295 [05:31<28:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 534/3295 [05:31<28:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 535/3295 [05:31<28:28,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▌        | 535/3295 [05:31<28:28,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 536/3295 [05:33<28:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 536/3295 [05:33<28:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 537/3295 [05:33<28:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 537/3295 [05:33<28:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 538/3295 [05:33<28:26,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 538/3295 [05:33<28:26,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 539/3295 [05:34<28:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 539/3295 [05:34<28:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 540/3295 [05:34<28:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 540/3295 [05:34<28:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 541/3295 [05:35<28:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 541/3295 [05:35<28:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 542/3295 [05:36<28:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 542/3295 [05:36<28:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 543/3295 [05:36<28:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  16%|█▋        | 543/3295 [05:36<28:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 544/3295 [05:36<28:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 544/3295 [05:36<28:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 545/3295 [05:38<28:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 545/3295 [05:38<28:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 546/3295 [05:38<28:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 546/3295 [05:38<28:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 547/3295 [05:38<28:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 547/3295 [05:38<28:20,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 548/3295 [05:40<28:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 548/3295 [05:40<28:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 549/3295 [05:40<28:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 549/3295 [05:40<28:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 550/3295 [05:40<28:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 550/3295 [05:40<28:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 551/3295 [05:42<28:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 551/3295 [05:42<28:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 552/3295 [05:42<28:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 552/3295 [05:42<28:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 553/3295 [05:42<28:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 553/3295 [05:42<28:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 554/3295 [05:44<28:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 554/3295 [05:44<28:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 555/3295 [05:44<28:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 555/3295 [05:44<28:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 556/3295 [05:44<28:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 556/3295 [05:44<28:15,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 557/3295 [05:45<28:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 557/3295 [05:45<28:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 558/3295 [05:45<28:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 558/3295 [05:45<28:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 559/3295 [05:45<28:13,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 559/3295 [05:45<28:13,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 560/3295 [05:47<28:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 560/3295 [05:47<28:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 561/3295 [05:47<28:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 561/3295 [05:47<28:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 562/3295 [05:47<28:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 562/3295 [05:47<28:11,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 563/3295 [05:49<28:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 563/3295 [05:49<28:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 564/3295 [05:49<28:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 564/3295 [05:49<28:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 565/3295 [05:49<28:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 565/3295 [05:49<28:09,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 566/3295 [05:51<28:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 566/3295 [05:51<28:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 567/3295 [05:51<28:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 567/3295 [05:51<28:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 568/3295 [05:51<28:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 568/3295 [05:51<28:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 569/3295 [05:53<28:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 569/3295 [05:53<28:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 570/3295 [05:53<28:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 570/3295 [05:53<28:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 571/3295 [05:53<28:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 571/3295 [05:53<28:06,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 572/3295 [05:55<28:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 572/3295 [05:55<28:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 573/3295 [05:55<28:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 573/3295 [05:55<28:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 574/3295 [05:55<28:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 574/3295 [05:55<28:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 575/3295 [05:57<28:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 575/3295 [05:57<28:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 576/3295 [05:57<28:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  17%|█▋        | 576/3295 [05:57<28:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 577/3295 [05:57<28:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 577/3295 [05:57<28:02,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 578/3295 [05:59<28:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 578/3295 [05:59<28:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 579/3295 [05:59<28:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 579/3295 [05:59<28:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 580/3295 [05:59<28:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 580/3295 [05:59<28:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 581/3295 [06:00<28:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 581/3295 [06:00<28:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 582/3295 [06:00<28:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 582/3295 [06:00<28:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 583/3295 [06:00<27:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 583/3295 [06:00<27:58,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 584/3295 [06:02<28:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 584/3295 [06:02<28:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 585/3295 [06:02<28:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 585/3295 [06:02<28:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 586/3295 [06:02<27:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 586/3295 [06:02<27:56,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 587/3295 [06:04<28:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 587/3295 [06:04<28:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 588/3295 [06:04<27:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 588/3295 [06:04<27:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 589/3295 [06:04<27:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 589/3295 [06:04<27:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 590/3295 [06:06<28:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 590/3295 [06:06<28:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 591/3295 [06:06<27:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 591/3295 [06:06<27:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 592/3295 [06:06<27:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 592/3295 [06:06<27:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 593/3295 [06:08<27:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 593/3295 [06:08<27:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 594/3295 [06:08<27:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 594/3295 [06:08<27:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 595/3295 [06:08<27:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 595/3295 [06:08<27:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 596/3295 [06:10<27:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 596/3295 [06:10<27:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 597/3295 [06:10<27:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 597/3295 [06:10<27:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 598/3295 [06:10<27:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 598/3295 [06:10<27:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 599/3295 [06:12<27:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 599/3295 [06:12<27:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 600/3295 [06:12<27:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 600/3295 [06:12<27:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 601/3295 [06:12<27:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 601/3295 [06:12<27:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 602/3295 [06:14<27:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 602/3295 [06:14<27:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 603/3295 [06:14<27:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 603/3295 [06:14<27:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 604/3295 [06:14<27:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 604/3295 [06:14<27:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 605/3295 [06:16<27:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 605/3295 [06:16<27:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 606/3295 [06:16<27:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 606/3295 [06:16<27:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 607/3295 [06:16<27:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 607/3295 [06:16<27:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 608/3295 [06:18<27:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 608/3295 [06:18<27:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 609/3295 [06:18<27:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  18%|█▊        | 609/3295 [06:18<27:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 610/3295 [06:18<27:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 610/3295 [06:18<27:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 611/3295 [06:19<27:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 611/3295 [06:19<27:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 612/3295 [06:19<27:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 612/3295 [06:19<27:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 613/3295 [06:19<27:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 613/3295 [06:19<27:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 614/3295 [06:21<27:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 614/3295 [06:21<27:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 615/3295 [06:21<27:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 615/3295 [06:21<27:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 616/3295 [06:21<27:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 616/3295 [06:21<27:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 617/3295 [06:23<27:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▊        | 617/3295 [06:23<27:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 618/3295 [06:23<27:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 618/3295 [06:23<27:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 619/3295 [06:23<27:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 619/3295 [06:23<27:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 620/3295 [06:25<27:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 620/3295 [06:25<27:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 621/3295 [06:25<27:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 621/3295 [06:25<27:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 622/3295 [06:25<27:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 622/3295 [06:25<27:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 623/3295 [06:27<27:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 623/3295 [06:27<27:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 624/3295 [06:27<27:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 624/3295 [06:27<27:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 625/3295 [06:27<27:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 625/3295 [06:27<27:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 626/3295 [06:28<27:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 626/3295 [06:28<27:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 627/3295 [06:28<27:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 627/3295 [06:28<27:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 628/3295 [06:29<27:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 628/3295 [06:29<27:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 629/3295 [06:30<27:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 629/3295 [06:30<27:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 630/3295 [06:30<27:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 630/3295 [06:30<27:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 631/3295 [06:30<27:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 631/3295 [06:30<27:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 632/3295 [06:32<27:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 632/3295 [06:32<27:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 633/3295 [06:32<27:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 633/3295 [06:32<27:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 634/3295 [06:32<27:27,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 634/3295 [06:32<27:27,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 635/3295 [06:34<27:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 635/3295 [06:34<27:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 636/3295 [06:34<27:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 636/3295 [06:34<27:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 637/3295 [06:34<27:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 637/3295 [06:34<27:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 638/3295 [06:36<27:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 638/3295 [06:36<27:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 639/3295 [06:36<27:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 639/3295 [06:36<27:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 640/3295 [06:36<27:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 640/3295 [06:36<27:23,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 641/3295 [06:37<27:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 641/3295 [06:37<27:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 642/3295 [06:38<27:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  19%|█▉        | 642/3295 [06:38<27:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 643/3295 [06:38<27:21,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 643/3295 [06:38<27:21,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 644/3295 [06:39<27:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 644/3295 [06:39<27:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 645/3295 [06:39<27:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 645/3295 [06:39<27:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 646/3295 [06:39<27:19,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 646/3295 [06:39<27:19,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 647/3295 [06:41<27:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 647/3295 [06:41<27:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 648/3295 [06:41<27:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 648/3295 [06:41<27:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 649/3295 [06:41<27:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 649/3295 [06:41<27:18,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 650/3295 [06:43<27:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 650/3295 [06:43<27:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 651/3295 [06:43<27:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 651/3295 [06:43<27:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 652/3295 [06:43<27:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 652/3295 [06:43<27:16,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 653/3295 [06:45<27:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 653/3295 [06:45<27:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 654/3295 [06:45<27:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 654/3295 [06:45<27:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 655/3295 [06:45<27:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 655/3295 [06:45<27:14,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 656/3295 [06:47<27:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 656/3295 [06:47<27:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 657/3295 [06:47<27:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 657/3295 [06:47<27:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 658/3295 [06:47<27:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|█▉        | 658/3295 [06:47<27:12,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 659/3295 [06:49<27:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 659/3295 [06:49<27:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 660/3295 [06:49<27:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 660/3295 [06:49<27:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 661/3295 [06:49<27:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 661/3295 [06:49<27:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 662/3295 [06:51<27:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 662/3295 [06:51<27:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 663/3295 [06:51<27:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 663/3295 [06:51<27:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 664/3295 [06:51<27:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 664/3295 [06:51<27:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 665/3295 [06:52<27:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 665/3295 [06:52<27:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 666/3295 [06:53<27:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 666/3295 [06:53<27:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 667/3295 [06:53<27:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 667/3295 [06:53<27:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 668/3295 [06:54<27:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 668/3295 [06:54<27:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 669/3295 [06:54<27:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 669/3295 [06:54<27:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 670/3295 [06:54<27:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 670/3295 [06:54<27:05,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 671/3295 [06:56<27:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 671/3295 [06:56<27:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 672/3295 [06:56<27:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 672/3295 [06:56<27:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 673/3295 [06:56<27:03,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 673/3295 [06:56<27:03,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 674/3295 [06:58<27:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 674/3295 [06:58<27:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 675/3295 [06:58<27:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  20%|██        | 675/3295 [06:58<27:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 676/3295 [06:58<27:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 676/3295 [06:58<27:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 677/3295 [07:00<27:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 677/3295 [07:00<27:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 678/3295 [07:00<27:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 678/3295 [07:00<27:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 679/3295 [07:00<26:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 679/3295 [07:00<26:59,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 680/3295 [07:02<27:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 680/3295 [07:02<27:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 681/3295 [07:02<27:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 681/3295 [07:02<27:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 682/3295 [07:02<26:57,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 682/3295 [07:02<26:57,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 683/3295 [07:03<27:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 683/3295 [07:03<27:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 684/3295 [07:03<26:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 684/3295 [07:03<26:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 685/3295 [07:03<26:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 685/3295 [07:03<26:55,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 686/3295 [07:05<26:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 686/3295 [07:05<26:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 687/3295 [07:05<26:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 687/3295 [07:05<26:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 688/3295 [07:05<26:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 688/3295 [07:05<26:53,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 689/3295 [07:07<26:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 689/3295 [07:07<26:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 690/3295 [07:07<26:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 690/3295 [07:07<26:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 691/3295 [07:07<26:51,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 691/3295 [07:07<26:51,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 692/3295 [07:09<26:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 692/3295 [07:09<26:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 693/3295 [07:09<26:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 693/3295 [07:09<26:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 694/3295 [07:09<26:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 694/3295 [07:09<26:49,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 695/3295 [07:11<26:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 695/3295 [07:11<26:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 696/3295 [07:11<26:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 696/3295 [07:11<26:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 697/3295 [07:11<26:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 697/3295 [07:11<26:47,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 698/3295 [07:13<26:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 698/3295 [07:13<26:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 699/3295 [07:13<26:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 699/3295 [07:13<26:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 700/3295 [07:13<26:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██        | 700/3295 [07:13<26:46,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 701/3295 [07:15<26:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 701/3295 [07:15<26:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 702/3295 [07:15<26:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 702/3295 [07:15<26:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 703/3295 [07:15<26:44,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 703/3295 [07:15<26:44,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 704/3295 [07:16<26:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 704/3295 [07:16<26:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 705/3295 [07:17<26:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 705/3295 [07:17<26:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 706/3295 [07:17<26:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 706/3295 [07:17<26:42,  1.62it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 707/3295 [07:19<26:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 707/3295 [07:19<26:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 708/3295 [07:19<26:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  21%|██▏       | 708/3295 [07:19<26:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 709/3295 [07:19<26:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 709/3295 [07:19<26:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 710/3295 [07:20<26:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 710/3295 [07:20<26:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 711/3295 [07:20<26:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 711/3295 [07:20<26:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 712/3295 [07:21<26:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 712/3295 [07:21<26:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 713/3295 [07:22<26:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 713/3295 [07:22<26:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 714/3295 [07:22<26:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 714/3295 [07:22<26:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 715/3295 [07:22<26:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 715/3295 [07:22<26:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 716/3295 [07:24<26:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 716/3295 [07:24<26:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 717/3295 [07:24<26:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 717/3295 [07:24<26:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 718/3295 [07:24<26:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 718/3295 [07:24<26:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 719/3295 [07:26<26:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 719/3295 [07:26<26:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 720/3295 [07:26<26:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 720/3295 [07:26<26:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 721/3295 [07:26<26:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 721/3295 [07:26<26:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 722/3295 [07:28<26:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 722/3295 [07:28<26:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 723/3295 [07:28<26:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 723/3295 [07:28<26:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 724/3295 [07:28<26:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 724/3295 [07:28<26:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 725/3295 [07:30<26:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 725/3295 [07:30<26:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 726/3295 [07:30<26:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 726/3295 [07:30<26:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 727/3295 [07:30<26:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 727/3295 [07:30<26:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 728/3295 [07:32<26:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 728/3295 [07:32<26:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 729/3295 [07:32<26:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 729/3295 [07:32<26:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 730/3295 [07:32<26:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 730/3295 [07:32<26:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 731/3295 [07:34<26:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 731/3295 [07:34<26:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 732/3295 [07:34<26:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 732/3295 [07:34<26:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 733/3295 [07:34<26:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 733/3295 [07:34<26:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 734/3295 [07:35<26:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 734/3295 [07:35<26:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 735/3295 [07:35<26:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 735/3295 [07:35<26:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 736/3295 [07:35<26:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 736/3295 [07:35<26:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 737/3295 [07:37<26:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 737/3295 [07:37<26:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 738/3295 [07:37<26:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 738/3295 [07:37<26:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 739/3295 [07:37<26:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 739/3295 [07:37<26:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 740/3295 [07:39<26:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 740/3295 [07:39<26:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 741/3295 [07:39<26:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  22%|██▏       | 741/3295 [07:39<26:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 742/3295 [07:39<26:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 742/3295 [07:39<26:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 743/3295 [07:41<26:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 743/3295 [07:41<26:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 744/3295 [07:41<26:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 744/3295 [07:41<26:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 745/3295 [07:41<26:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 745/3295 [07:41<26:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 746/3295 [07:43<26:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 746/3295 [07:43<26:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 747/3295 [07:43<26:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 747/3295 [07:43<26:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 748/3295 [07:43<26:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 748/3295 [07:43<26:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 749/3295 [07:45<26:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 749/3295 [07:45<26:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 750/3295 [07:45<26:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 750/3295 [07:45<26:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 751/3295 [07:45<26:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 751/3295 [07:45<26:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 752/3295 [07:47<26:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 752/3295 [07:47<26:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 753/3295 [07:47<26:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 753/3295 [07:47<26:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 754/3295 [07:47<26:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 754/3295 [07:47<26:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 755/3295 [07:49<26:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 755/3295 [07:49<26:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 756/3295 [07:49<26:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 756/3295 [07:49<26:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 757/3295 [07:49<26:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 757/3295 [07:49<26:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 758/3295 [07:50<26:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 758/3295 [07:50<26:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 759/3295 [07:50<26:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 759/3295 [07:50<26:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 760/3295 [07:50<26:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 760/3295 [07:50<26:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 761/3295 [07:52<26:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 761/3295 [07:52<26:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 762/3295 [07:52<26:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 762/3295 [07:52<26:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 763/3295 [07:52<26:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 763/3295 [07:52<26:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 764/3295 [07:54<26:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 764/3295 [07:54<26:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 765/3295 [07:54<26:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 765/3295 [07:54<26:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 766/3295 [07:54<26:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 766/3295 [07:54<26:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 767/3295 [07:56<26:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 767/3295 [07:56<26:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 768/3295 [07:56<26:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 768/3295 [07:56<26:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 769/3295 [07:56<26:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 769/3295 [07:56<26:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 770/3295 [07:58<26:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 770/3295 [07:58<26:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 771/3295 [07:58<26:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 771/3295 [07:58<26:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 772/3295 [07:58<26:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 772/3295 [07:58<26:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 773/3295 [08:00<26:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 773/3295 [08:00<26:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 774/3295 [08:00<26:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  23%|██▎       | 774/3295 [08:00<26:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 775/3295 [08:00<26:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 775/3295 [08:00<26:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 776/3295 [08:02<26:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 776/3295 [08:02<26:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 777/3295 [08:02<26:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 777/3295 [08:02<26:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 778/3295 [08:02<26:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 778/3295 [08:02<26:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 779/3295 [08:04<26:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 779/3295 [08:04<26:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 780/3295 [08:04<26:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 780/3295 [08:04<26:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 781/3295 [08:04<25:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 781/3295 [08:04<25:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 782/3295 [08:05<26:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▎       | 782/3295 [08:05<26:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 783/3295 [08:06<25:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 783/3295 [08:06<25:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 784/3295 [08:06<25:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 784/3295 [08:06<25:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 785/3295 [08:08<26:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 785/3295 [08:08<26:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 786/3295 [08:08<25:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 786/3295 [08:08<25:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 787/3295 [08:08<25:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 787/3295 [08:08<25:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 788/3295 [08:09<25:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 788/3295 [08:09<25:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 789/3295 [08:09<25:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 789/3295 [08:09<25:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 790/3295 [08:09<25:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 790/3295 [08:09<25:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 791/3295 [08:11<25:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 791/3295 [08:11<25:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 792/3295 [08:11<25:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 792/3295 [08:11<25:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 793/3295 [08:11<25:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 793/3295 [08:11<25:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 794/3295 [08:13<25:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 794/3295 [08:13<25:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 795/3295 [08:13<25:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 795/3295 [08:13<25:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 796/3295 [08:13<25:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 796/3295 [08:13<25:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 797/3295 [08:15<25:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 797/3295 [08:15<25:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 798/3295 [08:15<25:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 798/3295 [08:15<25:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 799/3295 [08:15<25:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 799/3295 [08:15<25:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 800/3295 [08:17<25:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 800/3295 [08:17<25:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 801/3295 [08:17<25:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 801/3295 [08:17<25:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 802/3295 [08:17<25:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 802/3295 [08:17<25:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 803/3295 [08:19<25:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 803/3295 [08:19<25:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 804/3295 [08:19<25:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 804/3295 [08:19<25:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 805/3295 [08:19<25:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 805/3295 [08:19<25:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 806/3295 [08:21<25:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 806/3295 [08:21<25:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 807/3295 [08:21<25:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  24%|██▍       | 807/3295 [08:21<25:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 808/3295 [08:21<25:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 808/3295 [08:21<25:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 809/3295 [08:23<25:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 809/3295 [08:23<25:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 810/3295 [08:23<25:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 810/3295 [08:23<25:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 811/3295 [08:23<25:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 811/3295 [08:23<25:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 812/3295 [08:24<25:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 812/3295 [08:24<25:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 813/3295 [08:25<25:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 813/3295 [08:25<25:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 814/3295 [08:25<25:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 814/3295 [08:25<25:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 815/3295 [08:26<25:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 815/3295 [08:26<25:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 816/3295 [08:26<25:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 816/3295 [08:26<25:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 817/3295 [08:26<25:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 817/3295 [08:26<25:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 818/3295 [08:28<25:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 818/3295 [08:28<25:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 819/3295 [08:28<25:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 819/3295 [08:28<25:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 820/3295 [08:28<25:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 820/3295 [08:28<25:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 821/3295 [08:30<25:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 821/3295 [08:30<25:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 822/3295 [08:30<25:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 822/3295 [08:30<25:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 823/3295 [08:30<25:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▍       | 823/3295 [08:30<25:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 824/3295 [08:32<25:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 824/3295 [08:32<25:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 825/3295 [08:32<25:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 825/3295 [08:32<25:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 826/3295 [08:33<25:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 826/3295 [08:33<25:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 827/3295 [08:34<25:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 827/3295 [08:34<25:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 828/3295 [08:34<25:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 828/3295 [08:34<25:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 829/3295 [08:34<25:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 829/3295 [08:34<25:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 830/3295 [08:36<25:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 830/3295 [08:36<25:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 831/3295 [08:36<25:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 831/3295 [08:36<25:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 832/3295 [08:36<25:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 832/3295 [08:36<25:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 833/3295 [08:38<25:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 833/3295 [08:38<25:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 834/3295 [08:38<25:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 834/3295 [08:38<25:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 835/3295 [08:38<25:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 835/3295 [08:38<25:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 836/3295 [08:40<25:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 836/3295 [08:40<25:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 837/3295 [08:40<25:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 837/3295 [08:40<25:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 838/3295 [08:40<25:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 838/3295 [08:40<25:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 839/3295 [08:42<25:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 839/3295 [08:42<25:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 840/3295 [08:42<25:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  25%|██▌       | 840/3295 [08:42<25:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 841/3295 [08:42<25:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 841/3295 [08:42<25:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 842/3295 [08:44<25:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 842/3295 [08:44<25:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 843/3295 [08:44<25:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 843/3295 [08:44<25:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 844/3295 [08:44<25:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 844/3295 [08:44<25:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 845/3295 [08:46<25:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 845/3295 [08:46<25:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 846/3295 [08:46<25:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 846/3295 [08:46<25:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 847/3295 [08:46<25:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 847/3295 [08:46<25:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 848/3295 [08:47<25:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 848/3295 [08:47<25:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 849/3295 [08:47<25:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 849/3295 [08:47<25:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 850/3295 [08:48<25:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 850/3295 [08:48<25:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 851/3295 [08:49<25:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 851/3295 [08:49<25:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 852/3295 [08:49<25:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 852/3295 [08:49<25:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 853/3295 [08:49<25:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 853/3295 [08:49<25:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 854/3295 [08:51<25:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 854/3295 [08:51<25:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 855/3295 [08:51<25:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 855/3295 [08:51<25:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 856/3295 [08:51<25:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 856/3295 [08:51<25:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 857/3295 [08:53<25:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 857/3295 [08:53<25:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 858/3295 [08:53<25:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 858/3295 [08:53<25:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 859/3295 [08:53<25:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 859/3295 [08:53<25:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 860/3295 [08:55<25:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 860/3295 [08:55<25:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 861/3295 [08:55<25:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 861/3295 [08:55<25:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 862/3295 [08:55<25:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 862/3295 [08:55<25:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 863/3295 [08:57<25:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 863/3295 [08:57<25:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 864/3295 [08:57<25:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▌       | 864/3295 [08:57<25:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 865/3295 [08:57<25:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 865/3295 [08:57<25:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 866/3295 [08:59<25:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 866/3295 [08:59<25:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 867/3295 [08:59<25:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 867/3295 [08:59<25:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 868/3295 [08:59<25:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 868/3295 [08:59<25:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 869/3295 [09:01<25:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 869/3295 [09:01<25:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 870/3295 [09:01<25:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 870/3295 [09:01<25:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 871/3295 [09:01<25:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 871/3295 [09:01<25:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 872/3295 [09:03<25:09,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 872/3295 [09:03<25:09,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 873/3295 [09:03<25:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  26%|██▋       | 873/3295 [09:03<25:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 874/3295 [09:03<25:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 874/3295 [09:03<25:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 875/3295 [09:05<25:07,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 875/3295 [09:05<25:07,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 876/3295 [09:05<25:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 876/3295 [09:05<25:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 877/3295 [09:05<25:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 877/3295 [09:05<25:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 878/3295 [09:07<25:06,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 878/3295 [09:07<25:06,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 879/3295 [09:07<25:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 879/3295 [09:07<25:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 880/3295 [09:07<25:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 880/3295 [09:07<25:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 881/3295 [09:08<25:04,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 881/3295 [09:08<25:04,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 882/3295 [09:09<25:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 882/3295 [09:09<25:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 883/3295 [09:09<24:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 883/3295 [09:09<24:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 884/3295 [09:10<25:02,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 884/3295 [09:10<25:02,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 885/3295 [09:10<24:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 885/3295 [09:10<24:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 886/3295 [09:10<24:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 886/3295 [09:10<24:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 887/3295 [09:12<25:00,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 887/3295 [09:12<25:00,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 888/3295 [09:12<24:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 888/3295 [09:12<24:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 889/3295 [09:12<24:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 889/3295 [09:12<24:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 890/3295 [09:14<24:58,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 890/3295 [09:14<24:58,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 891/3295 [09:14<24:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 891/3295 [09:14<24:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 892/3295 [09:14<24:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 892/3295 [09:14<24:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 893/3295 [09:16<24:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 893/3295 [09:16<24:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 894/3295 [09:16<24:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 894/3295 [09:16<24:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 895/3295 [09:16<24:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 895/3295 [09:16<24:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 896/3295 [09:18<24:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 896/3295 [09:18<24:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 897/3295 [09:18<24:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 897/3295 [09:18<24:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 898/3295 [09:18<24:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 898/3295 [09:18<24:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 899/3295 [09:19<24:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 899/3295 [09:19<24:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 900/3295 [09:19<24:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 900/3295 [09:19<24:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 901/3295 [09:19<24:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 901/3295 [09:19<24:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 902/3295 [09:21<24:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 902/3295 [09:21<24:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 903/3295 [09:21<24:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 903/3295 [09:21<24:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 904/3295 [09:21<24:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 904/3295 [09:21<24:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 905/3295 [09:23<24:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 905/3295 [09:23<24:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 906/3295 [09:23<24:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  27%|██▋       | 906/3295 [09:23<24:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 907/3295 [09:23<24:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 907/3295 [09:23<24:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 908/3295 [09:25<24:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 908/3295 [09:25<24:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 909/3295 [09:25<24:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 909/3295 [09:25<24:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 910/3295 [09:25<24:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 910/3295 [09:25<24:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 911/3295 [09:27<24:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 911/3295 [09:27<24:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 912/3295 [09:27<24:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 912/3295 [09:27<24:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 913/3295 [09:27<24:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 913/3295 [09:27<24:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 914/3295 [09:29<24:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 914/3295 [09:29<24:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 915/3295 [09:29<24:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 915/3295 [09:29<24:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 916/3295 [09:29<24:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 916/3295 [09:29<24:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 917/3295 [09:31<24:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 917/3295 [09:31<24:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 918/3295 [09:31<24:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 918/3295 [09:31<24:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 919/3295 [09:31<24:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 919/3295 [09:31<24:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 920/3295 [09:32<24:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 920/3295 [09:32<24:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 921/3295 [09:33<24:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 921/3295 [09:33<24:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 922/3295 [09:33<24:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 922/3295 [09:33<24:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 923/3295 [09:34<24:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 923/3295 [09:34<24:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 924/3295 [09:34<24:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 924/3295 [09:34<24:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 925/3295 [09:34<24:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 925/3295 [09:34<24:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 926/3295 [09:36<24:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 926/3295 [09:36<24:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 927/3295 [09:36<24:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 927/3295 [09:36<24:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 928/3295 [09:36<24:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 928/3295 [09:36<24:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 929/3295 [09:38<24:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 929/3295 [09:38<24:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 930/3295 [09:38<24:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 930/3295 [09:38<24:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 931/3295 [09:38<24:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 931/3295 [09:38<24:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 932/3295 [09:40<24:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 932/3295 [09:40<24:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 933/3295 [09:40<24:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 933/3295 [09:40<24:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 934/3295 [09:40<24:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 934/3295 [09:40<24:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 935/3295 [09:42<24:30,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 935/3295 [09:42<24:30,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 936/3295 [09:42<24:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 936/3295 [09:42<24:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 937/3295 [09:42<24:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 937/3295 [09:42<24:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 938/3295 [09:44<24:28,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 938/3295 [09:44<24:28,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 939/3295 [09:44<24:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  28%|██▊       | 939/3295 [09:44<24:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 940/3295 [09:44<24:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 940/3295 [09:44<24:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 941/3295 [09:46<24:26,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 941/3295 [09:46<24:26,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 942/3295 [09:46<24:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 942/3295 [09:46<24:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 943/3295 [09:46<24:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 943/3295 [09:46<24:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 944/3295 [09:48<24:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 944/3295 [09:48<24:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 945/3295 [09:48<24:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 945/3295 [09:48<24:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 946/3295 [09:48<24:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 946/3295 [09:48<24:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 947/3295 [09:49<24:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▊       | 947/3295 [09:49<24:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 948/3295 [09:49<24:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 948/3295 [09:49<24:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 949/3295 [09:49<24:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 949/3295 [09:49<24:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 950/3295 [09:51<24:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 950/3295 [09:51<24:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 951/3295 [09:51<24:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 951/3295 [09:51<24:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 952/3295 [09:51<24:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 952/3295 [09:51<24:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 953/3295 [09:53<24:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 953/3295 [09:53<24:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 954/3295 [09:53<24:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 954/3295 [09:53<24:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 955/3295 [09:53<24:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 955/3295 [09:53<24:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 956/3295 [09:55<24:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 956/3295 [09:55<24:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 957/3295 [09:55<24:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 957/3295 [09:55<24:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 958/3295 [09:55<24:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 958/3295 [09:55<24:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 959/3295 [09:57<24:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 959/3295 [09:57<24:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 960/3295 [09:57<24:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 960/3295 [09:57<24:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 961/3295 [09:57<24:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 961/3295 [09:57<24:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 962/3295 [09:59<24:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 962/3295 [09:59<24:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 963/3295 [09:59<24:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 963/3295 [09:59<24:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 964/3295 [09:59<24:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 964/3295 [09:59<24:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 965/3295 [10:00<24:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 965/3295 [10:00<24:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 966/3295 [10:01<24:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 966/3295 [10:01<24:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 967/3295 [10:01<24:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 967/3295 [10:01<24:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 968/3295 [10:02<24:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 968/3295 [10:02<24:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 969/3295 [10:02<24:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 969/3295 [10:02<24:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 970/3295 [10:02<24:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 970/3295 [10:02<24:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 971/3295 [10:04<24:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 971/3295 [10:04<24:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 972/3295 [10:04<24:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  29%|██▉       | 972/3295 [10:04<24:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 973/3295 [10:04<24:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 973/3295 [10:04<24:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 974/3295 [10:06<24:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 974/3295 [10:06<24:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 975/3295 [10:06<24:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 975/3295 [10:06<24:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 976/3295 [10:06<24:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 976/3295 [10:06<24:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 977/3295 [10:08<24:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 977/3295 [10:08<24:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 978/3295 [10:08<24:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 978/3295 [10:08<24:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 979/3295 [10:08<23:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 979/3295 [10:08<23:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 980/3295 [10:10<24:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 980/3295 [10:10<24:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 981/3295 [10:10<23:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 981/3295 [10:10<23:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 982/3295 [10:10<23:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 982/3295 [10:10<23:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 983/3295 [10:12<23:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 983/3295 [10:12<23:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 984/3295 [10:12<23:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 984/3295 [10:12<23:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 985/3295 [10:12<23:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 985/3295 [10:12<23:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 986/3295 [10:13<23:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 986/3295 [10:13<23:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 987/3295 [10:14<23:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 987/3295 [10:14<23:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 988/3295 [10:14<23:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|██▉       | 988/3295 [10:14<23:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 989/3295 [10:15<23:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 989/3295 [10:15<23:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 990/3295 [10:15<23:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 990/3295 [10:15<23:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 991/3295 [10:15<23:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 991/3295 [10:15<23:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 992/3295 [10:17<23:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 992/3295 [10:17<23:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 993/3295 [10:17<23:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 993/3295 [10:17<23:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 994/3295 [10:17<23:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 994/3295 [10:17<23:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 995/3295 [10:19<23:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 995/3295 [10:19<23:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 996/3295 [10:19<23:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 996/3295 [10:19<23:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 997/3295 [10:19<23:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 997/3295 [10:19<23:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 998/3295 [10:21<23:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 998/3295 [10:21<23:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 999/3295 [10:21<23:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 999/3295 [10:21<23:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1000/3295 [10:21<23:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1000/3295 [10:21<23:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1001/3295 [10:23<23:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1001/3295 [10:23<23:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1002/3295 [10:23<23:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1002/3295 [10:23<23:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1003/3295 [10:23<23:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1003/3295 [10:23<23:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1004/3295 [10:25<23:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  30%|███       | 1004/3295 [10:25<23:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1005/3295 [10:25<23:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1005/3295 [10:25<23:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1006/3295 [10:25<23:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1006/3295 [10:25<23:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1007/3295 [10:27<23:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1007/3295 [10:27<23:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1008/3295 [10:27<23:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1008/3295 [10:27<23:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1009/3295 [10:27<23:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1009/3295 [10:27<23:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1010/3295 [10:29<23:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1010/3295 [10:29<23:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1011/3295 [10:29<23:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1011/3295 [10:29<23:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1012/3295 [10:29<23:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1012/3295 [10:29<23:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1013/3295 [10:31<23:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1013/3295 [10:31<23:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1014/3295 [10:31<23:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1014/3295 [10:31<23:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1015/3295 [10:31<23:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1015/3295 [10:31<23:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1016/3295 [10:32<23:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1016/3295 [10:32<23:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1017/3295 [10:32<23:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1017/3295 [10:32<23:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1018/3295 [10:33<23:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1018/3295 [10:33<23:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1019/3295 [10:34<23:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1019/3295 [10:34<23:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1020/3295 [10:34<23:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1020/3295 [10:34<23:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1021/3295 [10:34<23:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1021/3295 [10:34<23:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1022/3295 [10:36<23:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1022/3295 [10:36<23:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1023/3295 [10:36<23:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1023/3295 [10:36<23:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1024/3295 [10:36<23:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1024/3295 [10:36<23:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1025/3295 [10:38<23:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1025/3295 [10:38<23:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1026/3295 [10:38<23:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1026/3295 [10:38<23:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1027/3295 [10:38<23:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1027/3295 [10:38<23:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1028/3295 [10:40<23:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1028/3295 [10:40<23:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1029/3295 [10:40<23:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███       | 1029/3295 [10:40<23:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1030/3295 [10:40<23:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1030/3295 [10:40<23:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1031/3295 [10:42<23:30,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1031/3295 [10:42<23:30,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1032/3295 [10:42<23:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1032/3295 [10:42<23:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1033/3295 [10:42<23:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1033/3295 [10:42<23:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1034/3295 [10:44<23:28,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1034/3295 [10:44<23:28,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1035/3295 [10:44<23:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1035/3295 [10:44<23:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1036/3295 [10:44<23:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1036/3295 [10:44<23:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1037/3295 [10:46<23:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  31%|███▏      | 1037/3295 [10:46<23:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1038/3295 [10:46<23:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1038/3295 [10:46<23:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1039/3295 [10:46<23:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1039/3295 [10:46<23:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1040/3295 [10:48<23:25,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1040/3295 [10:48<23:25,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1041/3295 [10:48<23:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1041/3295 [10:48<23:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1042/3295 [10:48<23:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1042/3295 [10:48<23:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1043/3295 [10:49<23:23,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1043/3295 [10:49<23:23,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1044/3295 [10:49<23:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1044/3295 [10:49<23:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1045/3295 [10:50<23:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1045/3295 [10:50<23:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1046/3295 [10:51<23:21,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1046/3295 [10:51<23:21,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1047/3295 [10:51<23:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1047/3295 [10:51<23:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1048/3295 [10:51<23:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1048/3295 [10:51<23:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1049/3295 [10:53<23:19,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1049/3295 [10:53<23:19,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1050/3295 [10:53<23:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1050/3295 [10:53<23:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1051/3295 [10:53<23:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1051/3295 [10:53<23:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1052/3295 [10:55<23:17,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1052/3295 [10:55<23:17,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1053/3295 [10:55<23:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1053/3295 [10:55<23:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1054/3295 [10:55<23:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1054/3295 [10:55<23:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1055/3295 [10:57<23:16,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1055/3295 [10:57<23:16,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1056/3295 [10:57<23:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1056/3295 [10:57<23:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1057/3295 [10:57<23:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1057/3295 [10:57<23:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1058/3295 [10:59<23:14,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1058/3295 [10:59<23:14,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1059/3295 [10:59<23:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1059/3295 [10:59<23:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1060/3295 [10:59<23:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1060/3295 [10:59<23:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1061/3295 [11:01<23:12,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1061/3295 [11:01<23:12,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1062/3295 [11:01<23:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1062/3295 [11:01<23:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1063/3295 [11:01<23:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1063/3295 [11:01<23:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1064/3295 [11:03<23:10,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1064/3295 [11:03<23:10,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1065/3295 [11:03<23:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1065/3295 [11:03<23:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1066/3295 [11:03<23:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1066/3295 [11:03<23:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1067/3295 [11:04<23:08,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1067/3295 [11:04<23:08,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1068/3295 [11:04<23:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1068/3295 [11:04<23:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1069/3295 [11:05<23:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1069/3295 [11:05<23:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1070/3295 [11:06<23:06,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  32%|███▏      | 1070/3295 [11:06<23:06,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1071/3295 [11:06<23:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1071/3295 [11:06<23:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1072/3295 [11:06<23:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1072/3295 [11:06<23:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1073/3295 [11:08<23:04,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1073/3295 [11:08<23:04,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1074/3295 [11:08<23:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1074/3295 [11:08<23:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1075/3295 [11:08<23:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1075/3295 [11:08<23:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1076/3295 [11:10<23:03,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1076/3295 [11:10<23:03,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1077/3295 [11:10<23:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1077/3295 [11:10<23:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1078/3295 [11:10<22:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1078/3295 [11:10<22:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1079/3295 [11:12<23:01,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1079/3295 [11:12<23:01,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1080/3295 [11:12<22:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1080/3295 [11:12<22:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1081/3295 [11:12<22:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1081/3295 [11:12<22:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1082/3295 [11:14<22:59,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1082/3295 [11:14<22:59,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1083/3295 [11:14<22:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1083/3295 [11:14<22:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1084/3295 [11:14<22:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1084/3295 [11:14<22:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1085/3295 [11:16<22:57,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1085/3295 [11:16<22:57,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1086/3295 [11:16<22:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1086/3295 [11:16<22:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1087/3295 [11:16<22:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1087/3295 [11:16<22:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1088/3295 [11:17<22:55,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1088/3295 [11:17<22:55,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1089/3295 [11:18<22:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1089/3295 [11:18<22:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1090/3295 [11:18<22:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1090/3295 [11:18<22:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1091/3295 [11:19<22:53,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1091/3295 [11:19<22:53,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1092/3295 [11:19<22:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1092/3295 [11:19<22:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1093/3295 [11:19<22:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1093/3295 [11:19<22:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1094/3295 [11:21<22:51,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1094/3295 [11:21<22:51,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1095/3295 [11:21<22:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1095/3295 [11:21<22:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1096/3295 [11:21<22:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1096/3295 [11:21<22:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1097/3295 [11:23<22:49,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1097/3295 [11:23<22:49,  1.60it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1098/3295 [11:23<22:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1098/3295 [11:23<22:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1099/3295 [11:23<22:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1099/3295 [11:23<22:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1100/3295 [11:25<22:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1100/3295 [11:25<22:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1101/3295 [11:25<22:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1101/3295 [11:25<22:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1102/3295 [11:25<22:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1102/3295 [11:25<22:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1103/3295 [11:27<22:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  33%|███▎      | 1103/3295 [11:27<22:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1104/3295 [11:27<22:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1104/3295 [11:27<22:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1105/3295 [11:27<22:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1105/3295 [11:27<22:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1106/3295 [11:29<22:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1106/3295 [11:29<22:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1107/3295 [11:29<22:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1107/3295 [11:29<22:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1108/3295 [11:29<22:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1108/3295 [11:29<22:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1109/3295 [11:30<22:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1109/3295 [11:30<22:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1110/3295 [11:30<22:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1110/3295 [11:30<22:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1111/3295 [11:30<22:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1111/3295 [11:30<22:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1112/3295 [11:32<22:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▎      | 1112/3295 [11:32<22:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1113/3295 [11:32<22:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1113/3295 [11:32<22:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1114/3295 [11:32<22:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1114/3295 [11:32<22:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1115/3295 [11:34<22:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1115/3295 [11:34<22:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1116/3295 [11:34<22:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1116/3295 [11:34<22:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1117/3295 [11:34<22:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1117/3295 [11:34<22:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1118/3295 [11:36<22:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1118/3295 [11:36<22:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1119/3295 [11:36<22:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1119/3295 [11:36<22:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1120/3295 [11:36<22:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1120/3295 [11:36<22:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1121/3295 [11:38<22:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1121/3295 [11:38<22:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1122/3295 [11:38<22:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1122/3295 [11:38<22:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1123/3295 [11:38<22:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1123/3295 [11:38<22:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1124/3295 [11:39<22:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1124/3295 [11:39<22:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1125/3295 [11:39<22:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1125/3295 [11:39<22:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1126/3295 [11:39<22:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1126/3295 [11:39<22:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1127/3295 [11:41<22:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1127/3295 [11:41<22:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1128/3295 [11:41<22:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1128/3295 [11:41<22:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1129/3295 [11:41<22:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1129/3295 [11:41<22:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1130/3295 [11:43<22:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1130/3295 [11:43<22:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1131/3295 [11:43<22:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1131/3295 [11:43<22:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1132/3295 [11:43<22:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1132/3295 [11:43<22:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1133/3295 [11:45<22:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1133/3295 [11:45<22:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1134/3295 [11:45<22:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1134/3295 [11:45<22:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1135/3295 [11:45<22:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1135/3295 [11:45<22:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1136/3295 [11:47<22:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  34%|███▍      | 1136/3295 [11:47<22:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1137/3295 [11:47<22:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1137/3295 [11:47<22:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1138/3295 [11:47<22:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1138/3295 [11:47<22:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1139/3295 [11:49<22:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1139/3295 [11:49<22:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1140/3295 [11:49<22:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1140/3295 [11:49<22:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1141/3295 [11:49<22:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1141/3295 [11:49<22:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1142/3295 [11:50<22:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1142/3295 [11:50<22:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1143/3295 [11:50<22:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1143/3295 [11:50<22:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1144/3295 [11:50<22:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1144/3295 [11:50<22:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1145/3295 [11:52<22:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1145/3295 [11:52<22:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1146/3295 [11:52<22:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1146/3295 [11:52<22:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1147/3295 [11:52<22:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1147/3295 [11:52<22:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1148/3295 [11:54<22:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1148/3295 [11:54<22:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1149/3295 [11:54<22:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1149/3295 [11:54<22:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1150/3295 [11:54<22:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1150/3295 [11:54<22:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1151/3295 [11:56<22:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1151/3295 [11:56<22:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1152/3295 [11:56<22:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1152/3295 [11:56<22:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1153/3295 [11:56<22:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▍      | 1153/3295 [11:56<22:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1154/3295 [11:58<22:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1154/3295 [11:58<22:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1155/3295 [11:58<22:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1155/3295 [11:58<22:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1156/3295 [11:58<22:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1156/3295 [11:58<22:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1157/3295 [12:00<22:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1157/3295 [12:00<22:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1158/3295 [12:00<22:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1158/3295 [12:00<22:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1159/3295 [12:00<22:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1159/3295 [12:00<22:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1160/3295 [12:02<22:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1160/3295 [12:02<22:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1161/3295 [12:02<22:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1161/3295 [12:02<22:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1162/3295 [12:02<22:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1162/3295 [12:02<22:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1163/3295 [12:04<22:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1163/3295 [12:04<22:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1164/3295 [12:04<22:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1164/3295 [12:04<22:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1165/3295 [12:04<22:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1165/3295 [12:04<22:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1166/3295 [12:05<22:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1166/3295 [12:05<22:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1167/3295 [12:05<22:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1167/3295 [12:05<22:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1168/3295 [12:05<22:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1168/3295 [12:05<22:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1169/3295 [12:07<22:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  35%|███▌      | 1169/3295 [12:07<22:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1170/3295 [12:07<22:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1170/3295 [12:07<22:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1171/3295 [12:07<22:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1171/3295 [12:07<22:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1172/3295 [12:09<22:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1172/3295 [12:09<22:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1173/3295 [12:09<21:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1173/3295 [12:09<21:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1174/3295 [12:09<21:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1174/3295 [12:09<21:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1175/3295 [12:11<21:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1175/3295 [12:11<21:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1176/3295 [12:11<21:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1176/3295 [12:11<21:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1177/3295 [12:11<21:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1177/3295 [12:11<21:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1178/3295 [12:13<21:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1178/3295 [12:13<21:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1179/3295 [12:13<21:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1179/3295 [12:13<21:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1180/3295 [12:13<21:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1180/3295 [12:13<21:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1181/3295 [12:14<21:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1181/3295 [12:14<21:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1182/3295 [12:14<21:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1182/3295 [12:14<21:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1183/3295 [12:14<21:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1183/3295 [12:14<21:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1184/3295 [12:16<21:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1184/3295 [12:16<21:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1185/3295 [12:16<21:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1185/3295 [12:16<21:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1186/3295 [12:16<21:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1186/3295 [12:16<21:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1187/3295 [12:18<21:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1187/3295 [12:18<21:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1188/3295 [12:18<21:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1188/3295 [12:18<21:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1189/3295 [12:18<21:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1189/3295 [12:18<21:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1190/3295 [12:20<21:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1190/3295 [12:20<21:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1191/3295 [12:20<21:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1191/3295 [12:20<21:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1192/3295 [12:20<21:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1192/3295 [12:20<21:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1193/3295 [12:22<21:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1193/3295 [12:22<21:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1194/3295 [12:22<21:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▌      | 1194/3295 [12:22<21:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1195/3295 [12:22<21:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1195/3295 [12:22<21:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1196/3295 [12:24<21:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1196/3295 [12:24<21:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1197/3295 [12:24<21:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1197/3295 [12:24<21:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1198/3295 [12:24<21:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1198/3295 [12:24<21:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1199/3295 [12:26<21:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1199/3295 [12:26<21:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1200/3295 [12:26<21:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1200/3295 [12:26<21:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1201/3295 [12:26<21:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1201/3295 [12:26<21:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1202/3295 [12:28<21:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  36%|███▋      | 1202/3295 [12:28<21:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1203/3295 [12:28<21:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1203/3295 [12:28<21:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1204/3295 [12:28<21:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1204/3295 [12:28<21:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1205/3295 [12:30<21:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1205/3295 [12:30<21:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1206/3295 [12:30<21:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1206/3295 [12:30<21:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1207/3295 [12:30<21:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1207/3295 [12:30<21:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1208/3295 [12:31<21:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1208/3295 [12:31<21:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1209/3295 [12:32<21:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1209/3295 [12:32<21:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1210/3295 [12:32<21:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1210/3295 [12:32<21:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1211/3295 [12:33<21:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1211/3295 [12:33<21:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1212/3295 [12:33<21:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1212/3295 [12:33<21:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1213/3295 [12:33<21:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1213/3295 [12:33<21:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1214/3295 [12:35<21:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1214/3295 [12:35<21:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1215/3295 [12:35<21:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1215/3295 [12:35<21:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1216/3295 [12:35<21:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1216/3295 [12:35<21:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1217/3295 [12:37<21:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1217/3295 [12:37<21:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1218/3295 [12:37<21:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1218/3295 [12:37<21:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1219/3295 [12:37<21:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1219/3295 [12:37<21:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1220/3295 [12:39<21:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1220/3295 [12:39<21:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1221/3295 [12:39<21:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1221/3295 [12:39<21:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1222/3295 [12:39<21:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1222/3295 [12:39<21:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1223/3295 [12:41<21:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1223/3295 [12:41<21:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1224/3295 [12:41<21:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1224/3295 [12:41<21:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1225/3295 [12:41<21:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1225/3295 [12:41<21:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1226/3295 [12:43<21:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1226/3295 [12:43<21:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1227/3295 [12:43<21:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1227/3295 [12:43<21:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1228/3295 [12:43<21:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1228/3295 [12:43<21:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1229/3295 [12:45<21:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1229/3295 [12:45<21:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1230/3295 [12:45<21:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1230/3295 [12:45<21:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1231/3295 [12:45<21:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1231/3295 [12:45<21:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1232/3295 [12:47<21:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1232/3295 [12:47<21:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1233/3295 [12:47<21:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1233/3295 [12:47<21:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1234/3295 [12:47<21:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1234/3295 [12:47<21:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1235/3295 [12:49<21:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  37%|███▋      | 1235/3295 [12:49<21:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1236/3295 [12:49<21:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1236/3295 [12:49<21:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1237/3295 [12:49<21:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1237/3295 [12:49<21:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1238/3295 [12:51<21:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1238/3295 [12:51<21:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1239/3295 [12:51<21:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1239/3295 [12:51<21:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1240/3295 [12:51<21:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1240/3295 [12:51<21:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1241/3295 [12:52<21:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1241/3295 [12:52<21:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1242/3295 [12:53<21:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1242/3295 [12:53<21:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1243/3295 [12:53<21:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1243/3295 [12:53<21:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1244/3295 [12:54<21:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1244/3295 [12:54<21:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1245/3295 [12:54<21:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1245/3295 [12:54<21:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1246/3295 [12:54<21:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1246/3295 [12:54<21:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1247/3295 [12:56<21:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1247/3295 [12:56<21:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1248/3295 [12:56<21:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1248/3295 [12:56<21:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1249/3295 [12:56<21:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1249/3295 [12:56<21:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1250/3295 [12:58<21:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1250/3295 [12:58<21:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1251/3295 [12:58<21:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1251/3295 [12:58<21:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1252/3295 [12:58<21:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1252/3295 [12:58<21:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1253/3295 [13:00<21:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1253/3295 [13:00<21:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1254/3295 [13:00<21:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1254/3295 [13:00<21:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1255/3295 [13:00<21:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1255/3295 [13:00<21:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1256/3295 [13:02<21:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1256/3295 [13:02<21:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1257/3295 [13:02<21:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1257/3295 [13:02<21:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1258/3295 [13:02<21:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1258/3295 [13:02<21:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1259/3295 [13:03<21:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1259/3295 [13:03<21:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1260/3295 [13:03<21:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1260/3295 [13:03<21:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1261/3295 [13:03<21:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1261/3295 [13:03<21:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1262/3295 [13:05<21:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1262/3295 [13:05<21:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1263/3295 [13:05<21:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1263/3295 [13:05<21:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1264/3295 [13:05<21:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1264/3295 [13:05<21:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1265/3295 [13:07<21:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1265/3295 [13:07<21:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1266/3295 [13:07<21:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1266/3295 [13:07<21:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1267/3295 [13:07<21:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1267/3295 [13:07<21:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1268/3295 [13:09<21:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  38%|███▊      | 1268/3295 [13:09<21:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1269/3295 [13:09<21:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1269/3295 [13:09<21:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1270/3295 [13:09<20:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1270/3295 [13:09<20:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1271/3295 [13:11<21:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1271/3295 [13:11<21:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1272/3295 [13:11<20:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1272/3295 [13:11<20:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1273/3295 [13:11<20:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1273/3295 [13:11<20:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1274/3295 [13:13<20:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1274/3295 [13:13<20:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1275/3295 [13:13<20:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1275/3295 [13:13<20:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1276/3295 [13:13<20:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▊      | 1276/3295 [13:13<20:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1277/3295 [13:15<20:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1277/3295 [13:15<20:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1278/3295 [13:15<20:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1278/3295 [13:15<20:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1279/3295 [13:15<20:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1279/3295 [13:15<20:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1280/3295 [13:17<20:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1280/3295 [13:17<20:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1281/3295 [13:17<20:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1281/3295 [13:17<20:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1282/3295 [13:17<20:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1282/3295 [13:17<20:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1283/3295 [13:18<20:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1283/3295 [13:18<20:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1284/3295 [13:18<20:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1284/3295 [13:18<20:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1285/3295 [13:19<20:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1285/3295 [13:19<20:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1286/3295 [13:20<20:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1286/3295 [13:20<20:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1287/3295 [13:20<20:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1287/3295 [13:20<20:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1288/3295 [13:20<20:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1288/3295 [13:20<20:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1289/3295 [13:22<20:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1289/3295 [13:22<20:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1290/3295 [13:22<20:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1290/3295 [13:22<20:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1291/3295 [13:22<20:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1291/3295 [13:22<20:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1292/3295 [13:24<20:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1292/3295 [13:24<20:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1293/3295 [13:24<20:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1293/3295 [13:24<20:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1294/3295 [13:24<20:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1294/3295 [13:24<20:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1295/3295 [13:26<20:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1295/3295 [13:26<20:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1296/3295 [13:26<20:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1296/3295 [13:26<20:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1297/3295 [13:26<20:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1297/3295 [13:26<20:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1298/3295 [13:28<20:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1298/3295 [13:28<20:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1299/3295 [13:28<20:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1299/3295 [13:28<20:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1300/3295 [13:28<20:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1300/3295 [13:28<20:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1301/3295 [13:30<20:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  39%|███▉      | 1301/3295 [13:30<20:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1302/3295 [13:30<20:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1302/3295 [13:30<20:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1303/3295 [13:30<20:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1303/3295 [13:30<20:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1304/3295 [13:31<20:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1304/3295 [13:31<20:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1305/3295 [13:31<20:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1305/3295 [13:31<20:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1306/3295 [13:32<20:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1306/3295 [13:32<20:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1307/3295 [13:33<20:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1307/3295 [13:33<20:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1308/3295 [13:33<20:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1308/3295 [13:33<20:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1309/3295 [13:33<20:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1309/3295 [13:33<20:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1310/3295 [13:35<20:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1310/3295 [13:35<20:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1311/3295 [13:35<20:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1311/3295 [13:35<20:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1312/3295 [13:35<20:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1312/3295 [13:35<20:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1313/3295 [13:37<20:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1313/3295 [13:37<20:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1314/3295 [13:37<20:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1314/3295 [13:37<20:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1315/3295 [13:37<20:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1315/3295 [13:37<20:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1316/3295 [13:39<20:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1316/3295 [13:39<20:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1317/3295 [13:39<20:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|███▉      | 1317/3295 [13:39<20:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1318/3295 [13:39<20:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1318/3295 [13:39<20:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1319/3295 [13:41<20:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1319/3295 [13:41<20:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1320/3295 [13:41<20:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1320/3295 [13:41<20:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1321/3295 [13:41<20:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1321/3295 [13:41<20:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1322/3295 [13:43<20:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1322/3295 [13:43<20:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1323/3295 [13:43<20:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1323/3295 [13:43<20:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1324/3295 [13:43<20:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1324/3295 [13:43<20:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1325/3295 [13:45<20:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1325/3295 [13:45<20:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1326/3295 [13:45<20:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1326/3295 [13:45<20:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1327/3295 [13:45<20:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1327/3295 [13:45<20:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1328/3295 [13:46<20:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1328/3295 [13:46<20:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1329/3295 [13:46<20:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1329/3295 [13:46<20:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1330/3295 [13:46<20:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1330/3295 [13:46<20:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1331/3295 [13:48<20:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1331/3295 [13:48<20:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1332/3295 [13:48<20:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1332/3295 [13:48<20:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1333/3295 [13:48<20:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1333/3295 [13:48<20:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1334/3295 [13:50<20:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  40%|████      | 1334/3295 [13:50<20:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1335/3295 [13:50<20:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1335/3295 [13:50<20:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1336/3295 [13:50<20:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1336/3295 [13:50<20:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1337/3295 [13:52<20:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1337/3295 [13:52<20:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1338/3295 [13:52<20:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1338/3295 [13:52<20:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1339/3295 [13:52<20:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1339/3295 [13:52<20:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1340/3295 [13:54<20:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1340/3295 [13:54<20:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1341/3295 [13:54<20:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1341/3295 [13:54<20:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1342/3295 [13:54<20:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1342/3295 [13:54<20:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1343/3295 [13:56<20:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1343/3295 [13:56<20:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1344/3295 [13:56<20:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1344/3295 [13:56<20:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1345/3295 [13:56<20:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1345/3295 [13:56<20:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1346/3295 [13:58<20:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1346/3295 [13:58<20:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1347/3295 [13:58<20:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1347/3295 [13:58<20:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1348/3295 [13:58<20:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1348/3295 [13:58<20:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1349/3295 [14:00<20:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1349/3295 [14:00<20:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1350/3295 [14:00<20:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1350/3295 [14:00<20:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1351/3295 [14:00<20:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1351/3295 [14:00<20:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1352/3295 [14:01<20:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1352/3295 [14:01<20:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1353/3295 [14:01<20:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1353/3295 [14:01<20:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1354/3295 [14:01<20:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1354/3295 [14:01<20:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1355/3295 [14:03<20:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1355/3295 [14:03<20:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1356/3295 [14:03<20:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1356/3295 [14:03<20:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1357/3295 [14:03<20:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1357/3295 [14:03<20:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1358/3295 [14:05<20:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1358/3295 [14:05<20:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1359/3295 [14:05<20:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████      | 1359/3295 [14:05<20:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1360/3295 [14:05<20:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1360/3295 [14:05<20:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1361/3295 [14:07<20:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1361/3295 [14:07<20:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1362/3295 [14:07<20:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1362/3295 [14:07<20:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1363/3295 [14:07<20:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1363/3295 [14:07<20:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1364/3295 [14:08<20:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1364/3295 [14:08<20:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1365/3295 [14:09<20:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1365/3295 [14:09<20:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1366/3295 [14:09<19:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1366/3295 [14:09<19:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1367/3295 [14:10<20:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  41%|████▏     | 1367/3295 [14:10<20:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1368/3295 [14:10<19:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1368/3295 [14:10<19:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1369/3295 [14:10<19:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1369/3295 [14:10<19:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1370/3295 [14:12<19:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1370/3295 [14:12<19:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1371/3295 [14:12<19:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1371/3295 [14:12<19:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1372/3295 [14:12<19:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1372/3295 [14:12<19:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1373/3295 [14:14<19:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1373/3295 [14:14<19:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1374/3295 [14:14<19:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1374/3295 [14:14<19:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1375/3295 [14:14<19:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1375/3295 [14:14<19:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1376/3295 [14:16<19:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1376/3295 [14:16<19:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1377/3295 [14:16<19:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1377/3295 [14:16<19:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1378/3295 [14:16<19:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1378/3295 [14:16<19:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1379/3295 [14:17<19:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1379/3295 [14:17<19:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1380/3295 [14:17<19:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1380/3295 [14:17<19:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1381/3295 [14:18<19:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1381/3295 [14:18<19:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1382/3295 [14:19<19:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1382/3295 [14:19<19:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1383/3295 [14:19<19:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1383/3295 [14:19<19:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1384/3295 [14:19<19:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1384/3295 [14:19<19:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1385/3295 [14:21<19:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1385/3295 [14:21<19:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1386/3295 [14:21<19:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1386/3295 [14:21<19:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1387/3295 [14:21<19:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1387/3295 [14:21<19:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1388/3295 [14:23<19:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1388/3295 [14:23<19:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1389/3295 [14:23<19:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1389/3295 [14:23<19:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1390/3295 [14:23<19:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1390/3295 [14:23<19:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1391/3295 [14:25<19:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1391/3295 [14:25<19:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1392/3295 [14:25<19:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1392/3295 [14:25<19:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1393/3295 [14:25<19:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1393/3295 [14:25<19:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1394/3295 [14:27<19:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1394/3295 [14:27<19:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1395/3295 [14:27<19:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1395/3295 [14:27<19:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1396/3295 [14:27<19:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1396/3295 [14:27<19:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1397/3295 [14:28<19:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1397/3295 [14:28<19:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1398/3295 [14:28<19:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1398/3295 [14:28<19:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1399/3295 [14:29<19:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1399/3295 [14:29<19:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1400/3295 [14:30<19:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  42%|████▏     | 1400/3295 [14:30<19:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1401/3295 [14:30<19:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1401/3295 [14:30<19:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1402/3295 [14:30<19:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1402/3295 [14:30<19:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1403/3295 [14:32<19:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1403/3295 [14:32<19:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1404/3295 [14:32<19:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1404/3295 [14:32<19:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1405/3295 [14:32<19:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1405/3295 [14:32<19:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1406/3295 [14:34<19:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1406/3295 [14:34<19:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1407/3295 [14:34<19:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1407/3295 [14:34<19:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1408/3295 [14:34<19:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1408/3295 [14:34<19:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1409/3295 [14:36<19:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1409/3295 [14:36<19:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1410/3295 [14:36<19:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1410/3295 [14:36<19:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1411/3295 [14:36<19:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1411/3295 [14:36<19:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1412/3295 [14:38<19:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1412/3295 [14:38<19:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1413/3295 [14:38<19:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1413/3295 [14:38<19:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1414/3295 [14:38<19:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1414/3295 [14:38<19:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1415/3295 [14:39<19:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1415/3295 [14:39<19:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1416/3295 [14:39<19:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1416/3295 [14:39<19:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1417/3295 [14:39<19:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1417/3295 [14:39<19:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1418/3295 [14:41<19:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1418/3295 [14:41<19:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1419/3295 [14:41<19:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1419/3295 [14:41<19:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1420/3295 [14:41<19:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1420/3295 [14:41<19:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1421/3295 [14:43<19:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1421/3295 [14:43<19:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1422/3295 [14:43<19:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1422/3295 [14:43<19:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1423/3295 [14:43<19:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1423/3295 [14:43<19:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1424/3295 [14:45<19:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1424/3295 [14:45<19:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1425/3295 [14:45<19:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1425/3295 [14:45<19:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1426/3295 [14:45<19:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1426/3295 [14:45<19:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1427/3295 [14:47<19:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1427/3295 [14:47<19:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1428/3295 [14:47<19:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1428/3295 [14:47<19:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1429/3295 [14:47<19:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1429/3295 [14:47<19:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1430/3295 [14:49<19:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1430/3295 [14:49<19:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1431/3295 [14:49<19:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1431/3295 [14:49<19:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1432/3295 [14:49<19:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1432/3295 [14:49<19:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1433/3295 [14:50<19:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  43%|████▎     | 1433/3295 [14:50<19:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1434/3295 [14:51<19:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1434/3295 [14:51<19:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1435/3295 [14:51<19:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1435/3295 [14:51<19:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1436/3295 [14:52<19:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1436/3295 [14:52<19:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1437/3295 [14:52<19:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1437/3295 [14:52<19:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1438/3295 [14:53<19:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1438/3295 [14:53<19:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1439/3295 [14:54<19:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1439/3295 [14:54<19:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1440/3295 [14:54<19:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1440/3295 [14:54<19:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1441/3295 [14:54<19:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▎     | 1441/3295 [14:54<19:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1442/3295 [14:56<19:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1442/3295 [14:56<19:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1443/3295 [14:56<19:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1443/3295 [14:56<19:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1444/3295 [14:56<19:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1444/3295 [14:56<19:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1445/3295 [14:58<19:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1445/3295 [14:58<19:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1446/3295 [14:58<19:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1446/3295 [14:58<19:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1447/3295 [14:58<19:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1447/3295 [14:58<19:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1448/3295 [15:00<19:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1448/3295 [15:00<19:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1449/3295 [15:00<19:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1449/3295 [15:00<19:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1450/3295 [15:00<19:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1450/3295 [15:00<19:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1451/3295 [15:02<19:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1451/3295 [15:02<19:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1452/3295 [15:02<19:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1452/3295 [15:02<19:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1453/3295 [15:02<19:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1453/3295 [15:02<19:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1454/3295 [15:03<19:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1454/3295 [15:03<19:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1455/3295 [15:04<19:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1455/3295 [15:04<19:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1456/3295 [15:04<19:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1456/3295 [15:04<19:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1457/3295 [15:05<19:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1457/3295 [15:05<19:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1458/3295 [15:05<19:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1458/3295 [15:05<19:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1459/3295 [15:05<18:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1459/3295 [15:05<18:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1460/3295 [15:07<19:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1460/3295 [15:07<19:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1461/3295 [15:07<18:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1461/3295 [15:07<18:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1462/3295 [15:07<18:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1462/3295 [15:07<18:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1463/3295 [15:09<18:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1463/3295 [15:09<18:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1464/3295 [15:09<18:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1464/3295 [15:09<18:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1465/3295 [15:09<18:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1465/3295 [15:09<18:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1466/3295 [15:11<18:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  44%|████▍     | 1466/3295 [15:11<18:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1467/3295 [15:11<18:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1467/3295 [15:11<18:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1468/3295 [15:11<18:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1468/3295 [15:11<18:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1469/3295 [15:13<18:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1469/3295 [15:13<18:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1470/3295 [15:13<18:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1470/3295 [15:13<18:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1471/3295 [15:13<18:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1471/3295 [15:13<18:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1472/3295 [15:14<18:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1472/3295 [15:14<18:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1473/3295 [15:14<18:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1473/3295 [15:14<18:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1474/3295 [15:14<18:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1474/3295 [15:14<18:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1475/3295 [15:16<18:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1475/3295 [15:16<18:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1476/3295 [15:16<18:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1476/3295 [15:16<18:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1477/3295 [15:16<18:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1477/3295 [15:16<18:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1478/3295 [15:18<18:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1478/3295 [15:18<18:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1479/3295 [15:18<18:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1479/3295 [15:18<18:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1480/3295 [15:18<18:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1480/3295 [15:18<18:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1481/3295 [15:20<18:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1481/3295 [15:20<18:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1482/3295 [15:20<18:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▍     | 1482/3295 [15:20<18:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1483/3295 [15:20<18:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1483/3295 [15:20<18:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1484/3295 [15:22<18:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1484/3295 [15:22<18:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1485/3295 [15:22<18:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1485/3295 [15:22<18:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1486/3295 [15:22<18:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1486/3295 [15:22<18:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1487/3295 [15:23<18:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1487/3295 [15:23<18:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1488/3295 [15:23<18:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1488/3295 [15:23<18:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1489/3295 [15:23<18:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1489/3295 [15:23<18:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1490/3295 [15:25<18:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1490/3295 [15:25<18:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1491/3295 [15:25<18:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1491/3295 [15:25<18:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1492/3295 [15:25<18:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1492/3295 [15:25<18:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1493/3295 [15:27<18:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1493/3295 [15:27<18:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1494/3295 [15:27<18:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1494/3295 [15:27<18:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1495/3295 [15:27<18:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1495/3295 [15:27<18:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1496/3295 [15:29<18:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1496/3295 [15:29<18:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1497/3295 [15:29<18:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1497/3295 [15:29<18:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1498/3295 [15:29<18:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1498/3295 [15:29<18:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1499/3295 [15:31<18:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  45%|████▌     | 1499/3295 [15:31<18:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1500/3295 [15:31<18:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1500/3295 [15:31<18:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1501/3295 [15:31<18:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1501/3295 [15:31<18:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1502/3295 [15:33<18:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1502/3295 [15:33<18:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1503/3295 [15:33<18:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1503/3295 [15:33<18:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1504/3295 [15:33<18:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1504/3295 [15:33<18:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1505/3295 [15:34<18:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1505/3295 [15:34<18:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1506/3295 [15:34<18:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1506/3295 [15:34<18:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1507/3295 [15:34<18:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1507/3295 [15:34<18:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1508/3295 [15:36<18:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1508/3295 [15:36<18:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1509/3295 [15:36<18:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1509/3295 [15:36<18:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1510/3295 [15:36<18:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1510/3295 [15:36<18:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1511/3295 [15:38<18:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1511/3295 [15:38<18:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1512/3295 [15:38<18:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1512/3295 [15:38<18:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1513/3295 [15:38<18:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1513/3295 [15:38<18:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1514/3295 [15:40<18:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1514/3295 [15:40<18:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1515/3295 [15:40<18:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1515/3295 [15:40<18:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1516/3295 [15:40<18:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1516/3295 [15:40<18:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1517/3295 [15:42<18:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1517/3295 [15:42<18:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1518/3295 [15:42<18:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1518/3295 [15:42<18:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1519/3295 [15:42<18:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1519/3295 [15:42<18:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1520/3295 [15:44<18:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1520/3295 [15:44<18:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1521/3295 [15:44<18:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1521/3295 [15:44<18:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1522/3295 [15:44<18:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1522/3295 [15:44<18:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1523/3295 [15:46<18:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▌     | 1523/3295 [15:46<18:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1524/3295 [15:46<18:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1524/3295 [15:46<18:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1525/3295 [15:46<18:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1525/3295 [15:46<18:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1526/3295 [15:48<18:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1526/3295 [15:48<18:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1527/3295 [15:48<18:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1527/3295 [15:48<18:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1528/3295 [15:48<18:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1528/3295 [15:48<18:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1529/3295 [15:50<18:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1529/3295 [15:50<18:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1530/3295 [15:50<18:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1530/3295 [15:50<18:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1531/3295 [15:50<18:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1531/3295 [15:50<18:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1532/3295 [15:51<18:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  46%|████▋     | 1532/3295 [15:51<18:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1533/3295 [15:51<18:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1533/3295 [15:51<18:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1534/3295 [15:51<18:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1534/3295 [15:51<18:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1535/3295 [15:53<18:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1535/3295 [15:53<18:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1536/3295 [15:53<18:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1536/3295 [15:53<18:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1537/3295 [15:53<18:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1537/3295 [15:53<18:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1538/3295 [15:55<18:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1538/3295 [15:55<18:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1539/3295 [15:55<18:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1539/3295 [15:55<18:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1540/3295 [15:55<18:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1540/3295 [15:55<18:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1541/3295 [15:57<18:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1541/3295 [15:57<18:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1542/3295 [15:57<18:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1542/3295 [15:57<18:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1543/3295 [15:57<18:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1543/3295 [15:57<18:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1544/3295 [15:59<18:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1544/3295 [15:59<18:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1545/3295 [15:59<18:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1545/3295 [15:59<18:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1546/3295 [15:59<18:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1546/3295 [15:59<18:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1547/3295 [16:00<18:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1547/3295 [16:00<18:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1548/3295 [16:00<18:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1548/3295 [16:00<18:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1549/3295 [16:00<18:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1549/3295 [16:00<18:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1550/3295 [16:02<18:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1550/3295 [16:02<18:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1551/3295 [16:02<18:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1551/3295 [16:02<18:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1552/3295 [16:02<18:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1552/3295 [16:02<18:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1553/3295 [16:04<18:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1553/3295 [16:04<18:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1554/3295 [16:04<18:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1554/3295 [16:04<18:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1555/3295 [16:04<17:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1555/3295 [16:04<17:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1556/3295 [16:06<18:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1556/3295 [16:06<18:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1557/3295 [16:06<17:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1557/3295 [16:06<17:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1558/3295 [16:06<17:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1558/3295 [16:06<17:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1559/3295 [16:08<17:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1559/3295 [16:08<17:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1560/3295 [16:08<17:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1560/3295 [16:08<17:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1561/3295 [16:08<17:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1561/3295 [16:08<17:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1562/3295 [16:10<17:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1562/3295 [16:10<17:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1563/3295 [16:10<17:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1563/3295 [16:10<17:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1564/3295 [16:10<17:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1564/3295 [16:10<17:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1565/3295 [16:12<17:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  47%|████▋     | 1565/3295 [16:12<17:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1566/3295 [16:12<17:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1566/3295 [16:12<17:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1567/3295 [16:12<17:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1567/3295 [16:12<17:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1568/3295 [16:14<17:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1568/3295 [16:14<17:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1569/3295 [16:14<17:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1569/3295 [16:14<17:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1570/3295 [16:14<17:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1570/3295 [16:14<17:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1571/3295 [16:16<17:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1571/3295 [16:16<17:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1572/3295 [16:16<17:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1572/3295 [16:16<17:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1573/3295 [16:16<17:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1573/3295 [16:16<17:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1574/3295 [16:17<17:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1574/3295 [16:17<17:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1575/3295 [16:17<17:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1575/3295 [16:17<17:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1576/3295 [16:17<17:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1576/3295 [16:17<17:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1577/3295 [16:19<17:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1577/3295 [16:19<17:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1578/3295 [16:19<17:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1578/3295 [16:19<17:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1579/3295 [16:19<17:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1579/3295 [16:19<17:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1580/3295 [16:21<17:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1580/3295 [16:21<17:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1581/3295 [16:21<17:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1581/3295 [16:21<17:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1582/3295 [16:21<17:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1582/3295 [16:21<17:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1583/3295 [16:23<17:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1583/3295 [16:23<17:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1584/3295 [16:23<17:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1584/3295 [16:23<17:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1585/3295 [16:23<17:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1585/3295 [16:23<17:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1586/3295 [16:25<17:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1586/3295 [16:25<17:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1587/3295 [16:25<17:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1587/3295 [16:25<17:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1588/3295 [16:25<17:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1588/3295 [16:25<17:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1589/3295 [16:27<17:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1589/3295 [16:27<17:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1590/3295 [16:27<17:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1590/3295 [16:27<17:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1591/3295 [16:27<17:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1591/3295 [16:27<17:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1592/3295 [16:28<17:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1592/3295 [16:28<17:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1593/3295 [16:28<17:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1593/3295 [16:28<17:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1594/3295 [16:29<17:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1594/3295 [16:29<17:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1595/3295 [16:30<17:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1595/3295 [16:30<17:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1596/3295 [16:30<17:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1596/3295 [16:30<17:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1597/3295 [16:30<17:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1597/3295 [16:30<17:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1598/3295 [16:32<17:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  48%|████▊     | 1598/3295 [16:32<17:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1599/3295 [16:32<17:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1599/3295 [16:32<17:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1600/3295 [16:32<17:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1600/3295 [16:32<17:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1601/3295 [16:34<17:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1601/3295 [16:34<17:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1602/3295 [16:34<17:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1602/3295 [16:34<17:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1603/3295 [16:34<17:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1603/3295 [16:34<17:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1604/3295 [16:36<17:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1604/3295 [16:36<17:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1605/3295 [16:36<17:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1605/3295 [16:36<17:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1606/3295 [16:36<17:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▊     | 1606/3295 [16:36<17:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1607/3295 [16:38<17:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1607/3295 [16:38<17:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1608/3295 [16:38<17:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1608/3295 [16:38<17:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1609/3295 [16:38<17:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1609/3295 [16:38<17:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1610/3295 [16:40<17:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1610/3295 [16:40<17:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1611/3295 [16:40<17:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1611/3295 [16:40<17:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1612/3295 [16:40<17:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1612/3295 [16:40<17:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1613/3295 [16:41<17:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1613/3295 [16:41<17:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1614/3295 [16:42<17:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1614/3295 [16:42<17:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1615/3295 [16:42<17:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1615/3295 [16:42<17:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1616/3295 [16:43<17:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1616/3295 [16:43<17:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1617/3295 [16:43<17:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1617/3295 [16:43<17:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1618/3295 [16:44<17:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1618/3295 [16:44<17:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1619/3295 [16:45<17:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1619/3295 [16:45<17:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1620/3295 [16:45<17:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1620/3295 [16:45<17:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1621/3295 [16:45<17:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1621/3295 [16:45<17:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1622/3295 [16:47<17:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1622/3295 [16:47<17:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1623/3295 [16:47<17:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1623/3295 [16:47<17:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1624/3295 [16:47<17:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1624/3295 [16:47<17:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1625/3295 [16:49<17:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1625/3295 [16:49<17:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1626/3295 [16:49<17:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1626/3295 [16:49<17:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1627/3295 [16:49<17:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1627/3295 [16:49<17:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1628/3295 [16:51<17:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1628/3295 [16:51<17:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1629/3295 [16:51<17:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1629/3295 [16:51<17:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1630/3295 [16:51<17:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1630/3295 [16:51<17:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1631/3295 [16:52<17:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  49%|████▉     | 1631/3295 [16:52<17:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1632/3295 [16:53<17:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1632/3295 [16:53<17:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1633/3295 [16:53<17:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1633/3295 [16:53<17:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1634/3295 [16:54<17:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1634/3295 [16:54<17:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1635/3295 [16:54<17:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1635/3295 [16:54<17:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1636/3295 [16:54<17:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1636/3295 [16:54<17:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1637/3295 [16:56<17:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1637/3295 [16:56<17:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1638/3295 [16:56<17:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1638/3295 [16:56<17:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1639/3295 [16:56<17:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1639/3295 [16:56<17:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1640/3295 [16:58<17:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1640/3295 [16:58<17:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1641/3295 [16:58<17:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1641/3295 [16:58<17:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1642/3295 [16:58<17:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1642/3295 [16:58<17:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1643/3295 [17:00<17:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1643/3295 [17:00<17:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1644/3295 [17:00<17:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1644/3295 [17:00<17:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1645/3295 [17:00<17:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1645/3295 [17:00<17:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1646/3295 [17:01<17:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1646/3295 [17:01<17:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1647/3295 [17:01<17:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|████▉     | 1647/3295 [17:01<17:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1648/3295 [17:01<17:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1648/3295 [17:01<17:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1649/3295 [17:03<17:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1649/3295 [17:03<17:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1650/3295 [17:03<17:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1650/3295 [17:03<17:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1651/3295 [17:03<16:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1651/3295 [17:03<16:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1652/3295 [17:05<17:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1652/3295 [17:05<17:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1653/3295 [17:05<16:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1653/3295 [17:05<16:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1654/3295 [17:05<16:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1654/3295 [17:05<16:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1655/3295 [17:07<16:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1655/3295 [17:07<16:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1656/3295 [17:07<16:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1656/3295 [17:07<16:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1657/3295 [17:07<16:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1657/3295 [17:07<16:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1658/3295 [17:09<16:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1658/3295 [17:09<16:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1659/3295 [17:09<16:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1659/3295 [17:09<16:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1660/3295 [17:09<16:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1660/3295 [17:09<16:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1661/3295 [17:11<16:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1661/3295 [17:11<16:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1662/3295 [17:11<16:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1662/3295 [17:11<16:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1663/3295 [17:11<16:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  50%|█████     | 1663/3295 [17:11<16:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1664/3295 [17:13<16:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1664/3295 [17:13<16:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1665/3295 [17:13<16:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1665/3295 [17:13<16:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1666/3295 [17:13<16:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1666/3295 [17:13<16:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1667/3295 [17:14<16:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1667/3295 [17:14<16:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1668/3295 [17:14<16:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1668/3295 [17:14<16:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1669/3295 [17:14<16:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1669/3295 [17:14<16:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1670/3295 [17:16<16:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1670/3295 [17:16<16:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1671/3295 [17:16<16:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1671/3295 [17:16<16:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1672/3295 [17:16<16:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1672/3295 [17:16<16:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1673/3295 [17:18<16:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1673/3295 [17:18<16:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1674/3295 [17:18<16:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1674/3295 [17:18<16:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1675/3295 [17:18<16:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1675/3295 [17:18<16:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1676/3295 [17:20<16:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1676/3295 [17:20<16:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1677/3295 [17:20<16:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1677/3295 [17:20<16:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1678/3295 [17:20<16:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1678/3295 [17:20<16:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1679/3295 [17:22<16:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1679/3295 [17:22<16:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1680/3295 [17:22<16:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1680/3295 [17:22<16:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1681/3295 [17:22<16:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1681/3295 [17:22<16:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1682/3295 [17:23<16:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1682/3295 [17:23<16:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1683/3295 [17:24<16:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1683/3295 [17:24<16:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1684/3295 [17:24<16:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1684/3295 [17:24<16:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1685/3295 [17:25<16:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1685/3295 [17:25<16:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1686/3295 [17:25<16:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1686/3295 [17:25<16:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1687/3295 [17:25<16:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1687/3295 [17:25<16:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1688/3295 [17:27<16:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████     | 1688/3295 [17:27<16:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1689/3295 [17:27<16:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1689/3295 [17:27<16:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1690/3295 [17:27<16:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1690/3295 [17:27<16:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1691/3295 [17:29<16:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1691/3295 [17:29<16:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1692/3295 [17:29<16:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1692/3295 [17:29<16:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1693/3295 [17:29<16:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1693/3295 [17:29<16:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1694/3295 [17:31<16:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1694/3295 [17:31<16:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1695/3295 [17:31<16:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1695/3295 [17:31<16:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1696/3295 [17:31<16:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  51%|█████▏    | 1696/3295 [17:31<16:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1697/3295 [17:33<16:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1697/3295 [17:33<16:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1698/3295 [17:33<16:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1698/3295 [17:33<16:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1699/3295 [17:33<16:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1699/3295 [17:33<16:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1700/3295 [17:35<16:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1700/3295 [17:35<16:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1701/3295 [17:35<16:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1701/3295 [17:35<16:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1702/3295 [17:35<16:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1702/3295 [17:35<16:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1703/3295 [17:36<16:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1703/3295 [17:36<16:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1704/3295 [17:37<16:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1704/3295 [17:37<16:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1705/3295 [17:37<16:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1705/3295 [17:37<16:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1706/3295 [17:38<16:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1706/3295 [17:38<16:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1707/3295 [17:38<16:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1707/3295 [17:38<16:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1708/3295 [17:38<16:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1708/3295 [17:38<16:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1709/3295 [17:40<16:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1709/3295 [17:40<16:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1710/3295 [17:40<16:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1710/3295 [17:40<16:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1711/3295 [17:40<16:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1711/3295 [17:40<16:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1712/3295 [17:42<16:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1712/3295 [17:42<16:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1713/3295 [17:42<16:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1713/3295 [17:42<16:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1714/3295 [17:42<16:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1714/3295 [17:42<16:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1715/3295 [17:44<16:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1715/3295 [17:44<16:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1716/3295 [17:44<16:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1716/3295 [17:44<16:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1717/3295 [17:44<16:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1717/3295 [17:44<16:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1718/3295 [17:45<16:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1718/3295 [17:45<16:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1719/3295 [17:46<16:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1719/3295 [17:46<16:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1720/3295 [17:46<16:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1720/3295 [17:46<16:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1721/3295 [17:47<16:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1721/3295 [17:47<16:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1722/3295 [17:47<16:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1722/3295 [17:47<16:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1723/3295 [17:48<16:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1723/3295 [17:48<16:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1724/3295 [17:49<16:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1724/3295 [17:49<16:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1725/3295 [17:49<16:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1725/3295 [17:49<16:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1726/3295 [17:49<16:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1726/3295 [17:49<16:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1727/3295 [17:51<16:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1727/3295 [17:51<16:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1728/3295 [17:51<16:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1728/3295 [17:51<16:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1729/3295 [17:51<16:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  52%|█████▏    | 1729/3295 [17:51<16:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1730/3295 [17:53<16:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1730/3295 [17:53<16:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1731/3295 [17:53<16:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1731/3295 [17:53<16:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1732/3295 [17:53<16:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1732/3295 [17:53<16:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1733/3295 [17:54<16:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1733/3295 [17:54<16:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1734/3295 [17:55<16:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1734/3295 [17:55<16:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1735/3295 [17:55<16:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1735/3295 [17:55<16:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1736/3295 [17:56<16:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1736/3295 [17:56<16:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1737/3295 [17:56<16:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1737/3295 [17:56<16:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1738/3295 [17:57<16:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1738/3295 [17:57<16:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1739/3295 [17:58<16:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1739/3295 [17:58<16:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1740/3295 [17:58<16:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1740/3295 [17:58<16:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1741/3295 [17:58<16:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1741/3295 [17:58<16:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1742/3295 [18:00<16:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1742/3295 [18:00<16:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1743/3295 [18:00<16:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1743/3295 [18:00<16:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1744/3295 [18:00<16:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1744/3295 [18:00<16:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1745/3295 [18:02<16:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1745/3295 [18:02<16:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1746/3295 [18:02<16:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1746/3295 [18:02<16:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1747/3295 [18:02<15:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1747/3295 [18:02<15:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1748/3295 [18:04<15:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1748/3295 [18:04<15:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1749/3295 [18:04<15:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1749/3295 [18:04<15:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1750/3295 [18:04<15:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1750/3295 [18:04<15:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1751/3295 [18:06<15:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1751/3295 [18:06<15:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1752/3295 [18:06<15:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1752/3295 [18:06<15:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1753/3295 [18:06<15:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1753/3295 [18:06<15:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1754/3295 [18:07<15:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1754/3295 [18:07<15:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1755/3295 [18:08<15:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1755/3295 [18:08<15:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1756/3295 [18:08<15:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1756/3295 [18:08<15:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1757/3295 [18:09<15:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1757/3295 [18:09<15:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1758/3295 [18:10<15:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1758/3295 [18:10<15:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1759/3295 [18:10<15:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1759/3295 [18:10<15:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1760/3295 [18:11<15:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1760/3295 [18:11<15:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1761/3295 [18:11<15:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1761/3295 [18:11<15:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1762/3295 [18:11<15:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  53%|█████▎    | 1762/3295 [18:11<15:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1763/3295 [18:13<15:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1763/3295 [18:13<15:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1764/3295 [18:13<15:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1764/3295 [18:13<15:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1765/3295 [18:13<15:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1765/3295 [18:13<15:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1766/3295 [18:15<15:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1766/3295 [18:15<15:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1767/3295 [18:15<15:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1767/3295 [18:15<15:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1768/3295 [18:15<15:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1768/3295 [18:15<15:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1769/3295 [18:17<15:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1769/3295 [18:17<15:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1770/3295 [18:17<15:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1770/3295 [18:17<15:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1771/3295 [18:17<15:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▎    | 1771/3295 [18:17<15:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1772/3295 [18:18<15:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1772/3295 [18:18<15:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1773/3295 [18:19<15:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1773/3295 [18:19<15:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1774/3295 [18:19<15:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1774/3295 [18:19<15:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1775/3295 [18:20<15:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1775/3295 [18:20<15:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1776/3295 [18:21<15:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1776/3295 [18:21<15:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1777/3295 [18:21<15:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1777/3295 [18:21<15:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1778/3295 [18:22<15:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1778/3295 [18:22<15:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1779/3295 [18:22<15:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1779/3295 [18:22<15:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1780/3295 [18:22<15:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1780/3295 [18:22<15:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1781/3295 [18:24<15:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1781/3295 [18:24<15:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1782/3295 [18:24<15:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1782/3295 [18:24<15:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1783/3295 [18:24<15:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1783/3295 [18:24<15:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1784/3295 [18:26<15:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1784/3295 [18:26<15:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1785/3295 [18:26<15:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1785/3295 [18:26<15:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1786/3295 [18:26<15:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1786/3295 [18:26<15:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1787/3295 [18:28<15:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1787/3295 [18:28<15:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1788/3295 [18:28<15:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1788/3295 [18:28<15:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1789/3295 [18:28<15:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1789/3295 [18:28<15:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1790/3295 [18:29<15:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1790/3295 [18:29<15:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1791/3295 [18:30<15:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1791/3295 [18:30<15:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1792/3295 [18:30<15:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1792/3295 [18:30<15:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1793/3295 [18:31<15:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1793/3295 [18:31<15:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1794/3295 [18:31<15:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1794/3295 [18:31<15:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1795/3295 [18:32<15:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  54%|█████▍    | 1795/3295 [18:32<15:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1796/3295 [18:33<15:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1796/3295 [18:33<15:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1797/3295 [18:33<15:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1797/3295 [18:33<15:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1798/3295 [18:33<15:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1798/3295 [18:33<15:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1799/3295 [18:35<15:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1799/3295 [18:35<15:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1800/3295 [18:35<15:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1800/3295 [18:35<15:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1801/3295 [18:35<15:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1801/3295 [18:35<15:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1802/3295 [18:37<15:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1802/3295 [18:37<15:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1803/3295 [18:37<15:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1803/3295 [18:37<15:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1804/3295 [18:37<15:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1804/3295 [18:37<15:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1805/3295 [18:39<15:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1805/3295 [18:39<15:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1806/3295 [18:39<15:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1806/3295 [18:39<15:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1807/3295 [18:39<15:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1807/3295 [18:39<15:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1808/3295 [18:40<15:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1808/3295 [18:40<15:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1809/3295 [18:41<15:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1809/3295 [18:41<15:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1810/3295 [18:41<15:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1810/3295 [18:41<15:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1811/3295 [18:42<15:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1811/3295 [18:42<15:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1812/3295 [18:43<15:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▍    | 1812/3295 [18:43<15:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1813/3295 [18:43<15:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1813/3295 [18:43<15:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1814/3295 [18:44<15:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1814/3295 [18:44<15:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1815/3295 [18:44<15:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1815/3295 [18:44<15:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1816/3295 [18:44<15:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1816/3295 [18:44<15:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1817/3295 [18:46<15:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1817/3295 [18:46<15:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1818/3295 [18:46<15:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1818/3295 [18:46<15:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1819/3295 [18:46<15:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1819/3295 [18:46<15:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1820/3295 [18:48<15:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1820/3295 [18:48<15:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1821/3295 [18:48<15:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1821/3295 [18:48<15:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1822/3295 [18:48<15:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1822/3295 [18:48<15:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1823/3295 [18:50<15:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1823/3295 [18:50<15:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1824/3295 [18:50<15:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1824/3295 [18:50<15:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1825/3295 [18:50<15:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1825/3295 [18:50<15:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1826/3295 [18:51<15:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1826/3295 [18:51<15:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1827/3295 [18:52<15:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1827/3295 [18:52<15:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1828/3295 [18:52<15:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  55%|█████▌    | 1828/3295 [18:52<15:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1829/3295 [18:53<15:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1829/3295 [18:53<15:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1830/3295 [18:54<15:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1830/3295 [18:54<15:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1831/3295 [18:54<15:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1831/3295 [18:54<15:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1832/3295 [18:55<15:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1832/3295 [18:55<15:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1833/3295 [18:55<15:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1833/3295 [18:55<15:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1834/3295 [18:55<15:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1834/3295 [18:55<15:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1835/3295 [18:57<15:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1835/3295 [18:57<15:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1836/3295 [18:57<15:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1836/3295 [18:57<15:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1837/3295 [18:57<15:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1837/3295 [18:57<15:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1838/3295 [18:59<15:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1838/3295 [18:59<15:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1839/3295 [18:59<15:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1839/3295 [18:59<15:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1840/3295 [18:59<15:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1840/3295 [18:59<15:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1841/3295 [19:01<15:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1841/3295 [19:01<15:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1842/3295 [19:01<15:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1842/3295 [19:01<15:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1843/3295 [19:01<14:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1843/3295 [19:01<14:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1844/3295 [19:03<14:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1844/3295 [19:03<14:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1845/3295 [19:03<14:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1845/3295 [19:03<14:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1846/3295 [19:03<14:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1846/3295 [19:03<14:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1847/3295 [19:05<14:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1847/3295 [19:05<14:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1848/3295 [19:05<14:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1848/3295 [19:05<14:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1849/3295 [19:05<14:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1849/3295 [19:05<14:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1850/3295 [19:07<14:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1850/3295 [19:07<14:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1851/3295 [19:07<14:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1851/3295 [19:07<14:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1852/3295 [19:07<14:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1852/3295 [19:07<14:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1853/3295 [19:09<14:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▌    | 1853/3295 [19:09<14:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1854/3295 [19:09<14:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1854/3295 [19:09<14:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1855/3295 [19:09<14:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1855/3295 [19:09<14:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1856/3295 [19:10<14:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1856/3295 [19:10<14:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1857/3295 [19:10<14:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1857/3295 [19:10<14:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1858/3295 [19:11<14:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1858/3295 [19:11<14:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1859/3295 [19:12<14:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1859/3295 [19:12<14:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1860/3295 [19:12<14:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1860/3295 [19:12<14:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1861/3295 [19:12<14:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  56%|█████▋    | 1861/3295 [19:12<14:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1862/3295 [19:14<14:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1862/3295 [19:14<14:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1863/3295 [19:14<14:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1863/3295 [19:14<14:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1864/3295 [19:14<14:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1864/3295 [19:14<14:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1865/3295 [19:16<14:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1865/3295 [19:16<14:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1866/3295 [19:16<14:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1866/3295 [19:16<14:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1867/3295 [19:16<14:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1867/3295 [19:16<14:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1868/3295 [19:18<14:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1868/3295 [19:18<14:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1869/3295 [19:18<14:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1869/3295 [19:18<14:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1870/3295 [19:18<14:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1870/3295 [19:18<14:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1871/3295 [19:20<14:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1871/3295 [19:20<14:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1872/3295 [19:20<14:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1872/3295 [19:20<14:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1873/3295 [19:20<14:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1873/3295 [19:20<14:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1874/3295 [19:22<14:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1874/3295 [19:22<14:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1875/3295 [19:22<14:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1875/3295 [19:22<14:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1876/3295 [19:22<14:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1876/3295 [19:22<14:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1877/3295 [19:24<14:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1877/3295 [19:24<14:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1878/3295 [19:24<14:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1878/3295 [19:24<14:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1879/3295 [19:24<14:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1879/3295 [19:24<14:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1880/3295 [19:25<14:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1880/3295 [19:25<14:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1881/3295 [19:25<14:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1881/3295 [19:25<14:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1882/3295 [19:25<14:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1882/3295 [19:25<14:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1883/3295 [19:27<14:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1883/3295 [19:27<14:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1884/3295 [19:27<14:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1884/3295 [19:27<14:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1885/3295 [19:27<14:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1885/3295 [19:27<14:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1886/3295 [19:29<14:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1886/3295 [19:29<14:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1887/3295 [19:29<14:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1887/3295 [19:29<14:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1888/3295 [19:29<14:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1888/3295 [19:29<14:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1889/3295 [19:31<14:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1889/3295 [19:31<14:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1890/3295 [19:31<14:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1890/3295 [19:31<14:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1891/3295 [19:31<14:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1891/3295 [19:31<14:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1892/3295 [19:33<14:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1892/3295 [19:33<14:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1893/3295 [19:33<14:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1893/3295 [19:33<14:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1894/3295 [19:33<14:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  57%|█████▋    | 1894/3295 [19:33<14:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1895/3295 [19:35<14:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1895/3295 [19:35<14:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1896/3295 [19:35<14:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1896/3295 [19:35<14:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1897/3295 [19:35<14:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1897/3295 [19:35<14:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1898/3295 [19:37<14:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1898/3295 [19:37<14:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1899/3295 [19:37<14:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1899/3295 [19:37<14:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1900/3295 [19:37<14:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1900/3295 [19:37<14:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1901/3295 [19:39<14:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1901/3295 [19:39<14:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1902/3295 [19:39<14:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1902/3295 [19:39<14:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1903/3295 [19:39<14:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1903/3295 [19:39<14:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1904/3295 [19:41<14:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1904/3295 [19:41<14:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1905/3295 [19:41<14:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1905/3295 [19:41<14:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1906/3295 [19:41<14:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1906/3295 [19:41<14:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1907/3295 [19:42<14:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1907/3295 [19:42<14:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1908/3295 [19:42<14:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1908/3295 [19:42<14:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1909/3295 [19:42<14:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1909/3295 [19:42<14:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1910/3295 [19:44<14:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1910/3295 [19:44<14:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1911/3295 [19:44<14:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1911/3295 [19:44<14:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1912/3295 [19:44<14:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1912/3295 [19:44<14:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1913/3295 [19:46<14:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1913/3295 [19:46<14:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1914/3295 [19:46<14:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1914/3295 [19:46<14:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1915/3295 [19:46<14:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1915/3295 [19:46<14:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1916/3295 [19:48<14:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1916/3295 [19:48<14:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1917/3295 [19:48<14:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1917/3295 [19:48<14:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1918/3295 [19:48<14:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1918/3295 [19:48<14:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1919/3295 [19:50<14:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1919/3295 [19:50<14:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1920/3295 [19:50<14:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1920/3295 [19:50<14:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1921/3295 [19:50<14:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1921/3295 [19:50<14:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1922/3295 [19:52<14:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1922/3295 [19:52<14:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1923/3295 [19:52<14:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1923/3295 [19:52<14:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1924/3295 [19:52<14:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1924/3295 [19:52<14:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1925/3295 [19:54<14:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1925/3295 [19:54<14:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1926/3295 [19:54<14:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1926/3295 [19:54<14:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1927/3295 [19:54<14:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  58%|█████▊    | 1927/3295 [19:54<14:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1928/3295 [19:56<14:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1928/3295 [19:56<14:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1929/3295 [19:56<14:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1929/3295 [19:56<14:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1930/3295 [19:56<14:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1930/3295 [19:56<14:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1931/3295 [19:57<14:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1931/3295 [19:57<14:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1932/3295 [19:57<14:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1932/3295 [19:57<14:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1933/3295 [19:58<14:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1933/3295 [19:58<14:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1934/3295 [19:59<14:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1934/3295 [19:59<14:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1935/3295 [19:59<14:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▊    | 1935/3295 [19:59<14:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1936/3295 [19:59<14:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1936/3295 [19:59<14:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1937/3295 [20:01<14:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1937/3295 [20:01<14:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1938/3295 [20:01<14:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1938/3295 [20:01<14:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1939/3295 [20:01<14:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1939/3295 [20:01<14:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1940/3295 [20:03<14:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1940/3295 [20:03<14:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1941/3295 [20:03<13:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1941/3295 [20:03<13:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1942/3295 [20:03<13:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1942/3295 [20:03<13:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1943/3295 [20:05<13:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1943/3295 [20:05<13:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1944/3295 [20:05<13:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1944/3295 [20:05<13:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1945/3295 [20:05<13:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1945/3295 [20:05<13:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1946/3295 [20:07<13:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1946/3295 [20:07<13:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1947/3295 [20:07<13:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1947/3295 [20:07<13:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1948/3295 [20:07<13:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1948/3295 [20:07<13:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1949/3295 [20:09<13:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1949/3295 [20:09<13:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1950/3295 [20:09<13:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1950/3295 [20:09<13:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1951/3295 [20:09<13:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1951/3295 [20:09<13:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1952/3295 [20:10<13:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1952/3295 [20:10<13:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1953/3295 [20:10<13:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1953/3295 [20:10<13:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1954/3295 [20:10<13:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1954/3295 [20:10<13:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1955/3295 [20:12<13:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1955/3295 [20:12<13:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1956/3295 [20:12<13:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1956/3295 [20:12<13:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1957/3295 [20:12<13:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1957/3295 [20:12<13:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1958/3295 [20:14<13:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1958/3295 [20:14<13:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1959/3295 [20:14<13:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1959/3295 [20:14<13:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1960/3295 [20:14<13:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  59%|█████▉    | 1960/3295 [20:14<13:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1961/3295 [20:16<13:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1961/3295 [20:16<13:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1962/3295 [20:16<13:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1962/3295 [20:16<13:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1963/3295 [20:16<13:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1963/3295 [20:16<13:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1964/3295 [20:18<13:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1964/3295 [20:18<13:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1965/3295 [20:18<13:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1965/3295 [20:18<13:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1966/3295 [20:18<13:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1966/3295 [20:18<13:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1967/3295 [20:20<13:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1967/3295 [20:20<13:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1968/3295 [20:20<13:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1968/3295 [20:20<13:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1969/3295 [20:20<13:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1969/3295 [20:20<13:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1970/3295 [20:21<13:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1970/3295 [20:21<13:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1971/3295 [20:21<13:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1971/3295 [20:21<13:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1972/3295 [20:22<13:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1972/3295 [20:22<13:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1973/3295 [20:23<13:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1973/3295 [20:23<13:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1974/3295 [20:23<13:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1974/3295 [20:23<13:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1975/3295 [20:23<13:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1975/3295 [20:23<13:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1976/3295 [20:25<13:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|█████▉    | 1976/3295 [20:25<13:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1977/3295 [20:25<13:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1977/3295 [20:25<13:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1978/3295 [20:25<13:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1978/3295 [20:25<13:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1979/3295 [20:27<13:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1979/3295 [20:27<13:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1980/3295 [20:27<13:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1980/3295 [20:27<13:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1981/3295 [20:27<13:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1981/3295 [20:27<13:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1982/3295 [20:29<13:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1982/3295 [20:29<13:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1983/3295 [20:29<13:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1983/3295 [20:29<13:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1984/3295 [20:29<13:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1984/3295 [20:29<13:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1985/3295 [20:31<13:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1985/3295 [20:31<13:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1986/3295 [20:31<13:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1986/3295 [20:31<13:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1987/3295 [20:31<13:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1987/3295 [20:31<13:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1988/3295 [20:33<13:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1988/3295 [20:33<13:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1989/3295 [20:33<13:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1989/3295 [20:33<13:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1990/3295 [20:33<13:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1990/3295 [20:33<13:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1991/3295 [20:35<13:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1991/3295 [20:35<13:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1992/3295 [20:35<13:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1992/3295 [20:35<13:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1993/3295 [20:35<13:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  60%|██████    | 1993/3295 [20:35<13:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1994/3295 [20:37<13:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1994/3295 [20:37<13:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1995/3295 [20:37<13:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1995/3295 [20:37<13:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1996/3295 [20:37<13:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1996/3295 [20:37<13:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1997/3295 [20:39<13:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1997/3295 [20:39<13:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1998/3295 [20:39<13:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1998/3295 [20:39<13:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1999/3295 [20:39<13:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 1999/3295 [20:39<13:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2000/3295 [20:40<13:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2000/3295 [20:40<13:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2001/3295 [20:40<13:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2001/3295 [20:40<13:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2002/3295 [20:40<13:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2002/3295 [20:40<13:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2003/3295 [20:42<13:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2003/3295 [20:42<13:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2004/3295 [20:42<13:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2004/3295 [20:42<13:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2005/3295 [20:42<13:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2005/3295 [20:42<13:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2006/3295 [20:44<13:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2006/3295 [20:44<13:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2007/3295 [20:44<13:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2007/3295 [20:44<13:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2008/3295 [20:44<13:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2008/3295 [20:44<13:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2009/3295 [20:46<13:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2009/3295 [20:46<13:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2010/3295 [20:46<13:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2010/3295 [20:46<13:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2011/3295 [20:46<13:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2011/3295 [20:46<13:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2012/3295 [20:48<13:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2012/3295 [20:48<13:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2013/3295 [20:48<13:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2013/3295 [20:48<13:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2014/3295 [20:48<13:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2014/3295 [20:48<13:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2015/3295 [20:49<13:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2015/3295 [20:49<13:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2016/3295 [20:49<13:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2016/3295 [20:49<13:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2017/3295 [20:49<13:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2017/3295 [20:49<13:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2018/3295 [20:51<13:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████    | 2018/3295 [20:51<13:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2019/3295 [20:51<13:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2019/3295 [20:51<13:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2020/3295 [20:51<13:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2020/3295 [20:51<13:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2021/3295 [20:53<13:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2021/3295 [20:53<13:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2022/3295 [20:53<13:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2022/3295 [20:53<13:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2023/3295 [20:53<13:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2023/3295 [20:53<13:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2024/3295 [20:55<13:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2024/3295 [20:55<13:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2025/3295 [20:55<13:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2025/3295 [20:55<13:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2026/3295 [20:55<13:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  61%|██████▏   | 2026/3295 [20:55<13:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2027/3295 [20:57<13:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2027/3295 [20:57<13:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2028/3295 [20:57<13:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2028/3295 [20:57<13:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2029/3295 [20:57<13:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2029/3295 [20:57<13:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2030/3295 [20:59<13:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2030/3295 [20:59<13:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2031/3295 [20:59<13:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2031/3295 [20:59<13:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2032/3295 [20:59<13:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2032/3295 [20:59<13:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2033/3295 [21:01<13:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2033/3295 [21:01<13:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2034/3295 [21:01<13:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2034/3295 [21:01<13:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2035/3295 [21:01<13:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2035/3295 [21:01<13:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2036/3295 [21:03<13:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2036/3295 [21:03<13:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2037/3295 [21:03<13:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2037/3295 [21:03<13:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2038/3295 [21:03<12:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2038/3295 [21:03<12:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2039/3295 [21:04<12:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2039/3295 [21:04<12:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2040/3295 [21:04<12:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2040/3295 [21:04<12:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2041/3295 [21:04<12:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2041/3295 [21:04<12:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2042/3295 [21:06<12:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2042/3295 [21:06<12:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2043/3295 [21:06<12:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2043/3295 [21:06<12:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2044/3295 [21:06<12:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2044/3295 [21:06<12:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2045/3295 [21:08<12:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2045/3295 [21:08<12:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2046/3295 [21:08<12:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2046/3295 [21:08<12:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2047/3295 [21:08<12:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2047/3295 [21:08<12:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2048/3295 [21:10<12:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2048/3295 [21:10<12:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2049/3295 [21:10<12:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2049/3295 [21:10<12:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2050/3295 [21:10<12:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2050/3295 [21:10<12:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2051/3295 [21:12<12:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2051/3295 [21:12<12:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2052/3295 [21:12<12:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2052/3295 [21:12<12:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2053/3295 [21:12<12:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2053/3295 [21:12<12:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2054/3295 [21:14<12:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2054/3295 [21:14<12:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2055/3295 [21:14<12:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2055/3295 [21:14<12:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2056/3295 [21:14<12:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2056/3295 [21:14<12:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2057/3295 [21:16<12:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2057/3295 [21:16<12:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2058/3295 [21:16<12:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2058/3295 [21:16<12:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2059/3295 [21:16<12:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  62%|██████▏   | 2059/3295 [21:16<12:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2060/3295 [21:18<12:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2060/3295 [21:18<12:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2061/3295 [21:18<12:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2061/3295 [21:18<12:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2062/3295 [21:18<12:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2062/3295 [21:18<12:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2063/3295 [21:19<12:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2063/3295 [21:19<12:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2064/3295 [21:20<12:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2064/3295 [21:20<12:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2065/3295 [21:20<12:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2065/3295 [21:20<12:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2066/3295 [21:21<12:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2066/3295 [21:21<12:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2067/3295 [21:21<12:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2067/3295 [21:21<12:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2068/3295 [21:21<12:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2068/3295 [21:21<12:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2069/3295 [21:23<12:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2069/3295 [21:23<12:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2070/3295 [21:23<12:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2070/3295 [21:23<12:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2071/3295 [21:23<12:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2071/3295 [21:23<12:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2072/3295 [21:25<12:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2072/3295 [21:25<12:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2073/3295 [21:25<12:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2073/3295 [21:25<12:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2074/3295 [21:25<12:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2074/3295 [21:25<12:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2075/3295 [21:27<12:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2075/3295 [21:27<12:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2076/3295 [21:27<12:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2076/3295 [21:27<12:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2077/3295 [21:27<12:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2077/3295 [21:27<12:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2078/3295 [21:29<12:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2078/3295 [21:29<12:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2079/3295 [21:29<12:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2079/3295 [21:29<12:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2080/3295 [21:29<12:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2080/3295 [21:29<12:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2081/3295 [21:31<12:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2081/3295 [21:31<12:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2082/3295 [21:31<12:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2082/3295 [21:31<12:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2083/3295 [21:31<12:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2083/3295 [21:31<12:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2084/3295 [21:32<12:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2084/3295 [21:32<12:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2085/3295 [21:32<12:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2085/3295 [21:32<12:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2086/3295 [21:32<12:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2086/3295 [21:32<12:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2087/3295 [21:34<12:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2087/3295 [21:34<12:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2088/3295 [21:34<12:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2088/3295 [21:34<12:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2089/3295 [21:34<12:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2089/3295 [21:34<12:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2090/3295 [21:36<12:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2090/3295 [21:36<12:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2091/3295 [21:36<12:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2091/3295 [21:36<12:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2092/3295 [21:36<12:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  63%|██████▎   | 2092/3295 [21:36<12:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2093/3295 [21:38<12:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2093/3295 [21:38<12:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2094/3295 [21:38<12:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2094/3295 [21:38<12:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2095/3295 [21:38<12:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2095/3295 [21:38<12:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2096/3295 [21:40<12:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2096/3295 [21:40<12:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2097/3295 [21:40<12:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2097/3295 [21:40<12:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2098/3295 [21:40<12:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2098/3295 [21:40<12:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2099/3295 [21:42<12:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2099/3295 [21:42<12:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2100/3295 [21:42<12:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▎   | 2100/3295 [21:42<12:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2101/3295 [21:42<12:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2101/3295 [21:42<12:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2102/3295 [21:44<12:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2102/3295 [21:44<12:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2103/3295 [21:44<12:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2103/3295 [21:44<12:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2104/3295 [21:44<12:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2104/3295 [21:44<12:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2105/3295 [21:45<12:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2105/3295 [21:45<12:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2106/3295 [21:45<12:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2106/3295 [21:45<12:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2107/3295 [21:45<12:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2107/3295 [21:45<12:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2108/3295 [21:47<12:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2108/3295 [21:47<12:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2109/3295 [21:47<12:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2109/3295 [21:47<12:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2110/3295 [21:47<12:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2110/3295 [21:47<12:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2111/3295 [21:49<12:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2111/3295 [21:49<12:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2112/3295 [21:49<12:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2112/3295 [21:49<12:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2113/3295 [21:49<12:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2113/3295 [21:49<12:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2114/3295 [21:51<12:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2114/3295 [21:51<12:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2115/3295 [21:51<12:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2115/3295 [21:51<12:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2116/3295 [21:51<12:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2116/3295 [21:51<12:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2117/3295 [21:53<12:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2117/3295 [21:53<12:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2118/3295 [21:53<12:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2118/3295 [21:53<12:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2119/3295 [21:53<12:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2119/3295 [21:53<12:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2120/3295 [21:55<12:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2120/3295 [21:55<12:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2121/3295 [21:55<12:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2121/3295 [21:55<12:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2122/3295 [21:55<12:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2122/3295 [21:55<12:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2123/3295 [21:56<12:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2123/3295 [21:56<12:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2124/3295 [21:57<12:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2124/3295 [21:57<12:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2125/3295 [21:57<12:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  64%|██████▍   | 2125/3295 [21:57<12:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2126/3295 [21:58<12:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2126/3295 [21:58<12:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2127/3295 [21:58<12:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2127/3295 [21:58<12:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2128/3295 [21:58<12:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2128/3295 [21:58<12:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2129/3295 [22:00<12:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2129/3295 [22:00<12:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2130/3295 [22:00<12:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2130/3295 [22:00<12:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2131/3295 [22:00<12:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2131/3295 [22:00<12:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2132/3295 [22:02<12:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2132/3295 [22:02<12:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2133/3295 [22:02<12:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2133/3295 [22:02<12:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2134/3295 [22:02<11:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2134/3295 [22:02<11:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2135/3295 [22:04<11:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2135/3295 [22:04<11:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2136/3295 [22:04<11:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2136/3295 [22:04<11:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2137/3295 [22:04<11:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2137/3295 [22:04<11:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2138/3295 [22:06<11:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2138/3295 [22:06<11:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2139/3295 [22:06<11:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2139/3295 [22:06<11:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2140/3295 [22:06<11:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2140/3295 [22:06<11:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2141/3295 [22:07<11:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▍   | 2141/3295 [22:07<11:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2142/3295 [22:07<11:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2142/3295 [22:07<11:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2143/3295 [22:07<11:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2143/3295 [22:07<11:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2144/3295 [22:09<11:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2144/3295 [22:09<11:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2145/3295 [22:09<11:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2145/3295 [22:09<11:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2146/3295 [22:09<11:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2146/3295 [22:09<11:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2147/3295 [22:11<11:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2147/3295 [22:11<11:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2148/3295 [22:11<11:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2148/3295 [22:11<11:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2149/3295 [22:11<11:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2149/3295 [22:11<11:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2150/3295 [22:13<11:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2150/3295 [22:13<11:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2151/3295 [22:13<11:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2151/3295 [22:13<11:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2152/3295 [22:13<11:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2152/3295 [22:13<11:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2153/3295 [22:15<11:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2153/3295 [22:15<11:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2154/3295 [22:15<11:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2154/3295 [22:15<11:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2155/3295 [22:15<11:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2155/3295 [22:15<11:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2156/3295 [22:17<11:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2156/3295 [22:17<11:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2157/3295 [22:17<11:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2157/3295 [22:17<11:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2158/3295 [22:17<11:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  65%|██████▌   | 2158/3295 [22:17<11:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2159/3295 [22:19<11:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2159/3295 [22:19<11:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2160/3295 [22:19<11:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2160/3295 [22:19<11:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2161/3295 [22:19<11:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2161/3295 [22:19<11:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2162/3295 [22:20<11:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2162/3295 [22:20<11:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2163/3295 [22:20<11:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2163/3295 [22:20<11:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2164/3295 [22:20<11:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2164/3295 [22:20<11:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2165/3295 [22:22<11:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2165/3295 [22:22<11:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2166/3295 [22:22<11:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2166/3295 [22:22<11:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2167/3295 [22:22<11:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2167/3295 [22:22<11:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2168/3295 [22:24<11:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2168/3295 [22:24<11:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2169/3295 [22:24<11:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2169/3295 [22:24<11:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2170/3295 [22:24<11:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2170/3295 [22:24<11:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2171/3295 [22:26<11:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2171/3295 [22:26<11:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2172/3295 [22:26<11:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2172/3295 [22:26<11:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2173/3295 [22:26<11:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2173/3295 [22:26<11:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2174/3295 [22:28<11:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2174/3295 [22:28<11:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2175/3295 [22:28<11:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2175/3295 [22:28<11:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2176/3295 [22:28<11:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2176/3295 [22:28<11:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2177/3295 [22:30<11:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2177/3295 [22:30<11:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2178/3295 [22:30<11:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2178/3295 [22:30<11:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2179/3295 [22:30<11:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2179/3295 [22:30<11:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2180/3295 [22:31<11:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2180/3295 [22:31<11:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2181/3295 [22:32<11:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2181/3295 [22:32<11:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2182/3295 [22:32<11:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▌   | 2182/3295 [22:32<11:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2183/3295 [22:33<11:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2183/3295 [22:33<11:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2184/3295 [22:33<11:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2184/3295 [22:33<11:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2185/3295 [22:33<11:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2185/3295 [22:33<11:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2186/3295 [22:35<11:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2186/3295 [22:35<11:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2187/3295 [22:35<11:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2187/3295 [22:35<11:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2188/3295 [22:35<11:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2188/3295 [22:35<11:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2189/3295 [22:37<11:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2189/3295 [22:37<11:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2190/3295 [22:37<11:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2190/3295 [22:37<11:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2191/3295 [22:37<11:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  66%|██████▋   | 2191/3295 [22:37<11:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2192/3295 [22:39<11:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2192/3295 [22:39<11:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2193/3295 [22:39<11:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2193/3295 [22:39<11:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2194/3295 [22:39<11:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2194/3295 [22:39<11:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2195/3295 [22:41<11:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2195/3295 [22:41<11:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2196/3295 [22:41<11:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2196/3295 [22:41<11:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2197/3295 [22:41<11:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2197/3295 [22:41<11:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2198/3295 [22:42<11:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2198/3295 [22:42<11:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2199/3295 [22:42<11:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2199/3295 [22:42<11:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2200/3295 [22:42<11:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2200/3295 [22:42<11:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2201/3295 [22:44<11:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2201/3295 [22:44<11:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2202/3295 [22:44<11:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2202/3295 [22:44<11:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2203/3295 [22:44<11:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2203/3295 [22:44<11:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2204/3295 [22:46<11:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2204/3295 [22:46<11:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2205/3295 [22:46<11:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2205/3295 [22:46<11:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2206/3295 [22:46<11:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2206/3295 [22:46<11:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2207/3295 [22:48<11:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2207/3295 [22:48<11:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2208/3295 [22:48<11:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2208/3295 [22:48<11:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2209/3295 [22:48<11:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2209/3295 [22:48<11:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2210/3295 [22:50<11:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2210/3295 [22:50<11:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2211/3295 [22:50<11:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2211/3295 [22:50<11:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2212/3295 [22:50<11:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2212/3295 [22:50<11:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2213/3295 [22:52<11:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2213/3295 [22:52<11:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2214/3295 [22:52<11:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2214/3295 [22:52<11:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2215/3295 [22:52<11:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2215/3295 [22:52<11:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2216/3295 [22:54<11:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2216/3295 [22:54<11:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2217/3295 [22:54<11:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2217/3295 [22:54<11:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2218/3295 [22:54<11:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2218/3295 [22:54<11:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2219/3295 [22:55<11:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2219/3295 [22:55<11:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2220/3295 [22:55<11:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2220/3295 [22:55<11:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2221/3295 [22:56<11:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2221/3295 [22:56<11:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2222/3295 [22:58<11:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2222/3295 [22:58<11:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2223/3295 [22:58<11:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2223/3295 [22:58<11:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2224/3295 [22:58<11:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  67%|██████▋   | 2224/3295 [22:58<11:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2225/3295 [22:59<11:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2225/3295 [22:59<11:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2226/3295 [22:59<11:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2226/3295 [22:59<11:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2227/3295 [22:59<11:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2227/3295 [22:59<11:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2228/3295 [23:01<11:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2228/3295 [23:01<11:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2229/3295 [23:01<11:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2229/3295 [23:01<11:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2230/3295 [23:01<10:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2230/3295 [23:01<10:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2231/3295 [23:03<10:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2231/3295 [23:03<10:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2232/3295 [23:03<10:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2232/3295 [23:03<10:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2233/3295 [23:03<10:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2233/3295 [23:03<10:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2234/3295 [23:05<10:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2234/3295 [23:05<10:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2235/3295 [23:05<10:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2235/3295 [23:05<10:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2236/3295 [23:05<10:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2236/3295 [23:05<10:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2237/3295 [23:07<10:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2237/3295 [23:07<10:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2238/3295 [23:07<10:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2238/3295 [23:07<10:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2239/3295 [23:07<10:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2239/3295 [23:07<10:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2240/3295 [23:09<10:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2240/3295 [23:09<10:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2241/3295 [23:09<10:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2241/3295 [23:09<10:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2242/3295 [23:09<10:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2242/3295 [23:09<10:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2243/3295 [23:11<10:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2243/3295 [23:11<10:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2244/3295 [23:11<10:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2244/3295 [23:11<10:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2245/3295 [23:11<10:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2245/3295 [23:11<10:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2246/3295 [23:13<10:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2246/3295 [23:13<10:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2247/3295 [23:13<10:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2247/3295 [23:13<10:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2248/3295 [23:13<10:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2248/3295 [23:13<10:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2249/3295 [23:15<10:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2249/3295 [23:15<10:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2250/3295 [23:15<10:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2250/3295 [23:15<10:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2251/3295 [23:15<10:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2251/3295 [23:15<10:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2252/3295 [23:16<10:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2252/3295 [23:16<10:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2253/3295 [23:16<10:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2253/3295 [23:16<10:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2254/3295 [23:16<10:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2254/3295 [23:16<10:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2255/3295 [23:18<10:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2255/3295 [23:18<10:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2256/3295 [23:18<10:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2256/3295 [23:18<10:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2257/3295 [23:18<10:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  68%|██████▊   | 2257/3295 [23:18<10:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2258/3295 [23:20<10:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2258/3295 [23:20<10:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2259/3295 [23:20<10:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2259/3295 [23:20<10:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2260/3295 [23:20<10:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2260/3295 [23:20<10:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2261/3295 [23:22<10:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2261/3295 [23:22<10:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2262/3295 [23:22<10:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2262/3295 [23:22<10:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2263/3295 [23:22<10:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2263/3295 [23:22<10:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2264/3295 [23:24<10:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2264/3295 [23:24<10:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2265/3295 [23:24<10:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▊   | 2265/3295 [23:24<10:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2266/3295 [23:24<10:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2266/3295 [23:24<10:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2267/3295 [23:25<10:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2267/3295 [23:25<10:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2268/3295 [23:26<10:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2268/3295 [23:26<10:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2269/3295 [23:26<10:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2269/3295 [23:26<10:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2270/3295 [23:27<10:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2270/3295 [23:27<10:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2271/3295 [23:27<10:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2271/3295 [23:27<10:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2272/3295 [23:27<10:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2272/3295 [23:27<10:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2273/3295 [23:29<10:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2273/3295 [23:29<10:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2274/3295 [23:29<10:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2274/3295 [23:29<10:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2275/3295 [23:29<10:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2275/3295 [23:29<10:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2276/3295 [23:31<10:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2276/3295 [23:31<10:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2277/3295 [23:31<10:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2277/3295 [23:31<10:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2278/3295 [23:31<10:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2278/3295 [23:31<10:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2279/3295 [23:33<10:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2279/3295 [23:33<10:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2280/3295 [23:33<10:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2280/3295 [23:33<10:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2281/3295 [23:33<10:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2281/3295 [23:33<10:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2282/3295 [23:35<10:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2282/3295 [23:35<10:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2283/3295 [23:35<10:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2283/3295 [23:35<10:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2284/3295 [23:35<10:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2284/3295 [23:35<10:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2285/3295 [23:37<10:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2285/3295 [23:37<10:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2286/3295 [23:37<10:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2286/3295 [23:37<10:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2287/3295 [23:37<10:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2287/3295 [23:37<10:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2288/3295 [23:38<10:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2288/3295 [23:38<10:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2289/3295 [23:38<10:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2289/3295 [23:38<10:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2290/3295 [23:39<10:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  69%|██████▉   | 2290/3295 [23:39<10:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2291/3295 [23:40<10:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2291/3295 [23:40<10:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2292/3295 [23:40<10:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2292/3295 [23:40<10:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2293/3295 [23:40<10:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2293/3295 [23:40<10:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2294/3295 [23:42<10:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2294/3295 [23:42<10:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2295/3295 [23:42<10:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2295/3295 [23:42<10:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2296/3295 [23:42<10:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2296/3295 [23:42<10:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2297/3295 [23:44<10:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2297/3295 [23:44<10:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2298/3295 [23:44<10:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2298/3295 [23:44<10:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2299/3295 [23:44<10:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2299/3295 [23:44<10:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2300/3295 [23:46<10:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2300/3295 [23:46<10:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2301/3295 [23:46<10:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2301/3295 [23:46<10:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2302/3295 [23:46<10:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2302/3295 [23:46<10:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2303/3295 [23:48<10:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2303/3295 [23:48<10:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2304/3295 [23:48<10:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2304/3295 [23:48<10:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2305/3295 [23:48<10:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2305/3295 [23:48<10:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2306/3295 [23:50<10:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|██████▉   | 2306/3295 [23:50<10:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2307/3295 [23:50<10:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2307/3295 [23:50<10:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2308/3295 [23:50<10:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2308/3295 [23:50<10:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2309/3295 [23:52<10:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2309/3295 [23:52<10:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2310/3295 [23:52<10:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2310/3295 [23:52<10:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2311/3295 [23:52<10:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2311/3295 [23:52<10:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2312/3295 [23:53<10:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2312/3295 [23:53<10:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2313/3295 [23:54<10:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2313/3295 [23:54<10:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2314/3295 [23:54<10:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2314/3295 [23:54<10:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2315/3295 [23:56<10:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2315/3295 [23:56<10:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2316/3295 [23:56<10:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2316/3295 [23:56<10:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2317/3295 [23:56<10:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2317/3295 [23:56<10:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2318/3295 [23:57<10:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2318/3295 [23:57<10:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2319/3295 [23:57<10:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2319/3295 [23:57<10:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2320/3295 [23:57<10:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2320/3295 [23:57<10:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2321/3295 [23:59<10:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2321/3295 [23:59<10:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2322/3295 [23:59<10:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  70%|███████   | 2322/3295 [23:59<10:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2323/3295 [23:59<10:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2323/3295 [23:59<10:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2324/3295 [24:01<10:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2324/3295 [24:01<10:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2325/3295 [24:01<10:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2325/3295 [24:01<10:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2326/3295 [24:01<10:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2326/3295 [24:01<10:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2327/3295 [24:03<10:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2327/3295 [24:03<10:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2328/3295 [24:03<09:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2328/3295 [24:03<09:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2329/3295 [24:03<09:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2329/3295 [24:03<09:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2330/3295 [24:05<09:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2330/3295 [24:05<09:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2331/3295 [24:05<09:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2331/3295 [24:05<09:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2332/3295 [24:05<09:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2332/3295 [24:05<09:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2333/3295 [24:06<09:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2333/3295 [24:06<09:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2334/3295 [24:06<09:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2334/3295 [24:06<09:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2335/3295 [24:06<09:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2335/3295 [24:06<09:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2336/3295 [24:08<09:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2336/3295 [24:08<09:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2337/3295 [24:08<09:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2337/3295 [24:08<09:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2338/3295 [24:08<09:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2338/3295 [24:08<09:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2339/3295 [24:10<09:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2339/3295 [24:10<09:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2340/3295 [24:10<09:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2340/3295 [24:10<09:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2341/3295 [24:10<09:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2341/3295 [24:10<09:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2342/3295 [24:12<09:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2342/3295 [24:12<09:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2343/3295 [24:12<09:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2343/3295 [24:12<09:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2344/3295 [24:12<09:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2344/3295 [24:12<09:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2345/3295 [24:14<09:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2345/3295 [24:14<09:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2346/3295 [24:14<09:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2346/3295 [24:14<09:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2347/3295 [24:14<09:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████   | 2347/3295 [24:14<09:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2348/3295 [24:16<09:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2348/3295 [24:16<09:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2349/3295 [24:16<09:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2349/3295 [24:16<09:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2350/3295 [24:16<09:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2350/3295 [24:16<09:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2351/3295 [24:18<09:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2351/3295 [24:18<09:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2352/3295 [24:18<09:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2352/3295 [24:18<09:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2353/3295 [24:18<09:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2353/3295 [24:18<09:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2354/3295 [24:20<09:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2354/3295 [24:20<09:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2355/3295 [24:20<09:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  71%|███████▏  | 2355/3295 [24:20<09:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2356/3295 [24:20<09:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2356/3295 [24:20<09:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2357/3295 [24:21<09:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2357/3295 [24:21<09:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2358/3295 [24:22<09:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2358/3295 [24:22<09:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2359/3295 [24:22<09:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2359/3295 [24:22<09:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2360/3295 [24:23<09:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2360/3295 [24:23<09:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2361/3295 [24:23<09:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2361/3295 [24:23<09:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2362/3295 [24:23<09:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2362/3295 [24:23<09:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2363/3295 [24:25<09:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2363/3295 [24:25<09:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2364/3295 [24:25<09:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2364/3295 [24:25<09:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2365/3295 [24:25<09:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2365/3295 [24:25<09:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2366/3295 [24:27<09:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2366/3295 [24:27<09:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2367/3295 [24:27<09:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2367/3295 [24:27<09:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2368/3295 [24:27<09:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2368/3295 [24:27<09:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2369/3295 [24:29<09:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2369/3295 [24:29<09:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2370/3295 [24:29<09:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2370/3295 [24:29<09:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2371/3295 [24:29<09:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2371/3295 [24:29<09:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2372/3295 [24:31<09:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2372/3295 [24:31<09:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2373/3295 [24:31<09:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2373/3295 [24:31<09:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2374/3295 [24:31<09:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2374/3295 [24:31<09:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2375/3295 [24:33<09:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2375/3295 [24:33<09:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2376/3295 [24:33<09:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2376/3295 [24:33<09:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2377/3295 [24:33<09:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2377/3295 [24:33<09:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2378/3295 [24:35<09:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2378/3295 [24:35<09:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2379/3295 [24:35<09:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2379/3295 [24:35<09:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2380/3295 [24:35<09:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2380/3295 [24:35<09:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2381/3295 [24:37<09:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2381/3295 [24:37<09:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2382/3295 [24:37<09:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2382/3295 [24:37<09:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2383/3295 [24:37<09:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2383/3295 [24:37<09:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2384/3295 [24:39<09:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2384/3295 [24:39<09:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2385/3295 [24:39<09:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2385/3295 [24:39<09:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2386/3295 [24:39<09:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2386/3295 [24:39<09:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2387/3295 [24:40<09:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2387/3295 [24:40<09:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2388/3295 [24:40<09:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  72%|███████▏  | 2388/3295 [24:40<09:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2389/3295 [24:40<09:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2389/3295 [24:40<09:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2390/3295 [24:42<09:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2390/3295 [24:42<09:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2391/3295 [24:42<09:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2391/3295 [24:42<09:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2392/3295 [24:42<09:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2392/3295 [24:42<09:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2393/3295 [24:44<09:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2393/3295 [24:44<09:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2394/3295 [24:44<09:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2394/3295 [24:44<09:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2395/3295 [24:44<09:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2395/3295 [24:44<09:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2396/3295 [24:46<09:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2396/3295 [24:46<09:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2397/3295 [24:46<09:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2397/3295 [24:46<09:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2398/3295 [24:46<09:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2398/3295 [24:46<09:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2399/3295 [24:48<09:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2399/3295 [24:48<09:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2400/3295 [24:48<09:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2400/3295 [24:48<09:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2401/3295 [24:48<09:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2401/3295 [24:48<09:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2402/3295 [24:50<09:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2402/3295 [24:50<09:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2403/3295 [24:50<09:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2403/3295 [24:50<09:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2404/3295 [24:50<09:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2404/3295 [24:50<09:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2405/3295 [24:52<09:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2405/3295 [24:52<09:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2406/3295 [24:52<09:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2406/3295 [24:52<09:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2407/3295 [24:52<09:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2407/3295 [24:52<09:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2408/3295 [24:54<09:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2408/3295 [24:54<09:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2409/3295 [24:54<09:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2409/3295 [24:54<09:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2410/3295 [24:54<09:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2410/3295 [24:54<09:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2411/3295 [24:55<09:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2411/3295 [24:55<09:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2412/3295 [24:55<09:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2412/3295 [24:55<09:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2413/3295 [24:55<09:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2413/3295 [24:55<09:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2414/3295 [24:57<09:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2414/3295 [24:57<09:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2415/3295 [24:57<09:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2415/3295 [24:57<09:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2416/3295 [24:57<09:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2416/3295 [24:57<09:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2417/3295 [24:59<09:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2417/3295 [24:59<09:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2418/3295 [24:59<09:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2418/3295 [24:59<09:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2419/3295 [24:59<09:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2419/3295 [24:59<09:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2420/3295 [25:01<09:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2420/3295 [25:01<09:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2421/3295 [25:01<09:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  73%|███████▎  | 2421/3295 [25:01<09:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2422/3295 [25:01<09:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2422/3295 [25:01<09:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2423/3295 [25:03<09:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2423/3295 [25:03<09:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2424/3295 [25:03<09:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2424/3295 [25:03<09:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2425/3295 [25:03<08:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2425/3295 [25:03<08:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2426/3295 [25:05<08:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2426/3295 [25:05<08:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2427/3295 [25:05<08:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2427/3295 [25:05<08:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2428/3295 [25:05<08:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2428/3295 [25:05<08:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2429/3295 [25:06<08:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2429/3295 [25:06<08:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2430/3295 [25:07<08:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▎  | 2430/3295 [25:07<08:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2431/3295 [25:07<08:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2431/3295 [25:07<08:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2432/3295 [25:08<08:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2432/3295 [25:08<08:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2433/3295 [25:08<08:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2433/3295 [25:08<08:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2434/3295 [25:08<08:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2434/3295 [25:08<08:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2435/3295 [25:10<08:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2435/3295 [25:10<08:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2436/3295 [25:10<08:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2436/3295 [25:10<08:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2437/3295 [25:10<08:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2437/3295 [25:10<08:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2438/3295 [25:12<08:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2438/3295 [25:12<08:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2439/3295 [25:12<08:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2439/3295 [25:12<08:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2440/3295 [25:12<08:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2440/3295 [25:12<08:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2441/3295 [25:14<08:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2441/3295 [25:14<08:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2442/3295 [25:14<08:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2442/3295 [25:14<08:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2443/3295 [25:14<08:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2443/3295 [25:14<08:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2444/3295 [25:16<08:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2444/3295 [25:16<08:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2445/3295 [25:16<08:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2445/3295 [25:16<08:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2446/3295 [25:16<08:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2446/3295 [25:16<08:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2447/3295 [25:18<08:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2447/3295 [25:18<08:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2448/3295 [25:18<08:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2448/3295 [25:18<08:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2449/3295 [25:18<08:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2449/3295 [25:18<08:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2450/3295 [25:20<08:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2450/3295 [25:20<08:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2451/3295 [25:20<08:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2451/3295 [25:20<08:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2452/3295 [25:20<08:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2452/3295 [25:20<08:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2453/3295 [25:22<08:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2453/3295 [25:22<08:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2454/3295 [25:22<08:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  74%|███████▍  | 2454/3295 [25:22<08:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2455/3295 [25:22<08:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2455/3295 [25:22<08:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2456/3295 [25:23<08:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2456/3295 [25:23<08:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2457/3295 [25:24<08:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2457/3295 [25:24<08:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2458/3295 [25:24<08:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2458/3295 [25:24<08:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2459/3295 [25:25<08:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2459/3295 [25:25<08:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2460/3295 [25:25<08:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2460/3295 [25:25<08:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2461/3295 [25:25<08:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2461/3295 [25:25<08:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2462/3295 [25:27<08:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2462/3295 [25:27<08:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2463/3295 [25:27<08:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2463/3295 [25:27<08:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2464/3295 [25:27<08:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2464/3295 [25:27<08:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2465/3295 [25:29<08:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2465/3295 [25:29<08:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2466/3295 [25:29<08:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2466/3295 [25:29<08:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2467/3295 [25:29<08:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2467/3295 [25:29<08:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2468/3295 [25:31<08:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2468/3295 [25:31<08:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2469/3295 [25:31<08:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2469/3295 [25:31<08:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2470/3295 [25:31<08:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2470/3295 [25:31<08:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2471/3295 [25:33<08:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▍  | 2471/3295 [25:33<08:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2472/3295 [25:33<08:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2472/3295 [25:33<08:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2473/3295 [25:33<08:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2473/3295 [25:33<08:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2474/3295 [25:35<08:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2474/3295 [25:35<08:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2475/3295 [25:35<08:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2475/3295 [25:35<08:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2476/3295 [25:35<08:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2476/3295 [25:35<08:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2477/3295 [25:36<08:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2477/3295 [25:36<08:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2478/3295 [25:36<08:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2478/3295 [25:36<08:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2479/3295 [25:36<08:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2479/3295 [25:36<08:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2480/3295 [25:38<08:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2480/3295 [25:38<08:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2481/3295 [25:38<08:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2481/3295 [25:38<08:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2482/3295 [25:38<08:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2482/3295 [25:38<08:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2483/3295 [25:40<08:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2483/3295 [25:40<08:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2484/3295 [25:40<08:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2484/3295 [25:40<08:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2485/3295 [25:40<08:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2485/3295 [25:40<08:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2486/3295 [25:42<08:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2486/3295 [25:42<08:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2487/3295 [25:42<08:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  75%|███████▌  | 2487/3295 [25:42<08:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2488/3295 [25:42<08:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2488/3295 [25:42<08:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2489/3295 [25:44<08:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2489/3295 [25:44<08:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2490/3295 [25:44<08:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2490/3295 [25:44<08:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2491/3295 [25:44<08:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2491/3295 [25:44<08:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2492/3295 [25:46<08:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2492/3295 [25:46<08:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2493/3295 [25:46<08:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2493/3295 [25:46<08:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2494/3295 [25:46<08:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2494/3295 [25:46<08:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2495/3295 [25:47<08:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2495/3295 [25:47<08:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2496/3295 [25:47<08:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2496/3295 [25:47<08:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2497/3295 [25:47<08:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2497/3295 [25:47<08:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2498/3295 [25:49<08:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2498/3295 [25:49<08:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2499/3295 [25:49<08:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2499/3295 [25:49<08:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2500/3295 [25:49<08:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2500/3295 [25:49<08:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2501/3295 [25:51<08:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2501/3295 [25:51<08:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2502/3295 [25:51<08:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2502/3295 [25:51<08:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2503/3295 [25:51<08:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2503/3295 [25:51<08:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2504/3295 [25:53<08:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2504/3295 [25:53<08:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2505/3295 [25:53<08:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2505/3295 [25:53<08:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2506/3295 [25:53<08:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2506/3295 [25:53<08:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2507/3295 [25:55<08:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2507/3295 [25:55<08:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2508/3295 [25:55<08:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2508/3295 [25:55<08:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2509/3295 [25:55<08:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2509/3295 [25:55<08:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2510/3295 [25:57<08:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2510/3295 [25:57<08:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2511/3295 [25:57<08:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2511/3295 [25:57<08:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2512/3295 [25:57<08:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▌  | 2512/3295 [25:57<08:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2513/3295 [25:59<08:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2513/3295 [25:59<08:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2514/3295 [25:59<08:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2514/3295 [25:59<08:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2515/3295 [25:59<08:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2515/3295 [25:59<08:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2516/3295 [26:01<08:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2516/3295 [26:01<08:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2517/3295 [26:01<08:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2517/3295 [26:01<08:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2518/3295 [26:01<08:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2518/3295 [26:01<08:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2519/3295 [26:02<08:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2519/3295 [26:02<08:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2520/3295 [26:03<08:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  76%|███████▋  | 2520/3295 [26:03<08:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2521/3295 [26:03<07:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2521/3295 [26:03<07:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2522/3295 [26:04<07:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2522/3295 [26:04<07:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2523/3295 [26:04<07:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2523/3295 [26:04<07:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2524/3295 [26:04<07:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2524/3295 [26:04<07:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2525/3295 [26:06<07:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2525/3295 [26:06<07:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2526/3295 [26:06<07:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2526/3295 [26:06<07:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2527/3295 [26:06<07:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2527/3295 [26:06<07:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2528/3295 [26:08<07:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2528/3295 [26:08<07:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2529/3295 [26:08<07:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2529/3295 [26:08<07:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2530/3295 [26:08<07:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2530/3295 [26:08<07:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2531/3295 [26:10<07:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2531/3295 [26:10<07:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2532/3295 [26:10<07:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2532/3295 [26:10<07:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2533/3295 [26:10<07:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2533/3295 [26:10<07:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2534/3295 [26:12<07:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2534/3295 [26:12<07:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2535/3295 [26:12<07:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2535/3295 [26:12<07:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2536/3295 [26:12<07:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2536/3295 [26:12<07:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2537/3295 [26:13<07:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2537/3295 [26:13<07:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2538/3295 [26:14<07:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2538/3295 [26:14<07:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2539/3295 [26:14<07:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2539/3295 [26:14<07:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2540/3295 [26:15<07:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2540/3295 [26:15<07:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2541/3295 [26:15<07:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2541/3295 [26:15<07:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2542/3295 [26:15<07:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2542/3295 [26:15<07:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2543/3295 [26:17<07:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2543/3295 [26:17<07:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2544/3295 [26:17<07:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2544/3295 [26:17<07:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2545/3295 [26:17<07:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2545/3295 [26:17<07:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2546/3295 [26:19<07:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2546/3295 [26:19<07:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2547/3295 [26:19<07:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2547/3295 [26:19<07:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2548/3295 [26:19<07:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2548/3295 [26:19<07:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2549/3295 [26:21<07:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2549/3295 [26:21<07:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2550/3295 [26:21<07:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2550/3295 [26:21<07:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2551/3295 [26:21<07:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2551/3295 [26:21<07:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2552/3295 [26:23<07:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2552/3295 [26:23<07:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2553/3295 [26:23<07:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  77%|███████▋  | 2553/3295 [26:23<07:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2554/3295 [26:23<07:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2554/3295 [26:23<07:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2555/3295 [26:25<07:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2555/3295 [26:25<07:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2556/3295 [26:25<07:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2556/3295 [26:25<07:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2557/3295 [26:25<07:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2557/3295 [26:25<07:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2558/3295 [26:27<07:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2558/3295 [26:27<07:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2559/3295 [26:27<07:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2559/3295 [26:27<07:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2560/3295 [26:27<07:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2560/3295 [26:27<07:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2561/3295 [26:28<07:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2561/3295 [26:28<07:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2562/3295 [26:28<07:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2562/3295 [26:28<07:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2563/3295 [26:28<07:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2563/3295 [26:28<07:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2564/3295 [26:30<07:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2564/3295 [26:30<07:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2565/3295 [26:30<07:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2565/3295 [26:30<07:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2566/3295 [26:30<07:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2566/3295 [26:30<07:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2567/3295 [26:32<07:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2567/3295 [26:32<07:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2568/3295 [26:32<07:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2568/3295 [26:32<07:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2569/3295 [26:32<07:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2569/3295 [26:32<07:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2570/3295 [26:34<07:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2570/3295 [26:34<07:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2571/3295 [26:34<07:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2571/3295 [26:34<07:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2572/3295 [26:34<07:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2572/3295 [26:34<07:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2573/3295 [26:36<07:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2573/3295 [26:36<07:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2574/3295 [26:36<07:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2574/3295 [26:36<07:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2575/3295 [26:36<07:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2575/3295 [26:36<07:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2576/3295 [26:38<07:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2576/3295 [26:38<07:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2577/3295 [26:38<07:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2577/3295 [26:38<07:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2578/3295 [26:38<07:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2578/3295 [26:38<07:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2579/3295 [26:40<07:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2579/3295 [26:40<07:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2580/3295 [26:40<07:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2580/3295 [26:40<07:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2581/3295 [26:40<07:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2581/3295 [26:40<07:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2582/3295 [26:42<07:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2582/3295 [26:42<07:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2583/3295 [26:42<07:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2583/3295 [26:42<07:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2584/3295 [26:42<07:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2584/3295 [26:42<07:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2585/3295 [26:44<07:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2585/3295 [26:44<07:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2586/3295 [26:44<07:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  78%|███████▊  | 2586/3295 [26:44<07:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2587/3295 [26:44<07:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2587/3295 [26:44<07:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2588/3295 [26:45<07:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2588/3295 [26:45<07:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2589/3295 [26:45<07:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2589/3295 [26:45<07:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2590/3295 [26:46<07:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2590/3295 [26:46<07:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2591/3295 [26:47<07:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2591/3295 [26:47<07:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2592/3295 [26:47<07:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2592/3295 [26:47<07:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2593/3295 [26:47<07:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2593/3295 [26:47<07:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2594/3295 [26:49<07:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▊  | 2594/3295 [26:49<07:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2595/3295 [26:49<07:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2595/3295 [26:49<07:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2596/3295 [26:49<07:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2596/3295 [26:49<07:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2597/3295 [26:51<07:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2597/3295 [26:51<07:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2598/3295 [26:51<07:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2598/3295 [26:51<07:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2599/3295 [26:51<07:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2599/3295 [26:51<07:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2600/3295 [26:53<07:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2600/3295 [26:53<07:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2601/3295 [26:53<07:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2601/3295 [26:53<07:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2602/3295 [26:53<07:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2602/3295 [26:53<07:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2603/3295 [26:55<07:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2603/3295 [26:55<07:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2604/3295 [26:55<07:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2604/3295 [26:55<07:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2605/3295 [26:55<07:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2605/3295 [26:55<07:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2606/3295 [26:57<07:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2606/3295 [26:57<07:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2607/3295 [26:57<07:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2607/3295 [26:57<07:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2608/3295 [26:57<07:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2608/3295 [26:57<07:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2609/3295 [26:58<07:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2609/3295 [26:58<07:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2610/3295 [26:58<07:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2610/3295 [26:58<07:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2611/3295 [26:58<07:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2611/3295 [26:58<07:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2612/3295 [27:00<07:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2612/3295 [27:00<07:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2613/3295 [27:00<07:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2613/3295 [27:00<07:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2614/3295 [27:01<07:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2614/3295 [27:01<07:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2615/3295 [27:03<07:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2615/3295 [27:03<07:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2616/3295 [27:03<07:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2616/3295 [27:03<07:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2617/3295 [27:03<07:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2617/3295 [27:03<07:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2618/3295 [27:04<07:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2618/3295 [27:04<07:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2619/3295 [27:04<06:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  79%|███████▉  | 2619/3295 [27:04<06:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2620/3295 [27:04<06:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2620/3295 [27:04<06:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2621/3295 [27:06<06:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2621/3295 [27:06<06:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2622/3295 [27:06<06:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2622/3295 [27:06<06:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2623/3295 [27:06<06:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2623/3295 [27:06<06:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2624/3295 [27:08<06:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2624/3295 [27:08<06:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2625/3295 [27:08<06:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2625/3295 [27:08<06:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2626/3295 [27:08<06:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2626/3295 [27:08<06:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2627/3295 [27:10<06:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2627/3295 [27:10<06:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2628/3295 [27:10<06:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2628/3295 [27:10<06:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2629/3295 [27:10<06:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2629/3295 [27:10<06:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2630/3295 [27:12<06:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2630/3295 [27:12<06:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2631/3295 [27:12<06:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2631/3295 [27:12<06:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2632/3295 [27:12<06:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2632/3295 [27:12<06:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2633/3295 [27:14<06:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2633/3295 [27:14<06:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2634/3295 [27:14<06:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2634/3295 [27:14<06:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2635/3295 [27:14<06:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|███████▉  | 2635/3295 [27:14<06:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2636/3295 [27:16<06:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2636/3295 [27:16<06:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2637/3295 [27:16<06:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2637/3295 [27:16<06:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2638/3295 [27:16<06:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2638/3295 [27:16<06:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2639/3295 [27:18<06:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2639/3295 [27:18<06:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2640/3295 [27:18<06:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2640/3295 [27:18<06:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2641/3295 [27:18<06:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2641/3295 [27:18<06:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2642/3295 [27:20<06:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2642/3295 [27:20<06:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2643/3295 [27:20<06:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2643/3295 [27:20<06:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2644/3295 [27:20<06:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2644/3295 [27:20<06:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2645/3295 [27:22<06:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2645/3295 [27:22<06:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2646/3295 [27:22<06:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2646/3295 [27:22<06:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2647/3295 [27:22<06:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2647/3295 [27:22<06:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2648/3295 [27:24<06:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2648/3295 [27:24<06:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2649/3295 [27:24<06:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2649/3295 [27:24<06:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2650/3295 [27:24<06:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2650/3295 [27:24<06:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2651/3295 [27:26<06:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2651/3295 [27:26<06:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2652/3295 [27:26<06:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  80%|████████  | 2652/3295 [27:26<06:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2653/3295 [27:26<06:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2653/3295 [27:26<06:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2654/3295 [27:27<06:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2654/3295 [27:27<06:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2655/3295 [27:27<06:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2655/3295 [27:27<06:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2656/3295 [27:27<06:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2656/3295 [27:27<06:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2657/3295 [27:29<06:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2657/3295 [27:29<06:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2658/3295 [27:29<06:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2658/3295 [27:29<06:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2659/3295 [27:29<06:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2659/3295 [27:29<06:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2660/3295 [27:31<06:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2660/3295 [27:31<06:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2661/3295 [27:31<06:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2661/3295 [27:31<06:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2662/3295 [27:31<06:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2662/3295 [27:31<06:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2663/3295 [27:33<06:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2663/3295 [27:33<06:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2664/3295 [27:33<06:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2664/3295 [27:33<06:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2665/3295 [27:33<06:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2665/3295 [27:33<06:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2666/3295 [27:35<06:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2666/3295 [27:35<06:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2667/3295 [27:35<06:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2667/3295 [27:35<06:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2668/3295 [27:35<06:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2668/3295 [27:35<06:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2669/3295 [27:37<06:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2669/3295 [27:37<06:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2670/3295 [27:37<06:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2670/3295 [27:37<06:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2671/3295 [27:37<06:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2671/3295 [27:37<06:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2672/3295 [27:39<06:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2672/3295 [27:39<06:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2673/3295 [27:39<06:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2673/3295 [27:39<06:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2674/3295 [27:39<06:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2674/3295 [27:39<06:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2675/3295 [27:40<06:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2675/3295 [27:40<06:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2676/3295 [27:40<06:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2676/3295 [27:40<06:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2677/3295 [27:40<06:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████  | 2677/3295 [27:40<06:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2678/3295 [27:42<06:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2678/3295 [27:42<06:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2679/3295 [27:42<06:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2679/3295 [27:42<06:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2680/3295 [27:42<06:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2680/3295 [27:42<06:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2681/3295 [27:44<06:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2681/3295 [27:44<06:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2682/3295 [27:44<06:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2682/3295 [27:44<06:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2683/3295 [27:44<06:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2683/3295 [27:44<06:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2684/3295 [27:46<06:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2684/3295 [27:46<06:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2685/3295 [27:46<06:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  81%|████████▏ | 2685/3295 [27:46<06:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2686/3295 [27:46<06:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2686/3295 [27:46<06:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2687/3295 [27:48<06:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2687/3295 [27:48<06:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2688/3295 [27:48<06:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2688/3295 [27:48<06:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2689/3295 [27:48<06:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2689/3295 [27:48<06:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2690/3295 [27:49<06:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2690/3295 [27:49<06:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2691/3295 [27:49<06:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2691/3295 [27:49<06:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2692/3295 [27:49<06:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2692/3295 [27:49<06:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2693/3295 [27:51<06:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2693/3295 [27:51<06:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2694/3295 [27:51<06:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2694/3295 [27:51<06:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2695/3295 [27:51<06:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2695/3295 [27:51<06:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2696/3295 [27:53<06:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2696/3295 [27:53<06:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2697/3295 [27:53<06:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2697/3295 [27:53<06:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2698/3295 [27:53<06:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2698/3295 [27:53<06:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2699/3295 [27:55<06:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2699/3295 [27:55<06:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2700/3295 [27:55<06:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2700/3295 [27:55<06:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2701/3295 [27:55<06:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2701/3295 [27:55<06:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2702/3295 [27:57<06:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2702/3295 [27:57<06:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2703/3295 [27:57<06:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2703/3295 [27:57<06:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2704/3295 [27:57<06:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2704/3295 [27:57<06:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2705/3295 [27:59<06:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2705/3295 [27:59<06:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2706/3295 [27:59<06:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2706/3295 [27:59<06:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2707/3295 [27:59<06:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2707/3295 [27:59<06:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2708/3295 [28:01<06:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2708/3295 [28:01<06:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2709/3295 [28:01<06:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2709/3295 [28:01<06:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2710/3295 [28:01<06:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2710/3295 [28:01<06:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2711/3295 [28:02<06:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2711/3295 [28:02<06:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2712/3295 [28:02<06:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2712/3295 [28:02<06:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2713/3295 [28:03<06:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2713/3295 [28:03<06:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2714/3295 [28:04<06:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2714/3295 [28:04<06:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2715/3295 [28:04<05:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2715/3295 [28:04<05:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2716/3295 [28:04<05:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2716/3295 [28:04<05:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2717/3295 [28:06<05:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2717/3295 [28:06<05:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2718/3295 [28:06<05:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  82%|████████▏ | 2718/3295 [28:06<05:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2719/3295 [28:06<05:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2719/3295 [28:06<05:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2720/3295 [28:08<05:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2720/3295 [28:08<05:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2721/3295 [28:08<05:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2721/3295 [28:08<05:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2722/3295 [28:08<05:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2722/3295 [28:08<05:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2723/3295 [28:10<05:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2723/3295 [28:10<05:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2724/3295 [28:10<05:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2724/3295 [28:10<05:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2725/3295 [28:10<05:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2725/3295 [28:10<05:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2726/3295 [28:12<05:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2726/3295 [28:12<05:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2727/3295 [28:12<05:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2727/3295 [28:12<05:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2728/3295 [28:12<05:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2728/3295 [28:12<05:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2729/3295 [28:14<05:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2729/3295 [28:14<05:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2730/3295 [28:14<05:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2730/3295 [28:14<05:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2731/3295 [28:14<05:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2731/3295 [28:14<05:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2732/3295 [28:15<05:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2732/3295 [28:15<05:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2733/3295 [28:15<05:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2733/3295 [28:15<05:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2734/3295 [28:15<05:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2734/3295 [28:15<05:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2735/3295 [28:17<05:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2735/3295 [28:17<05:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2736/3295 [28:17<05:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2736/3295 [28:17<05:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2737/3295 [28:17<05:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2737/3295 [28:17<05:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2738/3295 [28:19<05:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2738/3295 [28:19<05:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2739/3295 [28:19<05:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2739/3295 [28:19<05:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2740/3295 [28:19<05:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2740/3295 [28:19<05:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2741/3295 [28:21<05:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2741/3295 [28:21<05:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2742/3295 [28:21<05:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2742/3295 [28:21<05:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2743/3295 [28:21<05:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2743/3295 [28:21<05:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2744/3295 [28:23<05:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2744/3295 [28:23<05:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2745/3295 [28:23<05:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2745/3295 [28:23<05:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2746/3295 [28:23<05:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2746/3295 [28:23<05:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2747/3295 [28:25<05:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2747/3295 [28:25<05:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2748/3295 [28:25<05:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2748/3295 [28:25<05:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2749/3295 [28:25<05:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2749/3295 [28:25<05:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2750/3295 [28:26<05:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2750/3295 [28:26<05:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2751/3295 [28:27<05:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  83%|████████▎ | 2751/3295 [28:27<05:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2752/3295 [28:27<05:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2752/3295 [28:27<05:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2753/3295 [28:28<05:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2753/3295 [28:28<05:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2754/3295 [28:28<05:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2754/3295 [28:28<05:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2755/3295 [28:28<05:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2755/3295 [28:28<05:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2756/3295 [28:30<05:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2756/3295 [28:30<05:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2757/3295 [28:30<05:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2757/3295 [28:30<05:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2758/3295 [28:30<05:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2758/3295 [28:30<05:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2759/3295 [28:32<05:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▎ | 2759/3295 [28:32<05:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2760/3295 [28:32<05:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2760/3295 [28:32<05:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2761/3295 [28:32<05:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2761/3295 [28:32<05:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2762/3295 [28:34<05:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2762/3295 [28:34<05:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2763/3295 [28:34<05:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2763/3295 [28:34<05:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2764/3295 [28:34<05:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2764/3295 [28:34<05:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2765/3295 [28:35<05:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2765/3295 [28:35<05:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2766/3295 [28:36<05:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2766/3295 [28:36<05:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2767/3295 [28:36<05:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2767/3295 [28:36<05:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2768/3295 [28:37<05:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2768/3295 [28:37<05:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2769/3295 [28:37<05:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2769/3295 [28:37<05:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2770/3295 [28:37<05:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2770/3295 [28:37<05:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2771/3295 [28:39<05:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2771/3295 [28:39<05:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2772/3295 [28:39<05:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2772/3295 [28:39<05:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2773/3295 [28:39<05:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2773/3295 [28:39<05:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2774/3295 [28:41<05:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2774/3295 [28:41<05:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2775/3295 [28:41<05:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2775/3295 [28:41<05:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2776/3295 [28:41<05:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2776/3295 [28:41<05:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2777/3295 [28:43<05:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2777/3295 [28:43<05:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2778/3295 [28:43<05:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2778/3295 [28:43<05:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2779/3295 [28:43<05:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2779/3295 [28:43<05:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2780/3295 [28:45<05:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2780/3295 [28:45<05:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2781/3295 [28:45<05:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2781/3295 [28:45<05:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2782/3295 [28:45<05:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2782/3295 [28:45<05:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2783/3295 [28:46<05:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2783/3295 [28:46<05:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2784/3295 [28:46<05:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  84%|████████▍ | 2784/3295 [28:46<05:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2785/3295 [28:46<05:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2785/3295 [28:46<05:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2786/3295 [28:48<05:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2786/3295 [28:48<05:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2787/3295 [28:48<05:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2787/3295 [28:48<05:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2788/3295 [28:48<05:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2788/3295 [28:48<05:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2789/3295 [28:50<05:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2789/3295 [28:50<05:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2790/3295 [28:50<05:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2790/3295 [28:50<05:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2791/3295 [28:50<05:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2791/3295 [28:50<05:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2792/3295 [28:52<05:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2792/3295 [28:52<05:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2793/3295 [28:52<05:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2793/3295 [28:52<05:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2794/3295 [28:52<05:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2794/3295 [28:52<05:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2795/3295 [28:54<05:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2795/3295 [28:54<05:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2796/3295 [28:54<05:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2796/3295 [28:54<05:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2797/3295 [28:54<05:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2797/3295 [28:54<05:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2798/3295 [28:56<05:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2798/3295 [28:56<05:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2799/3295 [28:56<05:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2799/3295 [28:56<05:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2800/3295 [28:56<05:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▍ | 2800/3295 [28:56<05:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2801/3295 [28:58<05:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2801/3295 [28:58<05:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2802/3295 [28:58<05:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2802/3295 [28:58<05:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2803/3295 [28:58<05:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2803/3295 [28:58<05:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2804/3295 [28:59<05:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2804/3295 [28:59<05:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2805/3295 [28:59<05:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2805/3295 [28:59<05:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2806/3295 [28:59<05:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2806/3295 [28:59<05:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2807/3295 [29:01<05:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2807/3295 [29:01<05:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2808/3295 [29:01<05:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2808/3295 [29:01<05:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2809/3295 [29:01<05:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2809/3295 [29:01<05:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2810/3295 [29:03<05:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2810/3295 [29:03<05:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2811/3295 [29:03<05:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2811/3295 [29:03<05:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2812/3295 [29:03<04:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2812/3295 [29:03<04:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2813/3295 [29:05<04:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2813/3295 [29:05<04:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2814/3295 [29:05<04:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2814/3295 [29:05<04:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2815/3295 [29:05<04:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2815/3295 [29:05<04:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2816/3295 [29:07<04:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2816/3295 [29:07<04:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2817/3295 [29:07<04:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  85%|████████▌ | 2817/3295 [29:07<04:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2818/3295 [29:07<04:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2818/3295 [29:07<04:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2819/3295 [29:09<04:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2819/3295 [29:09<04:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2820/3295 [29:09<04:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2820/3295 [29:09<04:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2821/3295 [29:09<04:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2821/3295 [29:09<04:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2822/3295 [29:10<04:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2822/3295 [29:10<04:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2823/3295 [29:10<04:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2823/3295 [29:10<04:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2824/3295 [29:10<04:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2824/3295 [29:10<04:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2825/3295 [29:12<04:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2825/3295 [29:12<04:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2826/3295 [29:12<04:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2826/3295 [29:12<04:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2827/3295 [29:12<04:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2827/3295 [29:12<04:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2828/3295 [29:14<04:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2828/3295 [29:14<04:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2829/3295 [29:14<04:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2829/3295 [29:14<04:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2830/3295 [29:14<04:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2830/3295 [29:14<04:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2831/3295 [29:16<04:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2831/3295 [29:16<04:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2832/3295 [29:16<04:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2832/3295 [29:16<04:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2833/3295 [29:16<04:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2833/3295 [29:16<04:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2834/3295 [29:18<04:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2834/3295 [29:18<04:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2835/3295 [29:18<04:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2835/3295 [29:18<04:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2836/3295 [29:18<04:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2836/3295 [29:18<04:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2837/3295 [29:20<04:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2837/3295 [29:20<04:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2838/3295 [29:20<04:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2838/3295 [29:20<04:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2839/3295 [29:20<04:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2839/3295 [29:20<04:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2840/3295 [29:22<04:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2840/3295 [29:22<04:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2841/3295 [29:22<04:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▌ | 2841/3295 [29:22<04:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2842/3295 [29:22<04:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2842/3295 [29:22<04:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2843/3295 [29:23<04:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2843/3295 [29:23<04:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2844/3295 [29:24<04:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2844/3295 [29:24<04:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2845/3295 [29:24<04:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2845/3295 [29:24<04:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2846/3295 [29:25<04:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2846/3295 [29:25<04:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2847/3295 [29:25<04:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2847/3295 [29:25<04:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2848/3295 [29:25<04:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2848/3295 [29:25<04:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2849/3295 [29:27<04:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2849/3295 [29:27<04:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2850/3295 [29:27<04:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  86%|████████▋ | 2850/3295 [29:27<04:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2851/3295 [29:27<04:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2851/3295 [29:27<04:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2852/3295 [29:29<04:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2852/3295 [29:29<04:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2853/3295 [29:29<04:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2853/3295 [29:29<04:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2854/3295 [29:29<04:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2854/3295 [29:29<04:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2855/3295 [29:31<04:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2855/3295 [29:31<04:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2856/3295 [29:31<04:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2856/3295 [29:31<04:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2857/3295 [29:31<04:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2857/3295 [29:31<04:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2858/3295 [29:33<04:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2858/3295 [29:33<04:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2859/3295 [29:33<04:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2859/3295 [29:33<04:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2860/3295 [29:33<04:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2860/3295 [29:33<04:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2861/3295 [29:35<04:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2861/3295 [29:35<04:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2862/3295 [29:35<04:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2862/3295 [29:35<04:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2863/3295 [29:35<04:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2863/3295 [29:35<04:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2864/3295 [29:36<04:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2864/3295 [29:36<04:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2865/3295 [29:36<04:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2865/3295 [29:36<04:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2866/3295 [29:36<04:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2866/3295 [29:36<04:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2867/3295 [29:38<04:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2867/3295 [29:38<04:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2868/3295 [29:38<04:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2868/3295 [29:38<04:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2869/3295 [29:38<04:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2869/3295 [29:38<04:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2870/3295 [29:40<04:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2870/3295 [29:40<04:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2871/3295 [29:40<04:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2871/3295 [29:40<04:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2872/3295 [29:40<04:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2872/3295 [29:40<04:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2873/3295 [29:42<04:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2873/3295 [29:42<04:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2874/3295 [29:42<04:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2874/3295 [29:42<04:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2875/3295 [29:42<04:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2875/3295 [29:42<04:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2876/3295 [29:44<04:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2876/3295 [29:44<04:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2877/3295 [29:44<04:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2877/3295 [29:44<04:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2878/3295 [29:44<04:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2878/3295 [29:44<04:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2879/3295 [29:46<04:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2879/3295 [29:46<04:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2880/3295 [29:46<04:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2880/3295 [29:46<04:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2881/3295 [29:46<04:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2881/3295 [29:46<04:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2882/3295 [29:48<04:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2882/3295 [29:48<04:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2883/3295 [29:48<04:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  87%|████████▋ | 2883/3295 [29:48<04:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2884/3295 [29:48<04:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2884/3295 [29:48<04:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2885/3295 [29:49<04:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2885/3295 [29:49<04:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2886/3295 [29:49<04:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2886/3295 [29:49<04:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2887/3295 [29:49<04:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2887/3295 [29:49<04:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2888/3295 [29:51<04:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2888/3295 [29:51<04:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2889/3295 [29:51<04:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2889/3295 [29:51<04:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2890/3295 [29:51<04:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2890/3295 [29:51<04:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2891/3295 [29:53<04:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2891/3295 [29:53<04:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2892/3295 [29:53<04:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2892/3295 [29:53<04:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2893/3295 [29:53<04:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2893/3295 [29:53<04:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2894/3295 [29:55<04:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2894/3295 [29:55<04:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2895/3295 [29:55<04:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2895/3295 [29:55<04:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2896/3295 [29:55<04:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2896/3295 [29:55<04:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2897/3295 [29:57<04:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2897/3295 [29:57<04:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2898/3295 [29:57<04:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2898/3295 [29:57<04:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2899/3295 [29:57<04:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2899/3295 [29:57<04:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2900/3295 [29:59<04:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2900/3295 [29:59<04:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2901/3295 [29:59<04:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2901/3295 [29:59<04:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2902/3295 [29:59<04:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2902/3295 [29:59<04:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2903/3295 [30:00<04:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2903/3295 [30:00<04:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2904/3295 [30:00<04:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2904/3295 [30:00<04:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2905/3295 [30:00<04:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2905/3295 [30:00<04:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2906/3295 [30:02<04:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2906/3295 [30:02<04:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2907/3295 [30:02<04:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2907/3295 [30:02<04:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2908/3295 [30:02<03:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2908/3295 [30:02<03:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2909/3295 [30:04<03:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2909/3295 [30:04<03:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2910/3295 [30:04<03:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2910/3295 [30:04<03:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2911/3295 [30:04<03:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2911/3295 [30:04<03:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2912/3295 [30:06<03:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2912/3295 [30:06<03:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2913/3295 [30:06<03:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2913/3295 [30:06<03:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2914/3295 [30:06<03:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2914/3295 [30:06<03:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2915/3295 [30:08<03:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2915/3295 [30:08<03:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2916/3295 [30:08<03:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  88%|████████▊ | 2916/3295 [30:08<03:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2917/3295 [30:08<03:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2917/3295 [30:08<03:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2918/3295 [30:10<03:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2918/3295 [30:10<03:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2919/3295 [30:10<03:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2919/3295 [30:10<03:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2920/3295 [30:10<03:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2920/3295 [30:10<03:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2921/3295 [30:11<03:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2921/3295 [30:11<03:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2922/3295 [30:11<03:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2922/3295 [30:11<03:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2923/3295 [30:12<03:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2923/3295 [30:12<03:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2924/3295 [30:13<03:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▊ | 2924/3295 [30:13<03:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2925/3295 [30:13<03:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2925/3295 [30:13<03:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2926/3295 [30:13<03:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2926/3295 [30:13<03:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2927/3295 [30:15<03:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2927/3295 [30:15<03:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2928/3295 [30:15<03:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2928/3295 [30:15<03:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2929/3295 [30:15<03:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2929/3295 [30:15<03:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2930/3295 [30:17<03:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2930/3295 [30:17<03:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2931/3295 [30:17<03:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2931/3295 [30:17<03:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2932/3295 [30:17<03:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2932/3295 [30:17<03:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2933/3295 [30:19<03:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2933/3295 [30:19<03:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2934/3295 [30:19<03:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2934/3295 [30:19<03:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2935/3295 [30:19<03:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2935/3295 [30:19<03:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2936/3295 [30:21<03:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2936/3295 [30:21<03:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2937/3295 [30:21<03:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2937/3295 [30:21<03:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2938/3295 [30:21<03:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2938/3295 [30:21<03:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2939/3295 [30:23<03:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2939/3295 [30:23<03:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2940/3295 [30:23<03:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2940/3295 [30:23<03:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2941/3295 [30:23<03:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2941/3295 [30:23<03:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2942/3295 [30:25<03:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2942/3295 [30:25<03:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2943/3295 [30:25<03:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2943/3295 [30:25<03:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2944/3295 [30:25<03:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2944/3295 [30:25<03:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2945/3295 [30:27<03:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2945/3295 [30:27<03:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2946/3295 [30:27<03:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2946/3295 [30:27<03:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2947/3295 [30:27<03:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2947/3295 [30:27<03:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2948/3295 [30:29<03:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2948/3295 [30:29<03:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2949/3295 [30:29<03:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  89%|████████▉ | 2949/3295 [30:29<03:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2950/3295 [30:29<03:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2950/3295 [30:29<03:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2951/3295 [30:30<03:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2951/3295 [30:30<03:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2952/3295 [30:31<03:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2952/3295 [30:31<03:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2953/3295 [30:31<03:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2953/3295 [30:31<03:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2954/3295 [30:32<03:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2954/3295 [30:32<03:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2955/3295 [30:32<03:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2955/3295 [30:32<03:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2956/3295 [30:32<03:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2956/3295 [30:32<03:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2957/3295 [30:34<03:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2957/3295 [30:34<03:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2958/3295 [30:34<03:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2958/3295 [30:34<03:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2959/3295 [30:34<03:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2959/3295 [30:34<03:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2960/3295 [30:36<03:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2960/3295 [30:36<03:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2961/3295 [30:36<03:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2961/3295 [30:36<03:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2962/3295 [30:36<03:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2962/3295 [30:36<03:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2963/3295 [30:38<03:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2963/3295 [30:38<03:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2964/3295 [30:38<03:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2964/3295 [30:38<03:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2965/3295 [30:38<03:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|████████▉ | 2965/3295 [30:38<03:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2966/3295 [30:39<03:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2966/3295 [30:39<03:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2967/3295 [30:39<03:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2967/3295 [30:39<03:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2968/3295 [30:39<03:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2968/3295 [30:39<03:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2969/3295 [30:41<03:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2969/3295 [30:41<03:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2970/3295 [30:41<03:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2970/3295 [30:41<03:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2971/3295 [30:41<03:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2971/3295 [30:41<03:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2972/3295 [30:43<03:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2972/3295 [30:43<03:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2973/3295 [30:43<03:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2973/3295 [30:43<03:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2974/3295 [30:43<03:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2974/3295 [30:43<03:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2975/3295 [30:45<03:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2975/3295 [30:45<03:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2976/3295 [30:45<03:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2976/3295 [30:45<03:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2977/3295 [30:45<03:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2977/3295 [30:45<03:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2978/3295 [30:47<03:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2978/3295 [30:47<03:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2979/3295 [30:47<03:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2979/3295 [30:47<03:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2980/3295 [30:47<03:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2980/3295 [30:47<03:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2981/3295 [30:48<03:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  90%|█████████ | 2981/3295 [30:48<03:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2982/3295 [30:48<03:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2982/3295 [30:48<03:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2983/3295 [30:48<03:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2983/3295 [30:48<03:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2984/3295 [30:50<03:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2984/3295 [30:50<03:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2985/3295 [30:50<03:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2985/3295 [30:50<03:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2986/3295 [30:50<03:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2986/3295 [30:50<03:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2987/3295 [30:52<03:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2987/3295 [30:52<03:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2988/3295 [30:52<03:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2988/3295 [30:52<03:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2989/3295 [30:52<03:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2989/3295 [30:52<03:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2990/3295 [30:54<03:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2990/3295 [30:54<03:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2991/3295 [30:54<03:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2991/3295 [30:54<03:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2992/3295 [30:54<03:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2992/3295 [30:54<03:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2993/3295 [30:56<03:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2993/3295 [30:56<03:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2994/3295 [30:56<03:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2994/3295 [30:56<03:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2995/3295 [30:56<03:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2995/3295 [30:56<03:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2996/3295 [30:58<03:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2996/3295 [30:58<03:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2997/3295 [30:58<03:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2997/3295 [30:58<03:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2998/3295 [30:58<03:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2998/3295 [30:58<03:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2999/3295 [31:00<03:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 2999/3295 [31:00<03:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3000/3295 [31:00<03:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3000/3295 [31:00<03:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3001/3295 [31:00<03:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3001/3295 [31:00<03:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3002/3295 [31:02<03:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3002/3295 [31:02<03:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3003/3295 [31:02<03:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3003/3295 [31:02<03:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3004/3295 [31:02<03:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3004/3295 [31:02<03:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3005/3295 [31:03<02:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3005/3295 [31:03<02:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3006/3295 [31:03<02:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████ | 3006/3295 [31:03<02:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3007/3295 [31:03<02:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3007/3295 [31:03<02:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3008/3295 [31:05<02:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3008/3295 [31:05<02:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3009/3295 [31:05<02:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3009/3295 [31:05<02:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3010/3295 [31:05<02:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3010/3295 [31:05<02:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3011/3295 [31:07<02:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3011/3295 [31:07<02:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3012/3295 [31:07<02:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3012/3295 [31:07<02:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3013/3295 [31:07<02:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3013/3295 [31:07<02:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3014/3295 [31:09<02:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  91%|█████████▏| 3014/3295 [31:09<02:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3015/3295 [31:09<02:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3015/3295 [31:09<02:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3016/3295 [31:09<02:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3016/3295 [31:09<02:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3017/3295 [31:11<02:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3017/3295 [31:11<02:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3018/3295 [31:11<02:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3018/3295 [31:11<02:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3019/3295 [31:11<02:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3019/3295 [31:11<02:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3020/3295 [31:13<02:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3020/3295 [31:13<02:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3021/3295 [31:13<02:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3021/3295 [31:13<02:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3022/3295 [31:13<02:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3022/3295 [31:13<02:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3023/3295 [31:15<02:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3023/3295 [31:15<02:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3024/3295 [31:15<02:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3024/3295 [31:15<02:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3025/3295 [31:15<02:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3025/3295 [31:15<02:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3026/3295 [31:17<02:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3026/3295 [31:17<02:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3027/3295 [31:17<02:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3027/3295 [31:17<02:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3028/3295 [31:17<02:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3028/3295 [31:17<02:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3029/3295 [31:19<02:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3029/3295 [31:19<02:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3030/3295 [31:19<02:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3030/3295 [31:19<02:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3031/3295 [31:19<02:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3031/3295 [31:19<02:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3032/3295 [31:20<02:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3032/3295 [31:20<02:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3033/3295 [31:20<02:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3033/3295 [31:20<02:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3034/3295 [31:21<02:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3034/3295 [31:21<02:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3035/3295 [31:22<02:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3035/3295 [31:22<02:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3036/3295 [31:22<02:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3036/3295 [31:22<02:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3037/3295 [31:22<02:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3037/3295 [31:22<02:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3038/3295 [31:24<02:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3038/3295 [31:24<02:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3039/3295 [31:24<02:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3039/3295 [31:24<02:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3040/3295 [31:24<02:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3040/3295 [31:24<02:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3041/3295 [31:26<02:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3041/3295 [31:26<02:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3042/3295 [31:26<02:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3042/3295 [31:26<02:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3043/3295 [31:26<02:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3043/3295 [31:26<02:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3044/3295 [31:28<02:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3044/3295 [31:28<02:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3045/3295 [31:28<02:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3045/3295 [31:28<02:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3046/3295 [31:28<02:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3046/3295 [31:28<02:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3047/3295 [31:29<02:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  92%|█████████▏| 3047/3295 [31:29<02:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3048/3295 [31:29<02:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3048/3295 [31:29<02:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3049/3295 [31:29<02:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3049/3295 [31:29<02:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3050/3295 [31:31<02:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3050/3295 [31:31<02:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3051/3295 [31:31<02:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3051/3295 [31:31<02:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3052/3295 [31:31<02:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3052/3295 [31:31<02:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3053/3295 [31:33<02:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3053/3295 [31:33<02:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3054/3295 [31:33<02:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3054/3295 [31:33<02:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3055/3295 [31:33<02:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3055/3295 [31:33<02:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3056/3295 [31:35<02:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3056/3295 [31:35<02:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3057/3295 [31:35<02:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3057/3295 [31:35<02:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3058/3295 [31:35<02:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3058/3295 [31:35<02:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3059/3295 [31:37<02:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3059/3295 [31:37<02:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3060/3295 [31:37<02:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3060/3295 [31:37<02:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3061/3295 [31:37<02:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3061/3295 [31:37<02:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3062/3295 [31:39<02:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3062/3295 [31:39<02:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3063/3295 [31:39<02:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3063/3295 [31:39<02:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3064/3295 [31:39<02:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3064/3295 [31:39<02:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3065/3295 [31:41<02:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3065/3295 [31:41<02:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3066/3295 [31:41<02:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3066/3295 [31:41<02:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3067/3295 [31:41<02:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3067/3295 [31:41<02:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3068/3295 [31:42<02:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3068/3295 [31:42<02:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3069/3295 [31:42<02:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3069/3295 [31:42<02:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3070/3295 [31:42<02:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3070/3295 [31:42<02:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3071/3295 [31:44<02:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3071/3295 [31:44<02:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3072/3295 [31:44<02:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3072/3295 [31:44<02:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3073/3295 [31:44<02:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3073/3295 [31:44<02:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3074/3295 [31:46<02:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3074/3295 [31:46<02:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3075/3295 [31:46<02:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3075/3295 [31:46<02:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3076/3295 [31:46<02:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3076/3295 [31:46<02:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3077/3295 [31:48<02:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3077/3295 [31:48<02:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3078/3295 [31:48<02:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3078/3295 [31:48<02:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3079/3295 [31:48<02:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3079/3295 [31:48<02:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3080/3295 [31:50<02:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  93%|█████████▎| 3080/3295 [31:50<02:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3081/3295 [31:50<02:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3081/3295 [31:50<02:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3082/3295 [31:50<02:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3082/3295 [31:50<02:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3083/3295 [31:52<02:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3083/3295 [31:52<02:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3084/3295 [31:52<02:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3084/3295 [31:52<02:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3085/3295 [31:52<02:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3085/3295 [31:52<02:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3086/3295 [31:54<02:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3086/3295 [31:54<02:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3087/3295 [31:54<02:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3087/3295 [31:54<02:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3088/3295 [31:54<02:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3088/3295 [31:54<02:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3089/3295 [31:55<02:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▎| 3089/3295 [31:55<02:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3090/3295 [31:55<02:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3090/3295 [31:55<02:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3091/3295 [31:55<02:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3091/3295 [31:55<02:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3092/3295 [31:57<02:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3092/3295 [31:57<02:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3093/3295 [31:57<02:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3093/3295 [31:57<02:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3094/3295 [31:57<02:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3094/3295 [31:57<02:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3095/3295 [31:59<02:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3095/3295 [31:59<02:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3096/3295 [31:59<02:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3096/3295 [31:59<02:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3097/3295 [31:59<02:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3097/3295 [31:59<02:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3098/3295 [32:01<02:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3098/3295 [32:01<02:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3099/3295 [32:01<02:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3099/3295 [32:01<02:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3100/3295 [32:01<02:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3100/3295 [32:01<02:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3101/3295 [32:03<02:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3101/3295 [32:03<02:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3102/3295 [32:03<01:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3102/3295 [32:03<01:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3103/3295 [32:03<01:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3103/3295 [32:03<01:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3104/3295 [32:04<01:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3104/3295 [32:04<01:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3105/3295 [32:04<01:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3105/3295 [32:04<01:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3106/3295 [32:04<01:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3106/3295 [32:04<01:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3107/3295 [32:06<01:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3107/3295 [32:06<01:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3108/3295 [32:06<01:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3108/3295 [32:06<01:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3109/3295 [32:06<01:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3109/3295 [32:06<01:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3110/3295 [32:08<01:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3110/3295 [32:08<01:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3111/3295 [32:08<01:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3111/3295 [32:08<01:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3112/3295 [32:08<01:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3112/3295 [32:08<01:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3113/3295 [32:10<01:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  94%|█████████▍| 3113/3295 [32:10<01:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3114/3295 [32:10<01:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3114/3295 [32:10<01:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3115/3295 [32:10<01:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3115/3295 [32:10<01:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3116/3295 [32:12<01:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3116/3295 [32:12<01:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3117/3295 [32:12<01:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3117/3295 [32:12<01:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3118/3295 [32:12<01:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3118/3295 [32:12<01:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3119/3295 [32:14<01:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3119/3295 [32:14<01:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3120/3295 [32:14<01:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3120/3295 [32:14<01:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3121/3295 [32:14<01:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3121/3295 [32:14<01:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3122/3295 [32:15<01:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3122/3295 [32:15<01:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3123/3295 [32:15<01:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3123/3295 [32:15<01:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3124/3295 [32:15<01:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3124/3295 [32:15<01:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3125/3295 [32:17<01:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3125/3295 [32:17<01:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3126/3295 [32:17<01:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3126/3295 [32:17<01:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3127/3295 [32:17<01:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3127/3295 [32:17<01:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3128/3295 [32:19<01:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3128/3295 [32:19<01:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3129/3295 [32:19<01:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3129/3295 [32:19<01:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3130/3295 [32:19<01:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▍| 3130/3295 [32:19<01:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3131/3295 [32:21<01:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3131/3295 [32:21<01:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3132/3295 [32:21<01:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3132/3295 [32:21<01:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3133/3295 [32:21<01:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3133/3295 [32:21<01:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3134/3295 [32:23<01:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3134/3295 [32:23<01:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3135/3295 [32:23<01:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3135/3295 [32:23<01:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3136/3295 [32:23<01:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3136/3295 [32:23<01:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3137/3295 [32:24<01:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3137/3295 [32:24<01:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3138/3295 [32:24<01:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3138/3295 [32:24<01:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3139/3295 [32:24<01:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3139/3295 [32:24<01:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3140/3295 [32:26<01:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3140/3295 [32:26<01:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3141/3295 [32:26<01:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3141/3295 [32:26<01:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3142/3295 [32:26<01:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3142/3295 [32:26<01:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3143/3295 [32:28<01:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3143/3295 [32:28<01:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3144/3295 [32:28<01:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3144/3295 [32:28<01:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3145/3295 [32:28<01:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3145/3295 [32:28<01:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3146/3295 [32:30<01:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  95%|█████████▌| 3146/3295 [32:30<01:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3147/3295 [32:30<01:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3147/3295 [32:30<01:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3148/3295 [32:30<01:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3148/3295 [32:30<01:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3149/3295 [32:32<01:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3149/3295 [32:32<01:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3150/3295 [32:32<01:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3150/3295 [32:32<01:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3151/3295 [32:32<01:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3151/3295 [32:32<01:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3152/3295 [32:34<01:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3152/3295 [32:34<01:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3153/3295 [32:34<01:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3153/3295 [32:34<01:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3154/3295 [32:34<01:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3154/3295 [32:34<01:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3155/3295 [32:36<01:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3155/3295 [32:36<01:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3156/3295 [32:36<01:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3156/3295 [32:36<01:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3157/3295 [32:36<01:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3157/3295 [32:36<01:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3158/3295 [32:37<01:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3158/3295 [32:37<01:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3159/3295 [32:37<01:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3159/3295 [32:37<01:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3160/3295 [32:37<01:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3160/3295 [32:37<01:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3161/3295 [32:39<01:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3161/3295 [32:39<01:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3162/3295 [32:39<01:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3162/3295 [32:39<01:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3163/3295 [32:39<01:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3163/3295 [32:39<01:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3164/3295 [32:41<01:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3164/3295 [32:41<01:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3165/3295 [32:41<01:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3165/3295 [32:41<01:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3166/3295 [32:41<01:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3166/3295 [32:41<01:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3167/3295 [32:43<01:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3167/3295 [32:43<01:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3168/3295 [32:43<01:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3168/3295 [32:43<01:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3169/3295 [32:43<01:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3169/3295 [32:43<01:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3170/3295 [32:45<01:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3170/3295 [32:45<01:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3171/3295 [32:45<01:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▌| 3171/3295 [32:45<01:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3172/3295 [32:45<01:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3172/3295 [32:45<01:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3173/3295 [32:46<01:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3173/3295 [32:46<01:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3174/3295 [32:47<01:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3174/3295 [32:47<01:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3175/3295 [32:47<01:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3175/3295 [32:47<01:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3176/3295 [32:48<01:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3176/3295 [32:48<01:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3177/3295 [32:48<01:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3177/3295 [32:48<01:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3178/3295 [32:48<01:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3178/3295 [32:48<01:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3179/3295 [32:50<01:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  96%|█████████▋| 3179/3295 [32:50<01:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3180/3295 [32:50<01:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3180/3295 [32:50<01:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3181/3295 [32:50<01:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3181/3295 [32:50<01:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3182/3295 [32:52<01:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3182/3295 [32:52<01:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3183/3295 [32:52<01:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3183/3295 [32:52<01:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3184/3295 [32:52<01:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3184/3295 [32:52<01:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3185/3295 [32:54<01:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3185/3295 [32:54<01:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3186/3295 [32:54<01:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3186/3295 [32:54<01:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3187/3295 [32:54<01:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3187/3295 [32:54<01:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3188/3295 [32:56<01:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3188/3295 [32:56<01:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3189/3295 [32:56<01:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3189/3295 [32:56<01:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3190/3295 [32:56<01:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3190/3295 [32:56<01:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3191/3295 [32:57<01:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3191/3295 [32:57<01:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3192/3295 [32:57<01:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3192/3295 [32:57<01:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3193/3295 [32:57<01:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3193/3295 [32:57<01:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3194/3295 [32:59<01:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3194/3295 [32:59<01:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3195/3295 [32:59<01:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3195/3295 [32:59<01:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3196/3295 [32:59<01:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3196/3295 [32:59<01:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3197/3295 [33:01<01:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3197/3295 [33:01<01:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3198/3295 [33:01<01:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3198/3295 [33:01<01:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3199/3295 [33:01<00:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3199/3295 [33:01<00:59,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3200/3295 [33:03<00:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3200/3295 [33:03<00:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3201/3295 [33:03<00:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3201/3295 [33:03<00:58,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3202/3295 [33:03<00:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3202/3295 [33:03<00:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3203/3295 [33:05<00:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3203/3295 [33:05<00:57,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3204/3295 [33:05<00:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3204/3295 [33:05<00:56,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3205/3295 [33:05<00:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3205/3295 [33:05<00:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3206/3295 [33:06<00:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3206/3295 [33:06<00:55,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3207/3295 [33:06<00:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3207/3295 [33:06<00:54,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3208/3295 [33:07<00:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3208/3295 [33:07<00:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3209/3295 [33:08<00:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3209/3295 [33:08<00:53,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3210/3295 [33:08<00:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3210/3295 [33:08<00:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3211/3295 [33:08<00:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3211/3295 [33:08<00:52,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3212/3295 [33:10<00:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  97%|█████████▋| 3212/3295 [33:10<00:51,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3213/3295 [33:10<00:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3213/3295 [33:10<00:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3214/3295 [33:10<00:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3214/3295 [33:10<00:50,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3215/3295 [33:12<00:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3215/3295 [33:12<00:49,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3216/3295 [33:12<00:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3216/3295 [33:12<00:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3217/3295 [33:12<00:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3217/3295 [33:12<00:48,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3218/3295 [33:14<00:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3218/3295 [33:14<00:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3219/3295 [33:14<00:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3219/3295 [33:14<00:47,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3220/3295 [33:14<00:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3220/3295 [33:14<00:46,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3221/3295 [33:16<00:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3221/3295 [33:16<00:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3222/3295 [33:16<00:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3222/3295 [33:16<00:45,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3223/3295 [33:16<00:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3223/3295 [33:16<00:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3224/3295 [33:18<00:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3224/3295 [33:18<00:44,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3225/3295 [33:18<00:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3225/3295 [33:18<00:43,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3226/3295 [33:18<00:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3226/3295 [33:18<00:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3227/3295 [33:20<00:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3227/3295 [33:20<00:42,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3228/3295 [33:20<00:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3228/3295 [33:20<00:41,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3229/3295 [33:20<00:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3229/3295 [33:20<00:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3230/3295 [33:22<00:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3230/3295 [33:22<00:40,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3231/3295 [33:22<00:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3231/3295 [33:22<00:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3232/3295 [33:22<00:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3232/3295 [33:22<00:39,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3233/3295 [33:24<00:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3233/3295 [33:24<00:38,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3234/3295 [33:24<00:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3234/3295 [33:24<00:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3235/3295 [33:24<00:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3235/3295 [33:24<00:37,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3236/3295 [33:26<00:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3236/3295 [33:26<00:36,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3237/3295 [33:26<00:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3237/3295 [33:26<00:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3238/3295 [33:26<00:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3238/3295 [33:26<00:35,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3239/3295 [33:27<00:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3239/3295 [33:27<00:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3240/3295 [33:27<00:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3240/3295 [33:27<00:34,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3241/3295 [33:27<00:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3241/3295 [33:27<00:33,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3242/3295 [33:29<00:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3242/3295 [33:29<00:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3243/3295 [33:29<00:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3243/3295 [33:29<00:32,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3244/3295 [33:29<00:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3244/3295 [33:29<00:31,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3245/3295 [33:31<00:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  98%|█████████▊| 3245/3295 [33:31<00:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3246/3295 [33:31<00:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3246/3295 [33:31<00:30,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3247/3295 [33:31<00:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3247/3295 [33:31<00:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3248/3295 [33:33<00:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3248/3295 [33:33<00:29,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3249/3295 [33:33<00:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3249/3295 [33:33<00:28,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3250/3295 [33:33<00:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3250/3295 [33:33<00:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3251/3295 [33:35<00:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3251/3295 [33:35<00:27,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3252/3295 [33:35<00:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3252/3295 [33:35<00:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3253/3295 [33:35<00:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▊| 3253/3295 [33:35<00:26,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3254/3295 [33:37<00:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3254/3295 [33:37<00:25,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3255/3295 [33:37<00:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3255/3295 [33:37<00:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3256/3295 [33:37<00:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3256/3295 [33:37<00:24,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3257/3295 [33:39<00:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3257/3295 [33:39<00:23,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3258/3295 [33:39<00:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3258/3295 [33:39<00:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3259/3295 [33:39<00:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3259/3295 [33:39<00:22,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3260/3295 [33:41<00:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3260/3295 [33:41<00:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3261/3295 [33:41<00:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3261/3295 [33:41<00:21,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3262/3295 [33:41<00:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3262/3295 [33:41<00:20,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3263/3295 [33:43<00:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3263/3295 [33:43<00:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3264/3295 [33:43<00:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3264/3295 [33:43<00:19,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3265/3295 [33:43<00:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3265/3295 [33:43<00:18,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3266/3295 [33:45<00:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3266/3295 [33:45<00:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3267/3295 [33:45<00:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3267/3295 [33:45<00:17,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3268/3295 [33:45<00:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3268/3295 [33:45<00:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3269/3295 [33:47<00:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3269/3295 [33:47<00:16,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3270/3295 [33:47<00:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3270/3295 [33:47<00:15,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3271/3295 [33:47<00:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3271/3295 [33:47<00:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3272/3295 [33:49<00:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3272/3295 [33:49<00:14,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3273/3295 [33:49<00:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3273/3295 [33:49<00:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3274/3295 [33:49<00:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3274/3295 [33:49<00:13,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3275/3295 [33:50<00:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3275/3295 [33:50<00:12,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3276/3295 [33:51<00:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3276/3295 [33:51<00:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3277/3295 [33:51<00:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3277/3295 [33:51<00:11,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3278/3295 [33:52<00:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2:  99%|█████████▉| 3278/3295 [33:52<00:10,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3279/3295 [33:52<00:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3279/3295 [33:52<00:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3280/3295 [33:52<00:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3280/3295 [33:52<00:09,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3281/3295 [33:54<00:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3281/3295 [33:54<00:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3282/3295 [33:54<00:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3282/3295 [33:54<00:08,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3283/3295 [33:54<00:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3283/3295 [33:54<00:07,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3284/3295 [33:56<00:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3284/3295 [33:56<00:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3285/3295 [33:56<00:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3285/3295 [33:56<00:06,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3286/3295 [33:56<00:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3286/3295 [33:56<00:05,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3287/3295 [33:58<00:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3287/3295 [33:58<00:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3288/3295 [33:58<00:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3288/3295 [33:58<00:04,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3289/3295 [33:58<00:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3289/3295 [33:58<00:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3290/3295 [33:59<00:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3290/3295 [33:59<00:03,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3291/3295 [34:00<00:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3291/3295 [34:00<00:02,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3292/3295 [34:00<00:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3292/3295 [34:00<00:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3293/3295 [34:01<00:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3293/3295 [34:01<00:01,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3294/3295 [34:01<00:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|█████████▉| 3294/3295 [34:01<00:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|██████████| 3295/3295 [34:01<00:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|██████████| 3295/3295 [34:01<00:00,  1.61it/s, R@1 (val)=0.395, R@5 (val)=0.603, R@20 (val)=0.774, loss (train)=0.400, R@1 (train)=0.902]

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 0/602 [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 1/602 [00:00<00:09, 66.51it/s][A

Validation DataLoader 0:   0%|          | 2/602 [00:00<00:09, 64.20it/s][A

Validation DataLoader 0:   0%|          | 3/602 [00:00<00:09, 63.43it/s][A

Validation DataLoader 0:   1%|          | 4/602 [00:00<00:09, 64.98it/s][A

Validation DataLoader 0:   1%|          | 5/602 [00:00<00:09, 65.89it/s][A

Validation DataLoader 0:   1%|          | 6/602 [00:00<00:08, 67.17it/s][A

Validation DataLoader 0:   1%|          | 7/602 [00:00<00:39, 15.18it/s][A

Validation DataLoader 0:   1%|▏         | 8/602 [00:00<00:43, 13.71it/s][A

Validation DataLoader 0:   1%|▏         | 9/602 [00:00<00:39, 15.06it/s][A

Validation DataLoader 0:   2%|▏         | 10/602 [00:00<00:55, 10.58it/s][A

Validation DataLoader 0:   2%|▏         | 11/602 [00:00<00:53, 11.01it/s][A

Validation DataLoader 0:   2%|▏         | 12/602 [00:01<00:49, 11.85it/s][A

Validation DataLoader 0:   2%|▏         | 13/602 [00:01<01:00,  9.70it/s][A

Validation DataLoader 0:   2%|▏         | 14/602 [00:01<00:58, 10.05it/s][A

Validation DataLoader 0:   2%|▏         | 15/602 [00:01<00:56, 10.41it/s][A

Validation DataLoader 0:   3%|▎         | 16/602 [00:01<01:08,  8.62it/s][A

Validation DataLoader 0:   3%|▎         | 17/602 [00:01<01:04,  9.09it/s][A

Validation DataLoader 0:   3%|▎         | 18/602 [00:01<01:01,  9.55it/s][A

Validation DataLoader 0:   3%|▎         | 19/602 [00:02<01:11,  8.12it/s][A

Validation DataLoader 0:   3%|▎         | 20/602 [00:02<01:08,  8.50it/s][A

Validation DataLoader 0:   3%|▎         | 21/602 [00:02<01:05,  8.87it/s][A

Validation DataLoader 0:   4%|▎         | 22/602 [00:02<01:14,  7.83it/s][A

Validation DataLoader 0:   4%|▍         | 23/602 [00:02<01:11,  8.14it/s][A

Validation DataLoader 0:   4%|▍         | 24/602 [00:02<01:08,  8.46it/s][A

Validation DataLoader 0:   4%|▍         | 25/602 [00:03<01:16,  7.59it/s][A

Validation DataLoader 0:   4%|▍         | 26/602 [00:03<01:13,  7.86it/s][A

Validation DataLoader 0:   4%|▍         | 27/602 [00:03<01:10,  8.13it/s][A

Validation DataLoader 0:   5%|▍         | 28/602 [00:03<01:17,  7.43it/s][A

Validation DataLoader 0:   5%|▍         | 29/602 [00:03<01:14,  7.67it/s][A

Validation DataLoader 0:   5%|▍         | 30/602 [00:03<01:12,  7.90it/s][A

Validation DataLoader 0:   5%|▌         | 31/602 [00:04<01:17,  7.41it/s][A

Validation DataLoader 0:   5%|▌         | 32/602 [00:04<01:14,  7.63it/s][A

Validation DataLoader 0:   5%|▌         | 33/602 [00:04<01:12,  7.84it/s][A

Validation DataLoader 0:   6%|▌         | 34/602 [00:04<01:17,  7.34it/s][A

Validation DataLoader 0:   6%|▌         | 35/602 [00:04<01:15,  7.53it/s][A

Validation DataLoader 0:   6%|▌         | 36/602 [00:04<01:13,  7.72it/s][A

Validation DataLoader 0:   6%|▌         | 37/602 [00:05<01:18,  7.21it/s][A

Validation DataLoader 0:   6%|▋         | 38/602 [00:05<01:16,  7.39it/s][A

Validation DataLoader 0:   6%|▋         | 39/602 [00:05<01:14,  7.56it/s][A

Validation DataLoader 0:   7%|▋         | 40/602 [00:05<01:17,  7.22it/s][A

Validation DataLoader 0:   7%|▋         | 41/602 [00:05<01:15,  7.38it/s][A

Validation DataLoader 0:   7%|▋         | 42/602 [00:05<01:14,  7.54it/s][A

Validation DataLoader 0:   7%|▋         | 43/602 [00:05<01:17,  7.21it/s][A

Validation DataLoader 0:   7%|▋         | 44/602 [00:05<01:15,  7.36it/s][A

Validation DataLoader 0:   7%|▋         | 45/602 [00:05<01:14,  7.51it/s][A

Validation DataLoader 0:   8%|▊         | 46/602 [00:06<01:17,  7.20it/s][A

Validation DataLoader 0:   8%|▊         | 47/602 [00:06<01:15,  7.34it/s][A

Validation DataLoader 0:   8%|▊         | 48/602 [00:06<01:14,  7.49it/s][A

Validation DataLoader 0:   8%|▊         | 49/602 [00:06<01:17,  7.13it/s][A

Validation DataLoader 0:   8%|▊         | 50/602 [00:06<01:16,  7.26it/s][A

Validation DataLoader 0:   8%|▊         | 51/602 [00:06<01:14,  7.39it/s][A

Validation DataLoader 0:   9%|▊         | 52/602 [00:07<01:16,  7.15it/s][A

Validation DataLoader 0:   9%|▉         | 53/602 [00:07<01:15,  7.27it/s][A

Validation DataLoader 0:   9%|▉         | 54/602 [00:07<01:14,  7.39it/s][A

Validation DataLoader 0:   9%|▉         | 55/602 [00:07<01:16,  7.17it/s][A

Validation DataLoader 0:   9%|▉         | 56/602 [00:07<01:14,  7.29it/s][A

Validation DataLoader 0:   9%|▉         | 57/602 [00:07<01:13,  7.40it/s][A

Validation DataLoader 0:  10%|▉         | 58/602 [00:08<01:15,  7.23it/s][A

Validation DataLoader 0:  10%|▉         | 59/602 [00:08<01:13,  7.34it/s][A

Validation DataLoader 0:  10%|▉         | 60/602 [00:08<01:13,  7.41it/s][A

Validation DataLoader 0:  10%|█         | 61/602 [00:08<01:15,  7.21it/s][A

Validation DataLoader 0:  10%|█         | 62/602 [00:08<01:13,  7.31it/s][A

Validation DataLoader 0:  10%|█         | 63/602 [00:08<01:12,  7.42it/s][A

Validation DataLoader 0:  11%|█         | 64/602 [00:08<01:14,  7.21it/s][A

Validation DataLoader 0:  11%|█         | 65/602 [00:08<01:13,  7.31it/s][A

Validation DataLoader 0:  11%|█         | 66/602 [00:08<01:12,  7.41it/s][A

Validation DataLoader 0:  11%|█         | 67/602 [00:09<01:14,  7.15it/s][A

Validation DataLoader 0:  11%|█▏        | 68/602 [00:09<01:13,  7.25it/s][A

Validation DataLoader 0:  11%|█▏        | 69/602 [00:09<01:12,  7.34it/s][A

Validation DataLoader 0:  12%|█▏        | 70/602 [00:09<01:14,  7.18it/s][A

Validation DataLoader 0:  12%|█▏        | 71/602 [00:09<01:13,  7.27it/s][A

Validation DataLoader 0:  12%|█▏        | 72/602 [00:09<01:11,  7.36it/s][A

Validation DataLoader 0:  12%|█▏        | 73/602 [00:10<01:14,  7.14it/s][A

Validation DataLoader 0:  12%|█▏        | 74/602 [00:10<01:13,  7.22it/s][A

Validation DataLoader 0:  12%|█▏        | 75/602 [00:10<01:12,  7.31it/s][A

Validation DataLoader 0:  13%|█▎        | 76/602 [00:10<01:14,  7.09it/s][A

Validation DataLoader 0:  13%|█▎        | 77/602 [00:10<01:13,  7.17it/s][A

Validation DataLoader 0:  13%|█▎        | 78/602 [00:10<01:12,  7.26it/s][A

Validation DataLoader 0:  13%|█▎        | 79/602 [00:11<01:14,  6.98it/s][A

Validation DataLoader 0:  13%|█▎        | 80/602 [00:11<01:13,  7.06it/s][A

Validation DataLoader 0:  13%|█▎        | 81/602 [00:11<01:12,  7.14it/s][A

Validation DataLoader 0:  14%|█▎        | 82/602 [00:11<01:14,  6.95it/s][A

Validation DataLoader 0:  14%|█▍        | 83/602 [00:11<01:13,  7.03it/s][A

Validation DataLoader 0:  14%|█▍        | 84/602 [00:11<01:12,  7.11it/s][A

Validation DataLoader 0:  14%|█▍        | 85/602 [00:12<01:14,  6.95it/s][A

Validation DataLoader 0:  14%|█▍        | 86/602 [00:12<01:13,  7.03it/s][A

Validation DataLoader 0:  14%|█▍        | 87/602 [00:12<01:12,  7.10it/s][A

Validation DataLoader 0:  15%|█▍        | 88/602 [00:12<01:13,  6.96it/s][A

Validation DataLoader 0:  15%|█▍        | 89/602 [00:12<01:12,  7.03it/s][A

Validation DataLoader 0:  15%|█▍        | 90/602 [00:12<01:12,  7.10it/s][A

Validation DataLoader 0:  15%|█▌        | 91/602 [00:13<01:13,  6.92it/s][A

Validation DataLoader 0:  15%|█▌        | 92/602 [00:13<01:12,  6.99it/s][A

Validation DataLoader 0:  15%|█▌        | 93/602 [00:13<01:12,  7.06it/s][A

Validation DataLoader 0:  16%|█▌        | 94/602 [00:13<01:13,  6.95it/s][A

Validation DataLoader 0:  16%|█▌        | 95/602 [00:13<01:12,  7.01it/s][A

Validation DataLoader 0:  16%|█▌        | 96/602 [00:13<01:11,  7.08it/s][A

Validation DataLoader 0:  16%|█▌        | 97/602 [00:14<01:12,  6.93it/s][A

Validation DataLoader 0:  16%|█▋        | 98/602 [00:14<01:12,  6.99it/s][A

Validation DataLoader 0:  16%|█▋        | 99/602 [00:14<01:11,  7.05it/s][A

Validation DataLoader 0:  17%|█▋        | 100/602 [00:14<01:12,  6.93it/s][A

Validation DataLoader 0:  17%|█▋        | 101/602 [00:14<01:11,  6.99it/s][A

Validation DataLoader 0:  17%|█▋        | 102/602 [00:14<01:10,  7.05it/s][A

Validation DataLoader 0:  17%|█▋        | 103/602 [00:14<01:12,  6.93it/s][A

Validation DataLoader 0:  17%|█▋        | 104/602 [00:14<01:11,  6.99it/s][A

Validation DataLoader 0:  17%|█▋        | 105/602 [00:14<01:10,  7.05it/s][A

Validation DataLoader 0:  18%|█▊        | 106/602 [00:15<01:11,  6.94it/s][A

Validation DataLoader 0:  18%|█▊        | 107/602 [00:15<01:10,  7.00it/s][A

Validation DataLoader 0:  18%|█▊        | 108/602 [00:15<01:09,  7.06it/s][A

Validation DataLoader 0:  18%|█▊        | 109/602 [00:15<01:11,  6.90it/s][A

Validation DataLoader 0:  18%|█▊        | 110/602 [00:15<01:10,  6.96it/s][A

Validation DataLoader 0:  18%|█▊        | 111/602 [00:15<01:09,  7.02it/s][A

Validation DataLoader 0:  19%|█▊        | 112/602 [00:16<01:11,  6.84it/s][A

Validation DataLoader 0:  19%|█▉        | 113/602 [00:16<01:10,  6.90it/s][A

Validation DataLoader 0:  19%|█▉        | 114/602 [00:16<01:10,  6.95it/s][A

Validation DataLoader 0:  19%|█▉        | 115/602 [00:16<01:11,  6.85it/s][A

Validation DataLoader 0:  19%|█▉        | 116/602 [00:16<01:10,  6.90it/s][A

Validation DataLoader 0:  19%|█▉        | 117/602 [00:16<01:09,  6.95it/s][A

Validation DataLoader 0:  20%|█▉        | 118/602 [00:17<01:11,  6.81it/s][A

Validation DataLoader 0:  20%|█▉        | 119/602 [00:17<01:10,  6.86it/s][A

Validation DataLoader 0:  20%|█▉        | 120/602 [00:17<01:09,  6.91it/s][A

Validation DataLoader 0:  20%|██        | 121/602 [00:17<01:10,  6.80it/s][A

Validation DataLoader 0:  20%|██        | 122/602 [00:17<01:10,  6.85it/s][A

Validation DataLoader 0:  20%|██        | 123/602 [00:17<01:09,  6.90it/s][A

Validation DataLoader 0:  21%|██        | 124/602 [00:18<01:10,  6.80it/s][A

Validation DataLoader 0:  21%|██        | 125/602 [00:18<01:09,  6.85it/s][A

Validation DataLoader 0:  21%|██        | 126/602 [00:18<01:08,  6.90it/s][A

Validation DataLoader 0:  21%|██        | 127/602 [00:18<01:09,  6.79it/s][A

Validation DataLoader 0:  21%|██▏       | 128/602 [00:18<01:09,  6.84it/s][A

Validation DataLoader 0:  21%|██▏       | 129/602 [00:18<01:08,  6.88it/s][A

Validation DataLoader 0:  22%|██▏       | 130/602 [00:19<01:09,  6.80it/s][A

Validation DataLoader 0:  22%|██▏       | 131/602 [00:19<01:08,  6.85it/s][A

Validation DataLoader 0:  22%|██▏       | 132/602 [00:19<01:08,  6.90it/s][A

Validation DataLoader 0:  22%|██▏       | 133/602 [00:19<01:08,  6.80it/s][A

Validation DataLoader 0:  22%|██▏       | 134/602 [00:19<01:08,  6.85it/s][A

Validation DataLoader 0:  22%|██▏       | 135/602 [00:19<01:07,  6.89it/s][A

Validation DataLoader 0:  23%|██▎       | 136/602 [00:20<01:08,  6.80it/s][A

Validation DataLoader 0:  23%|██▎       | 137/602 [00:20<01:07,  6.84it/s][A

Validation DataLoader 0:  23%|██▎       | 138/602 [00:20<01:07,  6.89it/s][A

Validation DataLoader 0:  23%|██▎       | 139/602 [00:20<01:08,  6.80it/s][A

Validation DataLoader 0:  23%|██▎       | 140/602 [00:20<01:07,  6.84it/s][A

Validation DataLoader 0:  23%|██▎       | 141/602 [00:20<01:06,  6.89it/s][A

Validation DataLoader 0:  24%|██▎       | 142/602 [00:20<01:07,  6.81it/s][A

Validation DataLoader 0:  24%|██▍       | 143/602 [00:20<01:06,  6.85it/s][A

Validation DataLoader 0:  24%|██▍       | 144/602 [00:20<01:06,  6.90it/s][A

Validation DataLoader 0:  24%|██▍       | 145/602 [00:21<01:07,  6.82it/s][A

Validation DataLoader 0:  24%|██▍       | 146/602 [00:21<01:06,  6.86it/s][A

Validation DataLoader 0:  24%|██▍       | 147/602 [00:21<01:05,  6.91it/s][A

Validation DataLoader 0:  25%|██▍       | 148/602 [00:21<01:06,  6.80it/s][A

Validation DataLoader 0:  25%|██▍       | 149/602 [00:21<01:06,  6.84it/s][A

Validation DataLoader 0:  25%|██▍       | 150/602 [00:21<01:05,  6.88it/s][A

Validation DataLoader 0:  25%|██▌       | 151/602 [00:22<01:06,  6.80it/s][A

Validation DataLoader 0:  25%|██▌       | 152/602 [00:22<01:05,  6.84it/s][A

Validation DataLoader 0:  25%|██▌       | 153/602 [00:22<01:05,  6.88it/s][A

Validation DataLoader 0:  26%|██▌       | 154/602 [00:22<01:05,  6.80it/s][A

Validation DataLoader 0:  26%|██▌       | 155/602 [00:22<01:05,  6.84it/s][A

Validation DataLoader 0:  26%|██▌       | 156/602 [00:22<01:04,  6.88it/s][A

Validation DataLoader 0:  26%|██▌       | 157/602 [00:23<01:05,  6.79it/s][A

Validation DataLoader 0:  26%|██▌       | 158/602 [00:23<01:05,  6.83it/s][A

Validation DataLoader 0:  26%|██▋       | 159/602 [00:23<01:04,  6.87it/s][A

Validation DataLoader 0:  27%|██▋       | 160/602 [00:23<01:05,  6.79it/s][A

Validation DataLoader 0:  27%|██▋       | 161/602 [00:23<01:04,  6.83it/s][A

Validation DataLoader 0:  27%|██▋       | 162/602 [00:23<01:04,  6.87it/s][A

Validation DataLoader 0:  27%|██▋       | 163/602 [00:23<01:04,  6.82it/s][A

Validation DataLoader 0:  27%|██▋       | 164/602 [00:23<01:03,  6.85it/s][A

Validation DataLoader 0:  27%|██▋       | 165/602 [00:23<01:03,  6.89it/s][A

Validation DataLoader 0:  28%|██▊       | 166/602 [00:24<01:04,  6.81it/s][A

Validation DataLoader 0:  28%|██▊       | 167/602 [00:24<01:03,  6.85it/s][A

Validation DataLoader 0:  28%|██▊       | 168/602 [00:24<01:03,  6.89it/s][A

Validation DataLoader 0:  28%|██▊       | 169/602 [00:24<01:03,  6.80it/s][A

Validation DataLoader 0:  28%|██▊       | 170/602 [00:24<01:03,  6.84it/s][A

Validation DataLoader 0:  28%|██▊       | 171/602 [00:24<01:02,  6.88it/s][A

Validation DataLoader 0:  29%|██▊       | 172/602 [00:25<01:03,  6.80it/s][A

Validation DataLoader 0:  29%|██▊       | 173/602 [00:25<01:02,  6.83it/s][A

Validation DataLoader 0:  29%|██▉       | 174/602 [00:25<01:02,  6.86it/s][A

Validation DataLoader 0:  29%|██▉       | 175/602 [00:25<01:02,  6.80it/s][A

Validation DataLoader 0:  29%|██▉       | 176/602 [00:25<01:02,  6.84it/s][A

Validation DataLoader 0:  29%|██▉       | 177/602 [00:25<01:01,  6.87it/s][A

Validation DataLoader 0:  30%|██▉       | 178/602 [00:26<01:02,  6.80it/s][A

Validation DataLoader 0:  30%|██▉       | 179/602 [00:26<01:01,  6.83it/s][A

Validation DataLoader 0:  30%|██▉       | 180/602 [00:26<01:01,  6.87it/s][A

Validation DataLoader 0:  30%|███       | 181/602 [00:26<01:02,  6.79it/s][A

Validation DataLoader 0:  30%|███       | 182/602 [00:26<01:01,  6.81it/s][A

Validation DataLoader 0:  30%|███       | 183/602 [00:26<01:01,  6.85it/s][A

Validation DataLoader 0:  31%|███       | 184/602 [00:27<01:01,  6.79it/s][A

Validation DataLoader 0:  31%|███       | 185/602 [00:27<01:01,  6.82it/s][A

Validation DataLoader 0:  31%|███       | 186/602 [00:27<01:00,  6.86it/s][A

Validation DataLoader 0:  31%|███       | 187/602 [00:27<01:01,  6.80it/s][A

Validation DataLoader 0:  31%|███       | 188/602 [00:27<01:00,  6.84it/s][A

Validation DataLoader 0:  31%|███▏      | 189/602 [00:27<01:00,  6.87it/s][A

Validation DataLoader 0:  32%|███▏      | 190/602 [00:27<01:00,  6.80it/s][A

Validation DataLoader 0:  32%|███▏      | 191/602 [00:27<01:00,  6.83it/s][A

Validation DataLoader 0:  32%|███▏      | 192/602 [00:27<00:59,  6.87it/s][A

Validation DataLoader 0:  32%|███▏      | 193/602 [00:28<01:00,  6.80it/s][A

Validation DataLoader 0:  32%|███▏      | 194/602 [00:28<00:59,  6.83it/s][A

Validation DataLoader 0:  32%|███▏      | 195/602 [00:28<00:59,  6.86it/s][A

Validation DataLoader 0:  33%|███▎      | 196/602 [00:28<00:59,  6.80it/s][A

Validation DataLoader 0:  33%|███▎      | 197/602 [00:28<00:59,  6.82it/s][A

Validation DataLoader 0:  33%|███▎      | 198/602 [00:28<00:59,  6.85it/s][A

Validation DataLoader 0:  33%|███▎      | 199/602 [00:29<00:59,  6.80it/s][A

Validation DataLoader 0:  33%|███▎      | 200/602 [00:29<00:59,  6.80it/s][A

Validation DataLoader 0:  33%|███▎      | 201/602 [00:29<00:58,  6.83it/s][A

Validation DataLoader 0:  34%|███▎      | 202/602 [00:29<00:58,  6.79it/s][A

Validation DataLoader 0:  34%|███▎      | 203/602 [00:29<00:58,  6.80it/s][A

Validation DataLoader 0:  34%|███▍      | 204/602 [00:29<00:58,  6.83it/s][A

Validation DataLoader 0:  34%|███▍      | 205/602 [00:30<00:58,  6.79it/s][A

Validation DataLoader 0:  34%|███▍      | 206/602 [00:30<00:58,  6.80it/s][A

Validation DataLoader 0:  34%|███▍      | 207/602 [00:30<00:57,  6.83it/s][A

Validation DataLoader 0:  35%|███▍      | 208/602 [00:30<00:58,  6.77it/s][A

Validation DataLoader 0:  35%|███▍      | 209/602 [00:30<00:57,  6.80it/s][A

Validation DataLoader 0:  35%|███▍      | 210/602 [00:30<00:57,  6.83it/s][A

Validation DataLoader 0:  35%|███▌      | 211/602 [00:31<00:57,  6.77it/s][A

Validation DataLoader 0:  35%|███▌      | 212/602 [00:31<00:57,  6.80it/s][A

Validation DataLoader 0:  35%|███▌      | 213/602 [00:31<00:56,  6.83it/s][A

Validation DataLoader 0:  36%|███▌      | 214/602 [00:31<00:57,  6.77it/s][A

Validation DataLoader 0:  36%|███▌      | 215/602 [00:31<00:56,  6.80it/s][A

Validation DataLoader 0:  36%|███▌      | 216/602 [00:31<00:56,  6.83it/s][A

Validation DataLoader 0:  36%|███▌      | 217/602 [00:32<00:57,  6.75it/s][A

Validation DataLoader 0:  36%|███▌      | 218/602 [00:32<00:56,  6.78it/s][A

Validation DataLoader 0:  36%|███▋      | 219/602 [00:32<00:56,  6.81it/s][A

Validation DataLoader 0:  37%|███▋      | 220/602 [00:32<00:56,  6.75it/s][A

Validation DataLoader 0:  37%|███▋      | 221/602 [00:32<00:56,  6.78it/s][A

Validation DataLoader 0:  37%|███▋      | 222/602 [00:32<00:55,  6.81it/s][A

Validation DataLoader 0:  37%|███▋      | 223/602 [00:33<00:56,  6.74it/s][A

Validation DataLoader 0:  37%|███▋      | 224/602 [00:33<00:55,  6.77it/s][A

Validation DataLoader 0:  37%|███▋      | 225/602 [00:33<00:55,  6.80it/s][A

Validation DataLoader 0:  38%|███▊      | 226/602 [00:33<00:55,  6.74it/s][A

Validation DataLoader 0:  38%|███▊      | 227/602 [00:33<00:55,  6.77it/s][A

Validation DataLoader 0:  38%|███▊      | 228/602 [00:33<00:55,  6.80it/s][A

Validation DataLoader 0:  38%|███▊      | 229/602 [00:33<00:55,  6.75it/s][A

Validation DataLoader 0:  38%|███▊      | 230/602 [00:33<00:54,  6.77it/s][A

Validation DataLoader 0:  38%|███▊      | 231/602 [00:33<00:54,  6.80it/s][A

Validation DataLoader 0:  39%|███▊      | 232/602 [00:34<00:54,  6.75it/s][A

Validation DataLoader 0:  39%|███▊      | 233/602 [00:34<00:54,  6.78it/s][A

Validation DataLoader 0:  39%|███▉      | 234/602 [00:34<00:54,  6.80it/s][A

Validation DataLoader 0:  39%|███▉      | 235/602 [00:34<00:54,  6.74it/s][A

Validation DataLoader 0:  39%|███▉      | 236/602 [00:34<00:54,  6.77it/s][A

Validation DataLoader 0:  39%|███▉      | 237/602 [00:34<00:53,  6.80it/s][A

Validation DataLoader 0:  40%|███▉      | 238/602 [00:35<00:54,  6.73it/s][A

Validation DataLoader 0:  40%|███▉      | 239/602 [00:35<00:53,  6.76it/s][A

Validation DataLoader 0:  40%|███▉      | 240/602 [00:35<00:53,  6.78it/s][A

Validation DataLoader 0:  40%|████      | 241/602 [00:35<00:53,  6.75it/s][A

Validation DataLoader 0:  40%|████      | 242/602 [00:35<00:53,  6.78it/s][A

Validation DataLoader 0:  40%|████      | 243/602 [00:35<00:52,  6.80it/s][A

Validation DataLoader 0:  41%|████      | 244/602 [00:36<00:52,  6.76it/s][A

Validation DataLoader 0:  41%|████      | 245/602 [00:36<00:52,  6.79it/s][A

Validation DataLoader 0:  41%|████      | 246/602 [00:36<00:52,  6.81it/s][A

Validation DataLoader 0:  41%|████      | 247/602 [00:36<00:52,  6.76it/s][A

Validation DataLoader 0:  41%|████      | 248/602 [00:36<00:52,  6.77it/s][A

Validation DataLoader 0:  41%|████▏     | 249/602 [00:36<00:51,  6.80it/s][A

Validation DataLoader 0:  42%|████▏     | 250/602 [00:36<00:51,  6.77it/s][A

Validation DataLoader 0:  42%|████▏     | 251/602 [00:37<00:51,  6.78it/s][A

Validation DataLoader 0:  42%|████▏     | 252/602 [00:37<00:51,  6.80it/s][A

Validation DataLoader 0:  42%|████▏     | 253/602 [00:37<00:51,  6.77it/s][A

Validation DataLoader 0:  42%|████▏     | 254/602 [00:37<00:51,  6.78it/s][A

Validation DataLoader 0:  42%|████▏     | 255/602 [00:37<00:51,  6.80it/s][A

Validation DataLoader 0:  43%|████▎     | 256/602 [00:37<00:51,  6.77it/s][A

Validation DataLoader 0:  43%|████▎     | 257/602 [00:37<00:50,  6.78it/s][A

Validation DataLoader 0:  43%|████▎     | 258/602 [00:37<00:50,  6.80it/s][A

Validation DataLoader 0:  43%|████▎     | 259/602 [00:38<00:50,  6.78it/s][A

Validation DataLoader 0:  43%|████▎     | 260/602 [00:38<00:50,  6.77it/s][A

Validation DataLoader 0:  43%|████▎     | 261/602 [00:38<00:50,  6.80it/s][A

Validation DataLoader 0:  44%|████▎     | 262/602 [00:38<00:50,  6.78it/s][A

Validation DataLoader 0:  44%|████▎     | 263/602 [00:38<00:50,  6.78it/s][A

Validation DataLoader 0:  44%|████▍     | 264/602 [00:38<00:49,  6.80it/s][A

Validation DataLoader 0:  44%|████▍     | 265/602 [00:39<00:49,  6.76it/s][A

Validation DataLoader 0:  44%|████▍     | 266/602 [00:39<00:49,  6.79it/s][A

Validation DataLoader 0:  44%|████▍     | 267/602 [00:39<00:49,  6.81it/s][A

Validation DataLoader 0:  45%|████▍     | 268/602 [00:39<00:49,  6.78it/s][A

Validation DataLoader 0:  45%|████▍     | 269/602 [00:39<00:49,  6.77it/s][A

Validation DataLoader 0:  45%|████▍     | 270/602 [00:39<00:48,  6.79it/s][A

Validation DataLoader 0:  45%|████▌     | 271/602 [00:39<00:48,  6.78it/s][A

Validation DataLoader 0:  45%|████▌     | 272/602 [00:40<00:48,  6.77it/s][A

Validation DataLoader 0:  45%|████▌     | 273/602 [00:40<00:48,  6.79it/s][A

Validation DataLoader 0:  46%|████▌     | 274/602 [00:40<00:48,  6.77it/s][A

Validation DataLoader 0:  46%|████▌     | 275/602 [00:40<00:48,  6.77it/s][A

Validation DataLoader 0:  46%|████▌     | 276/602 [00:40<00:47,  6.79it/s][A

Validation DataLoader 0:  46%|████▌     | 277/602 [00:40<00:48,  6.76it/s][A

Validation DataLoader 0:  46%|████▌     | 278/602 [00:41<00:47,  6.77it/s][A

Validation DataLoader 0:  46%|████▋     | 279/602 [00:41<00:47,  6.79it/s][A

Validation DataLoader 0:  47%|████▋     | 280/602 [00:41<00:47,  6.76it/s][A

Validation DataLoader 0:  47%|████▋     | 281/602 [00:41<00:47,  6.77it/s][A

Validation DataLoader 0:  47%|████▋     | 282/602 [00:41<00:47,  6.79it/s][A

Validation DataLoader 0:  47%|████▋     | 283/602 [00:41<00:47,  6.77it/s][A

Validation DataLoader 0:  47%|████▋     | 284/602 [00:42<00:47,  6.76it/s][A

Validation DataLoader 0:  47%|████▋     | 285/602 [00:42<00:46,  6.78it/s][A

Validation DataLoader 0:  48%|████▊     | 286/602 [00:42<00:46,  6.78it/s][A

Validation DataLoader 0:  48%|████▊     | 287/602 [00:42<00:46,  6.76it/s][A

Validation DataLoader 0:  48%|████▊     | 288/602 [00:42<00:46,  6.78it/s][A

Validation DataLoader 0:  48%|████▊     | 289/602 [00:42<00:46,  6.78it/s][A

Validation DataLoader 0:  48%|████▊     | 290/602 [00:42<00:46,  6.75it/s][A

Validation DataLoader 0:  48%|████▊     | 291/602 [00:43<00:45,  6.77it/s][A

Validation DataLoader 0:  49%|████▊     | 292/602 [00:43<00:45,  6.79it/s][A

Validation DataLoader 0:  49%|████▊     | 293/602 [00:43<00:45,  6.74it/s][A

Validation DataLoader 0:  49%|████▉     | 294/602 [00:43<00:45,  6.76it/s][A

Validation DataLoader 0:  49%|████▉     | 295/602 [00:43<00:45,  6.78it/s][A

Validation DataLoader 0:  49%|████▉     | 296/602 [00:43<00:45,  6.74it/s][A

Validation DataLoader 0:  49%|████▉     | 297/602 [00:43<00:45,  6.76it/s][A

Validation DataLoader 0:  50%|████▉     | 298/602 [00:43<00:44,  6.78it/s][A

Validation DataLoader 0:  50%|████▉     | 299/602 [00:44<00:44,  6.73it/s][A

Validation DataLoader 0:  50%|████▉     | 300/602 [00:44<00:44,  6.75it/s][A

Validation DataLoader 0:  50%|█████     | 301/602 [00:44<00:44,  6.77it/s][A

Validation DataLoader 0:  50%|█████     | 302/602 [00:44<00:44,  6.73it/s][A

Validation DataLoader 0:  50%|█████     | 303/602 [00:44<00:44,  6.75it/s][A

Validation DataLoader 0:  50%|█████     | 304/602 [00:44<00:44,  6.77it/s][A

Validation DataLoader 0:  51%|█████     | 305/602 [00:45<00:44,  6.73it/s][A

Validation DataLoader 0:  51%|█████     | 306/602 [00:45<00:43,  6.75it/s][A

Validation DataLoader 0:  51%|█████     | 307/602 [00:45<00:43,  6.77it/s][A

Validation DataLoader 0:  51%|█████     | 308/602 [00:45<00:43,  6.73it/s][A

Validation DataLoader 0:  51%|█████▏    | 309/602 [00:45<00:43,  6.75it/s][A

Validation DataLoader 0:  51%|█████▏    | 310/602 [00:45<00:43,  6.77it/s][A

Validation DataLoader 0:  52%|█████▏    | 311/602 [00:46<00:43,  6.73it/s][A

Validation DataLoader 0:  52%|█████▏    | 312/602 [00:46<00:42,  6.75it/s][A

Validation DataLoader 0:  52%|█████▏    | 313/602 [00:46<00:42,  6.77it/s][A

Validation DataLoader 0:  52%|█████▏    | 314/602 [00:46<00:42,  6.73it/s][A

Validation DataLoader 0:  52%|█████▏    | 315/602 [00:46<00:42,  6.75it/s][A

Validation DataLoader 0:  52%|█████▏    | 316/602 [00:46<00:42,  6.77it/s][A

Validation DataLoader 0:  53%|█████▎    | 317/602 [00:47<00:42,  6.72it/s][A

Validation DataLoader 0:  53%|█████▎    | 318/602 [00:47<00:42,  6.73it/s][A

Validation DataLoader 0:  53%|█████▎    | 319/602 [00:47<00:41,  6.75it/s][A

Validation DataLoader 0:  53%|█████▎    | 320/602 [00:47<00:42,  6.70it/s][A

Validation DataLoader 0:  53%|█████▎    | 321/602 [00:47<00:41,  6.72it/s][A

Validation DataLoader 0:  53%|█████▎    | 322/602 [00:47<00:41,  6.74it/s][A

Validation DataLoader 0:  54%|█████▎    | 323/602 [00:48<00:41,  6.70it/s][A

Validation DataLoader 0:  54%|█████▍    | 324/602 [00:48<00:41,  6.72it/s][A

Validation DataLoader 0:  54%|█████▍    | 325/602 [00:48<00:41,  6.74it/s][A

Validation DataLoader 0:  54%|█████▍    | 326/602 [00:48<00:41,  6.70it/s][A

Validation DataLoader 0:  54%|█████▍    | 327/602 [00:48<00:40,  6.72it/s][A

Validation DataLoader 0:  54%|█████▍    | 328/602 [00:48<00:40,  6.74it/s][A

Validation DataLoader 0:  55%|█████▍    | 329/602 [00:49<00:40,  6.70it/s][A

Validation DataLoader 0:  55%|█████▍    | 330/602 [00:49<00:40,  6.72it/s][A

Validation DataLoader 0:  55%|█████▍    | 331/602 [00:49<00:40,  6.74it/s][A

Validation DataLoader 0:  55%|█████▌    | 332/602 [00:49<00:40,  6.71it/s][A

Validation DataLoader 0:  55%|█████▌    | 333/602 [00:49<00:39,  6.73it/s][A

Validation DataLoader 0:  55%|█████▌    | 334/602 [00:49<00:39,  6.74it/s][A

Validation DataLoader 0:  56%|█████▌    | 335/602 [00:49<00:39,  6.71it/s][A

Validation DataLoader 0:  56%|█████▌    | 336/602 [00:49<00:39,  6.73it/s][A

Validation DataLoader 0:  56%|█████▌    | 337/602 [00:49<00:39,  6.75it/s][A

Validation DataLoader 0:  56%|█████▌    | 338/602 [00:50<00:39,  6.72it/s][A

Validation DataLoader 0:  56%|█████▋    | 339/602 [00:50<00:39,  6.73it/s][A

Validation DataLoader 0:  56%|█████▋    | 340/602 [00:50<00:38,  6.75it/s][A

Validation DataLoader 0:  57%|█████▋    | 341/602 [00:50<00:38,  6.72it/s][A

Validation DataLoader 0:  57%|█████▋    | 342/602 [00:50<00:38,  6.74it/s][A

Validation DataLoader 0:  57%|█████▋    | 343/602 [00:50<00:38,  6.76it/s][A

Validation DataLoader 0:  57%|█████▋    | 344/602 [00:51<00:38,  6.73it/s][A

Validation DataLoader 0:  57%|█████▋    | 345/602 [00:51<00:38,  6.75it/s][A

Validation DataLoader 0:  57%|█████▋    | 346/602 [00:51<00:37,  6.76it/s][A

Validation DataLoader 0:  58%|█████▊    | 347/602 [00:51<00:37,  6.73it/s][A

Validation DataLoader 0:  58%|█████▊    | 348/602 [00:51<00:37,  6.75it/s][A

Validation DataLoader 0:  58%|█████▊    | 349/602 [00:51<00:37,  6.77it/s][A

Validation DataLoader 0:  58%|█████▊    | 350/602 [00:51<00:37,  6.74it/s][A

Validation DataLoader 0:  58%|█████▊    | 351/602 [00:52<00:37,  6.74it/s][A

Validation DataLoader 0:  58%|█████▊    | 352/602 [00:52<00:37,  6.76it/s][A

Validation DataLoader 0:  59%|█████▊    | 353/602 [00:52<00:36,  6.74it/s][A

Validation DataLoader 0:  59%|█████▉    | 354/602 [00:52<00:36,  6.74it/s][A

Validation DataLoader 0:  59%|█████▉    | 355/602 [00:52<00:36,  6.75it/s][A

Validation DataLoader 0:  59%|█████▉    | 356/602 [00:52<00:36,  6.75it/s][A

Validation DataLoader 0:  59%|█████▉    | 357/602 [00:52<00:36,  6.74it/s][A

Validation DataLoader 0:  59%|█████▉    | 358/602 [00:52<00:36,  6.76it/s][A

Validation DataLoader 0:  60%|█████▉    | 359/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|█████▉    | 360/602 [00:53<00:35,  6.74it/s][A

Validation DataLoader 0:  60%|█████▉    | 361/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|██████    | 362/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  60%|██████    | 363/602 [00:53<00:35,  6.75it/s][A

Validation DataLoader 0:  60%|██████    | 364/602 [00:53<00:35,  6.76it/s][A

Validation DataLoader 0:  61%|██████    | 365/602 [00:53<00:35,  6.77it/s][A

Validation DataLoader 0:  61%|██████    | 366/602 [00:54<00:35,  6.74it/s][A

Validation DataLoader 0:  61%|██████    | 367/602 [00:54<00:34,  6.76it/s][A

Validation DataLoader 0:  61%|██████    | 368/602 [00:54<00:34,  6.77it/s][A

Validation DataLoader 0:  61%|██████▏   | 369/602 [00:54<00:34,  6.75it/s][A

Validation DataLoader 0:  61%|██████▏   | 370/602 [00:54<00:34,  6.76it/s][A

Validation DataLoader 0:  62%|██████▏   | 371/602 [00:54<00:34,  6.77it/s][A

Validation DataLoader 0:  62%|██████▏   | 372/602 [00:55<00:34,  6.75it/s][A

Validation DataLoader 0:  62%|██████▏   | 373/602 [00:55<00:33,  6.76it/s][A

Validation DataLoader 0:  62%|██████▏   | 374/602 [00:55<00:33,  6.77it/s][A

Validation DataLoader 0:  62%|██████▏   | 375/602 [00:55<00:33,  6.76it/s][A

Validation DataLoader 0:  62%|██████▏   | 376/602 [00:55<00:33,  6.77it/s][A

Validation DataLoader 0:  63%|██████▎   | 377/602 [00:55<00:33,  6.77it/s][A

Validation DataLoader 0:  63%|██████▎   | 378/602 [00:55<00:33,  6.76it/s][A

Validation DataLoader 0:  63%|██████▎   | 379/602 [00:55<00:32,  6.77it/s][A

Validation DataLoader 0:  63%|██████▎   | 380/602 [00:56<00:32,  6.77it/s][A

Validation DataLoader 0:  63%|██████▎   | 381/602 [00:56<00:32,  6.76it/s][A

Validation DataLoader 0:  63%|██████▎   | 382/602 [00:56<00:32,  6.77it/s][A

Validation DataLoader 0:  64%|██████▎   | 383/602 [00:56<00:32,  6.78it/s][A

Validation DataLoader 0:  64%|██████▍   | 384/602 [00:56<00:32,  6.76it/s][A

Validation DataLoader 0:  64%|██████▍   | 385/602 [00:56<00:32,  6.76it/s][A

Validation DataLoader 0:  64%|██████▍   | 386/602 [00:56<00:31,  6.78it/s][A

Validation DataLoader 0:  64%|██████▍   | 387/602 [00:57<00:31,  6.76it/s][A

Validation DataLoader 0:  64%|██████▍   | 388/602 [00:57<00:31,  6.76it/s][A

Validation DataLoader 0:  65%|██████▍   | 389/602 [00:57<00:31,  6.78it/s][A

Validation DataLoader 0:  65%|██████▍   | 390/602 [00:57<00:31,  6.77it/s][A

Validation DataLoader 0:  65%|██████▍   | 391/602 [00:57<00:31,  6.76it/s][A

Validation DataLoader 0:  65%|██████▌   | 392/602 [00:57<00:31,  6.77it/s][A

Validation DataLoader 0:  65%|██████▌   | 393/602 [00:58<00:30,  6.75it/s][A

Validation DataLoader 0:  65%|██████▌   | 394/602 [00:58<00:30,  6.76it/s][A

Validation DataLoader 0:  66%|██████▌   | 395/602 [00:58<00:30,  6.77it/s][A

Validation DataLoader 0:  66%|██████▌   | 396/602 [00:58<00:30,  6.75it/s][A

Validation DataLoader 0:  66%|██████▌   | 397/602 [00:58<00:30,  6.76it/s][A

Validation DataLoader 0:  66%|██████▌   | 398/602 [00:58<00:30,  6.78it/s][A

Validation DataLoader 0:  66%|██████▋   | 399/602 [00:59<00:30,  6.76it/s][A

Validation DataLoader 0:  66%|██████▋   | 400/602 [00:59<00:29,  6.77it/s][A

Validation DataLoader 0:  67%|██████▋   | 401/602 [00:59<00:29,  6.79it/s][A

Validation DataLoader 0:  67%|██████▋   | 402/602 [00:59<00:29,  6.76it/s][A

Validation DataLoader 0:  67%|██████▋   | 403/602 [00:59<00:29,  6.77it/s][A

Validation DataLoader 0:  67%|██████▋   | 404/602 [00:59<00:29,  6.79it/s][A

Validation DataLoader 0:  67%|██████▋   | 405/602 [00:59<00:29,  6.75it/s][A

Validation DataLoader 0:  67%|██████▋   | 406/602 [00:59<00:28,  6.77it/s][A

Validation DataLoader 0:  68%|██████▊   | 407/602 [01:00<00:28,  6.78it/s][A

Validation DataLoader 0:  68%|██████▊   | 408/602 [01:00<00:28,  6.74it/s][A

Validation DataLoader 0:  68%|██████▊   | 409/602 [01:00<00:28,  6.75it/s][A

Validation DataLoader 0:  68%|██████▊   | 410/602 [01:00<00:28,  6.77it/s][A

Validation DataLoader 0:  68%|██████▊   | 411/602 [01:01<00:28,  6.73it/s][A

Validation DataLoader 0:  68%|██████▊   | 412/602 [01:01<00:28,  6.75it/s][A

Validation DataLoader 0:  69%|██████▊   | 413/602 [01:01<00:27,  6.76it/s][A

Validation DataLoader 0:  69%|██████▉   | 414/602 [01:01<00:27,  6.74it/s][A

Validation DataLoader 0:  69%|██████▉   | 415/602 [01:01<00:27,  6.75it/s][A

Validation DataLoader 0:  69%|██████▉   | 416/602 [01:01<00:27,  6.76it/s][A

Validation DataLoader 0:  69%|██████▉   | 417/602 [01:01<00:27,  6.73it/s][A

Validation DataLoader 0:  69%|██████▉   | 418/602 [01:01<00:27,  6.75it/s][A

Validation DataLoader 0:  70%|██████▉   | 419/602 [01:01<00:27,  6.76it/s][A

Validation DataLoader 0:  70%|██████▉   | 420/602 [01:02<00:27,  6.74it/s][A

Validation DataLoader 0:  70%|██████▉   | 421/602 [01:02<00:26,  6.75it/s][A

Validation DataLoader 0:  70%|███████   | 422/602 [01:02<00:26,  6.77it/s][A

Validation DataLoader 0:  70%|███████   | 423/602 [01:02<00:26,  6.73it/s][A

Validation DataLoader 0:  70%|███████   | 424/602 [01:02<00:26,  6.75it/s][A

Validation DataLoader 0:  71%|███████   | 425/602 [01:02<00:26,  6.76it/s][A

Validation DataLoader 0:  71%|███████   | 426/602 [01:03<00:26,  6.74it/s][A

Validation DataLoader 0:  71%|███████   | 427/602 [01:03<00:25,  6.75it/s][A

Validation DataLoader 0:  71%|███████   | 428/602 [01:03<00:25,  6.76it/s][A

Validation DataLoader 0:  71%|███████▏  | 429/602 [01:03<00:25,  6.73it/s][A

Validation DataLoader 0:  71%|███████▏  | 430/602 [01:03<00:25,  6.75it/s][A

Validation DataLoader 0:  72%|███████▏  | 431/602 [01:03<00:25,  6.76it/s][A

Validation DataLoader 0:  72%|███████▏  | 432/602 [01:04<00:25,  6.73it/s][A

Validation DataLoader 0:  72%|███████▏  | 433/602 [01:04<00:25,  6.75it/s][A

Validation DataLoader 0:  72%|███████▏  | 434/602 [01:04<00:24,  6.76it/s][A

Validation DataLoader 0:  72%|███████▏  | 435/602 [01:04<00:24,  6.74it/s][A

Validation DataLoader 0:  72%|███████▏  | 436/602 [01:04<00:24,  6.76it/s][A

Validation DataLoader 0:  73%|███████▎  | 437/602 [01:04<00:24,  6.77it/s][A

Validation DataLoader 0:  73%|███████▎  | 438/602 [01:04<00:24,  6.75it/s][A

Validation DataLoader 0:  73%|███████▎  | 439/602 [01:04<00:24,  6.76it/s][A

Validation DataLoader 0:  73%|███████▎  | 440/602 [01:04<00:23,  6.78it/s][A

Validation DataLoader 0:  73%|███████▎  | 441/602 [01:05<00:23,  6.74it/s][A

Validation DataLoader 0:  73%|███████▎  | 442/602 [01:05<00:23,  6.76it/s][A

Validation DataLoader 0:  74%|███████▎  | 443/602 [01:05<00:23,  6.77it/s][A

Validation DataLoader 0:  74%|███████▍  | 444/602 [01:05<00:23,  6.74it/s][A

Validation DataLoader 0:  74%|███████▍  | 445/602 [01:05<00:23,  6.75it/s][A

Validation DataLoader 0:  74%|███████▍  | 446/602 [01:05<00:23,  6.77it/s][A

Validation DataLoader 0:  74%|███████▍  | 447/602 [01:06<00:22,  6.74it/s][A

Validation DataLoader 0:  74%|███████▍  | 448/602 [01:06<00:22,  6.75it/s][A

Validation DataLoader 0:  75%|███████▍  | 449/602 [01:06<00:22,  6.77it/s][A

Validation DataLoader 0:  75%|███████▍  | 450/602 [01:06<00:22,  6.74it/s][A

Validation DataLoader 0:  75%|███████▍  | 451/602 [01:06<00:22,  6.75it/s][A

Validation DataLoader 0:  75%|███████▌  | 452/602 [01:06<00:22,  6.77it/s][A

Validation DataLoader 0:  75%|███████▌  | 453/602 [01:07<00:22,  6.74it/s][A

Validation DataLoader 0:  75%|███████▌  | 454/602 [01:07<00:21,  6.76it/s][A

Validation DataLoader 0:  76%|███████▌  | 455/602 [01:07<00:21,  6.77it/s][A

Validation DataLoader 0:  76%|███████▌  | 456/602 [01:07<00:21,  6.74it/s][A

Validation DataLoader 0:  76%|███████▌  | 457/602 [01:07<00:21,  6.76it/s][A

Validation DataLoader 0:  76%|███████▌  | 458/602 [01:07<00:21,  6.77it/s][A

Validation DataLoader 0:  76%|███████▌  | 459/602 [01:08<00:21,  6.74it/s][A

Validation DataLoader 0:  76%|███████▋  | 460/602 [01:08<00:21,  6.76it/s][A

Validation DataLoader 0:  77%|███████▋  | 461/602 [01:08<00:20,  6.77it/s][A

Validation DataLoader 0:  77%|███████▋  | 462/602 [01:08<00:20,  6.74it/s][A

Validation DataLoader 0:  77%|███████▋  | 463/602 [01:08<00:20,  6.75it/s][A

Validation DataLoader 0:  77%|███████▋  | 464/602 [01:08<00:20,  6.76it/s][A

Validation DataLoader 0:  77%|███████▋  | 465/602 [01:08<00:20,  6.74it/s][A

Validation DataLoader 0:  77%|███████▋  | 466/602 [01:09<00:20,  6.75it/s][A

Validation DataLoader 0:  78%|███████▊  | 467/602 [01:09<00:19,  6.77it/s][A

Validation DataLoader 0:  78%|███████▊  | 468/602 [01:09<00:19,  6.73it/s][A

Validation DataLoader 0:  78%|███████▊  | 469/602 [01:09<00:19,  6.75it/s][A

Validation DataLoader 0:  78%|███████▊  | 470/602 [01:09<00:19,  6.76it/s][A

Validation DataLoader 0:  78%|███████▊  | 471/602 [01:09<00:19,  6.74it/s][A

Validation DataLoader 0:  78%|███████▊  | 472/602 [01:09<00:19,  6.75it/s][A

Validation DataLoader 0:  79%|███████▊  | 473/602 [01:09<00:19,  6.76it/s][A

Validation DataLoader 0:  79%|███████▊  | 474/602 [01:10<00:19,  6.74it/s][A

Validation DataLoader 0:  79%|███████▉  | 475/602 [01:10<00:18,  6.75it/s][A

Validation DataLoader 0:  79%|███████▉  | 476/602 [01:10<00:18,  6.76it/s][A

Validation DataLoader 0:  79%|███████▉  | 477/602 [01:10<00:18,  6.74it/s][A

Validation DataLoader 0:  79%|███████▉  | 478/602 [01:10<00:18,  6.75it/s][A

Validation DataLoader 0:  80%|███████▉  | 479/602 [01:10<00:18,  6.76it/s][A

Validation DataLoader 0:  80%|███████▉  | 480/602 [01:11<00:18,  6.73it/s][A

Validation DataLoader 0:  80%|███████▉  | 481/602 [01:11<00:17,  6.75it/s][A

Validation DataLoader 0:  80%|████████  | 482/602 [01:11<00:17,  6.76it/s][A

Validation DataLoader 0:  80%|████████  | 483/602 [01:11<00:17,  6.73it/s][A

Validation DataLoader 0:  80%|████████  | 484/602 [01:11<00:17,  6.75it/s][A

Validation DataLoader 0:  81%|████████  | 485/602 [01:11<00:17,  6.76it/s][A

Validation DataLoader 0:  81%|████████  | 486/602 [01:12<00:17,  6.74it/s][A

Validation DataLoader 0:  81%|████████  | 487/602 [01:12<00:17,  6.75it/s][A

Validation DataLoader 0:  81%|████████  | 488/602 [01:12<00:16,  6.76it/s][A

Validation DataLoader 0:  81%|████████  | 489/602 [01:12<00:16,  6.73it/s][A

Validation DataLoader 0:  81%|████████▏ | 490/602 [01:12<00:16,  6.74it/s][A

Validation DataLoader 0:  82%|████████▏ | 491/602 [01:12<00:16,  6.76it/s][A

Validation DataLoader 0:  82%|████████▏ | 492/602 [01:13<00:16,  6.73it/s][A

Validation DataLoader 0:  82%|████████▏ | 493/602 [01:13<00:16,  6.75it/s][A

Validation DataLoader 0:  82%|████████▏ | 494/602 [01:13<00:15,  6.76it/s][A

Validation DataLoader 0:  82%|████████▏ | 495/602 [01:13<00:15,  6.74it/s][A

Validation DataLoader 0:  82%|████████▏ | 496/602 [01:13<00:15,  6.75it/s][A

Validation DataLoader 0:  83%|████████▎ | 497/602 [01:13<00:15,  6.77it/s][A

Validation DataLoader 0:  83%|████████▎ | 498/602 [01:13<00:15,  6.74it/s][A

Validation DataLoader 0:  83%|████████▎ | 499/602 [01:13<00:15,  6.75it/s][A

Validation DataLoader 0:  83%|████████▎ | 500/602 [01:13<00:15,  6.77it/s][A

Validation DataLoader 0:  83%|████████▎ | 501/602 [01:14<00:14,  6.75it/s][A

Validation DataLoader 0:  83%|████████▎ | 502/602 [01:14<00:14,  6.76it/s][A

Validation DataLoader 0:  84%|████████▎ | 503/602 [01:14<00:14,  6.77it/s][A

Validation DataLoader 0:  84%|████████▎ | 504/602 [01:14<00:14,  6.75it/s][A

Validation DataLoader 0:  84%|████████▍ | 505/602 [01:14<00:14,  6.76it/s][A

Validation DataLoader 0:  84%|████████▍ | 506/602 [01:14<00:14,  6.77it/s][A

Validation DataLoader 0:  84%|████████▍ | 507/602 [01:15<00:14,  6.75it/s][A

Validation DataLoader 0:  84%|████████▍ | 508/602 [01:15<00:13,  6.76it/s][A

Validation DataLoader 0:  85%|████████▍ | 509/602 [01:15<00:13,  6.77it/s][A

Validation DataLoader 0:  85%|████████▍ | 510/602 [01:15<00:13,  6.75it/s][A

Validation DataLoader 0:  85%|████████▍ | 511/602 [01:15<00:13,  6.76it/s][A

Validation DataLoader 0:  85%|████████▌ | 512/602 [01:15<00:13,  6.77it/s][A

Validation DataLoader 0:  85%|████████▌ | 513/602 [01:15<00:13,  6.76it/s][A

Validation DataLoader 0:  85%|████████▌ | 514/602 [01:15<00:13,  6.76it/s][A

Validation DataLoader 0:  86%|████████▌ | 515/602 [01:16<00:12,  6.78it/s][A

Validation DataLoader 0:  86%|████████▌ | 516/602 [01:16<00:12,  6.75it/s][A

Validation DataLoader 0:  86%|████████▌ | 517/602 [01:16<00:12,  6.76it/s][A

Validation DataLoader 0:  86%|████████▌ | 518/602 [01:16<00:12,  6.77it/s][A

Validation DataLoader 0:  86%|████████▌ | 519/602 [01:16<00:12,  6.74it/s][A

Validation DataLoader 0:  86%|████████▋ | 520/602 [01:16<00:12,  6.75it/s][A

Validation DataLoader 0:  87%|████████▋ | 521/602 [01:17<00:11,  6.77it/s][A

Validation DataLoader 0:  87%|████████▋ | 522/602 [01:17<00:11,  6.75it/s][A

Validation DataLoader 0:  87%|████████▋ | 523/602 [01:17<00:11,  6.76it/s][A

Validation DataLoader 0:  87%|████████▋ | 524/602 [01:17<00:11,  6.77it/s][A

Validation DataLoader 0:  87%|████████▋ | 525/602 [01:17<00:11,  6.74it/s][A

Validation DataLoader 0:  87%|████████▋ | 526/602 [01:17<00:11,  6.75it/s][A

Validation DataLoader 0:  88%|████████▊ | 527/602 [01:17<00:11,  6.77it/s][A

Validation DataLoader 0:  88%|████████▊ | 528/602 [01:18<00:10,  6.74it/s][A

Validation DataLoader 0:  88%|████████▊ | 529/602 [01:18<00:10,  6.75it/s][A

Validation DataLoader 0:  88%|████████▊ | 530/602 [01:18<00:10,  6.77it/s][A

Validation DataLoader 0:  88%|████████▊ | 531/602 [01:18<00:10,  6.74it/s][A

Validation DataLoader 0:  88%|████████▊ | 532/602 [01:18<00:10,  6.76it/s][A

Validation DataLoader 0:  89%|████████▊ | 533/602 [01:18<00:10,  6.77it/s][A

Validation DataLoader 0:  89%|████████▊ | 534/602 [01:19<00:10,  6.75it/s][A

Validation DataLoader 0:  89%|████████▉ | 535/602 [01:19<00:09,  6.76it/s][A

Validation DataLoader 0:  89%|████████▉ | 536/602 [01:19<00:09,  6.77it/s][A

Validation DataLoader 0:  89%|████████▉ | 537/602 [01:19<00:09,  6.74it/s][A

Validation DataLoader 0:  89%|████████▉ | 538/602 [01:19<00:09,  6.75it/s][A

Validation DataLoader 0:  90%|████████▉ | 539/602 [01:19<00:09,  6.76it/s][A

Validation DataLoader 0:  90%|████████▉ | 540/602 [01:20<00:09,  6.74it/s][A

Validation DataLoader 0:  90%|████████▉ | 541/602 [01:20<00:09,  6.75it/s][A

Validation DataLoader 0:  90%|█████████ | 542/602 [01:20<00:08,  6.76it/s][A

Validation DataLoader 0:  90%|█████████ | 543/602 [01:20<00:08,  6.74it/s][A

Validation DataLoader 0:  90%|█████████ | 544/602 [01:20<00:08,  6.75it/s][A

Validation DataLoader 0:  91%|█████████ | 545/602 [01:20<00:08,  6.76it/s][A

Validation DataLoader 0:  91%|█████████ | 546/602 [01:20<00:08,  6.74it/s][A

Validation DataLoader 0:  91%|█████████ | 547/602 [01:20<00:08,  6.75it/s][A

Validation DataLoader 0:  91%|█████████ | 548/602 [01:21<00:07,  6.76it/s][A

Validation DataLoader 0:  91%|█████████ | 549/602 [01:21<00:07,  6.74it/s][A

Validation DataLoader 0:  91%|█████████▏| 550/602 [01:21<00:07,  6.75it/s][A

Validation DataLoader 0:  92%|█████████▏| 551/602 [01:21<00:07,  6.76it/s][A

Validation DataLoader 0:  92%|█████████▏| 552/602 [01:21<00:07,  6.74it/s][A

Validation DataLoader 0:  92%|█████████▏| 553/602 [01:21<00:07,  6.75it/s][A

Validation DataLoader 0:  92%|█████████▏| 554/602 [01:21<00:07,  6.76it/s][A

Validation DataLoader 0:  92%|█████████▏| 555/602 [01:22<00:06,  6.75it/s][A

Validation DataLoader 0:  92%|█████████▏| 556/602 [01:22<00:06,  6.76it/s][A

Validation DataLoader 0:  93%|█████████▎| 557/602 [01:22<00:06,  6.77it/s][A

Validation DataLoader 0:  93%|█████████▎| 558/602 [01:22<00:06,  6.75it/s][A

Validation DataLoader 0:  93%|█████████▎| 559/602 [01:22<00:06,  6.76it/s][A

Validation DataLoader 0:  93%|█████████▎| 560/602 [01:22<00:06,  6.77it/s][A

Validation DataLoader 0:  93%|█████████▎| 561/602 [01:23<00:06,  6.75it/s][A

Validation DataLoader 0:  93%|█████████▎| 562/602 [01:23<00:05,  6.75it/s][A

Validation DataLoader 0:  94%|█████████▎| 563/602 [01:23<00:05,  6.76it/s][A

Validation DataLoader 0:  94%|█████████▎| 564/602 [01:23<00:05,  6.74it/s][A

Validation DataLoader 0:  94%|█████████▍| 565/602 [01:23<00:05,  6.74it/s][A

Validation DataLoader 0:  94%|█████████▍| 566/602 [01:23<00:05,  6.76it/s][A

Validation DataLoader 0:  94%|█████████▍| 567/602 [01:24<00:05,  6.74it/s][A

Validation DataLoader 0:  94%|█████████▍| 568/602 [01:24<00:05,  6.75it/s][A

Validation DataLoader 0:  95%|█████████▍| 569/602 [01:24<00:04,  6.76it/s][A

Validation DataLoader 0:  95%|█████████▍| 570/602 [01:24<00:04,  6.74it/s][A

Validation DataLoader 0:  95%|█████████▍| 571/602 [01:24<00:04,  6.74it/s][A

Validation DataLoader 0:  95%|█████████▌| 572/602 [01:24<00:04,  6.75it/s][A

Validation DataLoader 0:  95%|█████████▌| 573/602 [01:25<00:04,  6.74it/s][A

Validation DataLoader 0:  95%|█████████▌| 574/602 [01:25<00:04,  6.74it/s][A

Validation DataLoader 0:  96%|█████████▌| 575/602 [01:25<00:03,  6.75it/s][A

Validation DataLoader 0:  96%|█████████▌| 576/602 [01:25<00:03,  6.74it/s][A

Validation DataLoader 0:  96%|█████████▌| 577/602 [01:25<00:03,  6.74it/s][A

Validation DataLoader 0:  96%|█████████▌| 578/602 [01:25<00:03,  6.75it/s][A

Validation DataLoader 0:  96%|█████████▌| 579/602 [01:26<00:03,  6.73it/s][A

Validation DataLoader 0:  96%|█████████▋| 580/602 [01:26<00:03,  6.74it/s][A

Validation DataLoader 0:  97%|█████████▋| 581/602 [01:26<00:03,  6.75it/s][A

Validation DataLoader 0:  97%|█████████▋| 582/602 [01:26<00:02,  6.74it/s][A

Validation DataLoader 0:  97%|█████████▋| 583/602 [01:26<00:02,  6.74it/s][A

Validation DataLoader 0:  97%|█████████▋| 584/602 [01:26<00:02,  6.75it/s][A

Validation DataLoader 0:  97%|█████████▋| 585/602 [01:26<00:02,  6.73it/s][A

Validation DataLoader 0:  97%|█████████▋| 586/602 [01:26<00:02,  6.74it/s][A

Validation DataLoader 0:  98%|█████████▊| 587/602 [01:26<00:02,  6.75it/s][A

Validation DataLoader 0:  98%|█████████▊| 588/602 [01:27<00:02,  6.73it/s][A

Validation DataLoader 0:  98%|█████████▊| 589/602 [01:27<00:01,  6.74it/s][A

Validation DataLoader 0:  98%|█████████▊| 590/602 [01:27<00:01,  6.75it/s][A

Validation DataLoader 0:  98%|█████████▊| 591/602 [01:27<00:01,  6.73it/s][A

Validation DataLoader 0:  98%|█████████▊| 592/602 [01:27<00:01,  6.74it/s][A

Validation DataLoader 0:  99%|█████████▊| 593/602 [01:27<00:01,  6.75it/s][A

Validation DataLoader 0:  99%|█████████▊| 594/602 [01:28<00:01,  6.73it/s][A

Validation DataLoader 0:  99%|█████████▉| 595/602 [01:28<00:01,  6.74it/s][A

Validation DataLoader 0:  99%|█████████▉| 596/602 [01:28<00:00,  6.75it/s][A

Validation DataLoader 0:  99%|█████████▉| 597/602 [01:28<00:00,  6.73it/s][A

Validation DataLoader 0:  99%|█████████▉| 598/602 [01:28<00:00,  6.74it/s][A

Validation DataLoader 0: 100%|█████████▉| 599/602 [01:28<00:00,  6.75it/s][A

Validation DataLoader 0: 100%|█████████▉| 600/602 [01:29<00:00,  6.73it/s][A

Validation DataLoader 0: 100%|█████████▉| 601/602 [01:29<00:00,  6.74it/s][A

Validation DataLoader 0: 100%|██████████| 602/602 [01:29<00:00,  6.75it/s][A

                                                                          [A
Epoch 2: 100%|██████████| 3295/3295 [35:55<00:00,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.400, R@1 (train)=0.902]
Epoch 2: 100%|██████████| 3295/3295 [35:55<00:00,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]Epoch 2, global step 9885: 'R@1 (val)' reached 0.40328 (best 0.40328), saving model to '/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=02-R@1 (val)=0.403.ckpt' as top 1

Epoch 2:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]           
Epoch 3:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 1/3295 [00:02<1:51:21,  0.49it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 1/3295 [00:02<1:51:22,  0.49it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 2/3295 [00:02<56:19,  0.97it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]  
Epoch 3:   0%|          | 2/3295 [00:02<56:20,  0.97it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 3/3295 [00:02<38:04,  1.44it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 3/3295 [00:02<38:05,  1.44it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 4/3295 [00:03<53:46,  1.02it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 4/3295 [00:03<53:46,  1.02it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 5/3295 [00:03<43:19,  1.27it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 5/3295 [00:03<43:19,  1.27it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 6/3295 [00:03<36:21,  1.51it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 6/3295 [00:03<36:21,  1.51it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 7/3295 [00:05<45:33,  1.20it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 7/3295 [00:05<45:33,  1.20it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 8/3295 [00:05<40:02,  1.37it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 8/3295 [00:05<40:02,  1.37it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 9/3295 [00:05<35:45,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 9/3295 [00:05<35:45,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 10/3295 [00:07<41:58,  1.30it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 10/3295 [00:07<41:58,  1.30it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 11/3295 [00:07<38:16,  1.43it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 11/3295 [00:07<38:17,  1.43it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 12/3295 [00:07<35:11,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 12/3295 [00:07<35:12,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 13/3295 [00:09<39:15,  1.39it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 13/3295 [00:09<39:15,  1.39it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 14/3295 [00:09<36:33,  1.50it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 14/3295 [00:09<36:33,  1.50it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 15/3295 [00:09<34:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 15/3295 [00:09<34:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 16/3295 [00:11<38:24,  1.42it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   0%|          | 16/3295 [00:11<38:24,  1.42it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 17/3295 [00:11<36:14,  1.51it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 17/3295 [00:11<36:14,  1.51it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 18/3295 [00:11<34:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 18/3295 [00:11<34:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 19/3295 [00:13<37:41,  1.45it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 19/3295 [00:13<37:41,  1.45it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 20/3295 [00:13<35:52,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 20/3295 [00:13<35:52,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 21/3295 [00:13<34:14,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 21/3295 [00:13<34:14,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 22/3295 [00:15<37:17,  1.46it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 22/3295 [00:15<37:17,  1.46it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 23/3295 [00:15<35:44,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 23/3295 [00:15<35:44,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 24/3295 [00:15<34:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 24/3295 [00:15<34:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 25/3295 [00:17<37:06,  1.47it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 25/3295 [00:17<37:06,  1.47it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 26/3295 [00:17<35:43,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 26/3295 [00:17<35:43,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 27/3295 [00:17<34:27,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 27/3295 [00:17<34:27,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 28/3295 [00:18<36:48,  1.48it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 28/3295 [00:18<36:48,  1.48it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 29/3295 [00:18<35:35,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 29/3295 [00:18<35:35,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 30/3295 [00:18<34:26,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 30/3295 [00:18<34:26,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 31/3295 [00:20<36:26,  1.49it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 31/3295 [00:20<36:26,  1.49it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 32/3295 [00:20<35:20,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 32/3295 [00:20<35:20,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 33/3295 [00:20<34:18,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 33/3295 [00:20<34:18,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 34/3295 [00:22<36:23,  1.49it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 34/3295 [00:22<36:23,  1.49it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 35/3295 [00:22<35:23,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 35/3295 [00:22<35:23,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 36/3295 [00:22<34:25,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 36/3295 [00:22<34:25,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 37/3295 [00:24<36:07,  1.50it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 37/3295 [00:24<36:07,  1.50it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 38/3295 [00:24<35:11,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 38/3295 [00:24<35:11,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 39/3295 [00:24<34:19,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 39/3295 [00:24<34:19,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 40/3295 [00:26<35:50,  1.51it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 40/3295 [00:26<35:50,  1.51it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 41/3295 [00:26<34:59,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|          | 41/3295 [00:26<34:59,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 42/3295 [00:26<34:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 42/3295 [00:26<34:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 43/3295 [00:28<35:37,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 43/3295 [00:28<35:37,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 44/3295 [00:28<34:50,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 44/3295 [00:28<34:50,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 45/3295 [00:28<34:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 45/3295 [00:28<34:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 46/3295 [00:30<35:31,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 46/3295 [00:30<35:31,  1.52it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 47/3295 [00:30<34:47,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 47/3295 [00:30<34:47,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 48/3295 [00:30<34:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 48/3295 [00:30<34:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 49/3295 [00:31<35:13,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   1%|▏         | 49/3295 [00:31<35:13,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 50/3295 [00:31<34:32,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 50/3295 [00:31<34:32,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 51/3295 [00:31<33:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 51/3295 [00:31<33:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 52/3295 [00:33<35:14,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 52/3295 [00:33<35:14,  1.53it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 53/3295 [00:33<34:35,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 53/3295 [00:33<34:35,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 54/3295 [00:33<33:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 54/3295 [00:33<33:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 55/3295 [00:35<35:07,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 55/3295 [00:35<35:07,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 56/3295 [00:35<34:30,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 56/3295 [00:35<34:30,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 57/3295 [00:35<33:55,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 57/3295 [00:35<33:55,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 58/3295 [00:37<35:05,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 58/3295 [00:37<35:05,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 59/3295 [00:37<34:30,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 59/3295 [00:37<34:30,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 60/3295 [00:37<33:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 60/3295 [00:37<33:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 61/3295 [00:39<34:58,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 61/3295 [00:39<34:58,  1.54it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 62/3295 [00:39<34:25,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 62/3295 [00:39<34:25,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 63/3295 [00:39<33:53,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 63/3295 [00:39<33:53,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 64/3295 [00:41<34:49,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 64/3295 [00:41<34:49,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 65/3295 [00:41<34:17,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 65/3295 [00:41<34:17,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 66/3295 [00:41<33:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 66/3295 [00:41<33:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 67/3295 [00:43<34:36,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 67/3295 [00:43<34:36,  1.55it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 68/3295 [00:43<34:06,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 68/3295 [00:43<34:06,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 69/3295 [00:43<33:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 69/3295 [00:43<33:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 70/3295 [00:45<34:33,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 70/3295 [00:45<34:33,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 71/3295 [00:45<34:05,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 71/3295 [00:45<34:05,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 72/3295 [00:45<33:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 72/3295 [00:45<33:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 73/3295 [00:46<34:31,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 73/3295 [00:46<34:31,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 74/3295 [00:46<34:04,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 74/3295 [00:46<34:04,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 75/3295 [00:47<33:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 75/3295 [00:47<33:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 76/3295 [00:48<34:28,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 76/3295 [00:48<34:28,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 77/3295 [00:48<34:02,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 77/3295 [00:48<34:02,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 78/3295 [00:48<33:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 78/3295 [00:48<33:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 79/3295 [00:50<34:22,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 79/3295 [00:50<34:22,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 80/3295 [00:50<33:57,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 80/3295 [00:50<33:57,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 81/3295 [00:50<33:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 81/3295 [00:50<33:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 82/3295 [00:52<34:12,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   2%|▏         | 82/3295 [00:52<34:12,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 83/3295 [00:52<33:48,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 83/3295 [00:52<33:48,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 84/3295 [00:52<33:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 84/3295 [00:52<33:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 85/3295 [00:54<34:12,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 85/3295 [00:54<34:12,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 86/3295 [00:54<33:48,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 86/3295 [00:54<33:48,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 87/3295 [00:54<33:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 87/3295 [00:54<33:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 88/3295 [00:56<34:08,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 88/3295 [00:56<34:08,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 89/3295 [00:56<33:45,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 89/3295 [00:56<33:45,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 90/3295 [00:56<33:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 90/3295 [00:56<33:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 91/3295 [00:57<34:00,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 91/3295 [00:57<34:00,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 92/3295 [00:57<33:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 92/3295 [00:57<33:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 93/3295 [00:58<33:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 93/3295 [00:58<33:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 94/3295 [01:00<34:03,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 94/3295 [01:00<34:03,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 95/3295 [01:00<33:42,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 95/3295 [01:00<33:42,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 96/3295 [01:00<33:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 96/3295 [01:00<33:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 97/3295 [01:02<34:07,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 97/3295 [01:02<34:07,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 98/3295 [01:02<33:46,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 98/3295 [01:02<33:46,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 99/3295 [01:02<33:26,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 99/3295 [01:02<33:26,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 100/3295 [01:03<34:02,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 100/3295 [01:03<34:02,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 101/3295 [01:03<33:41,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 101/3295 [01:03<33:42,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 102/3295 [01:03<33:22,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 102/3295 [01:03<33:22,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 103/3295 [01:05<34:00,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 103/3295 [01:05<34:00,  1.56it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 104/3295 [01:05<33:41,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 104/3295 [01:05<33:41,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 105/3295 [01:05<33:22,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 105/3295 [01:05<33:22,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 106/3295 [01:07<33:53,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 106/3295 [01:07<33:53,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 107/3295 [01:07<33:35,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 107/3295 [01:07<33:35,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 108/3295 [01:07<33:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 108/3295 [01:07<33:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 109/3295 [01:09<33:50,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 109/3295 [01:09<33:50,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 110/3295 [01:09<33:32,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 110/3295 [01:09<33:32,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 111/3295 [01:09<33:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 111/3295 [01:09<33:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 112/3295 [01:11<33:47,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 112/3295 [01:11<33:47,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 113/3295 [01:11<33:29,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 113/3295 [01:11<33:29,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 114/3295 [01:11<33:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 114/3295 [01:11<33:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 115/3295 [01:13<33:43,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   3%|▎         | 115/3295 [01:13<33:43,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 116/3295 [01:13<33:26,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 116/3295 [01:13<33:26,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 117/3295 [01:13<33:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 117/3295 [01:13<33:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 118/3295 [01:14<33:38,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 118/3295 [01:14<33:38,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 119/3295 [01:14<33:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 119/3295 [01:14<33:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 120/3295 [01:15<33:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 120/3295 [01:15<33:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 121/3295 [01:16<33:36,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 121/3295 [01:16<33:36,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 122/3295 [01:16<33:19,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 122/3295 [01:16<33:19,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 123/3295 [01:16<33:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▎         | 123/3295 [01:16<33:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 124/3295 [01:18<33:34,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 124/3295 [01:18<33:34,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 125/3295 [01:18<33:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 125/3295 [01:18<33:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 126/3295 [01:18<33:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 126/3295 [01:18<33:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 127/3295 [01:20<33:32,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 127/3295 [01:20<33:32,  1.57it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 128/3295 [01:20<33:17,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 128/3295 [01:20<33:17,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 129/3295 [01:20<33:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 129/3295 [01:20<33:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 130/3295 [01:22<33:28,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 130/3295 [01:22<33:28,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 131/3295 [01:22<33:13,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 131/3295 [01:22<33:13,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 132/3295 [01:22<32:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 132/3295 [01:22<32:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 133/3295 [01:24<33:26,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 133/3295 [01:24<33:26,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 134/3295 [01:24<33:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 134/3295 [01:24<33:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 135/3295 [01:24<32:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 135/3295 [01:24<32:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 136/3295 [01:26<33:17,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 136/3295 [01:26<33:17,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 137/3295 [01:26<33:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 137/3295 [01:26<33:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 138/3295 [01:26<32:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 138/3295 [01:26<32:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 139/3295 [01:27<33:16,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 139/3295 [01:27<33:16,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 140/3295 [01:27<33:02,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 140/3295 [01:27<33:02,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 141/3295 [01:27<32:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 141/3295 [01:27<32:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 142/3295 [01:29<33:14,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 142/3295 [01:29<33:14,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 143/3295 [01:29<33:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 143/3295 [01:29<33:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 144/3295 [01:29<32:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 144/3295 [01:29<32:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 145/3295 [01:31<33:08,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 145/3295 [01:31<33:08,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 146/3295 [01:31<32:55,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 146/3295 [01:31<32:55,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 147/3295 [01:31<32:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 147/3295 [01:31<32:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 148/3295 [01:33<33:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   4%|▍         | 148/3295 [01:33<33:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 149/3295 [01:33<32:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 149/3295 [01:33<32:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 150/3295 [01:33<32:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 150/3295 [01:33<32:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 151/3295 [01:35<32:58,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 151/3295 [01:35<32:58,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 152/3295 [01:35<32:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 152/3295 [01:35<32:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 153/3295 [01:35<32:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 153/3295 [01:35<32:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 154/3295 [01:36<32:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 154/3295 [01:36<32:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 155/3295 [01:36<32:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 155/3295 [01:36<32:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 156/3295 [01:37<32:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 156/3295 [01:37<32:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 157/3295 [01:38<32:53,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 157/3295 [01:38<32:53,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 158/3295 [01:38<32:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 158/3295 [01:38<32:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 159/3295 [01:38<32:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 159/3295 [01:38<32:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 160/3295 [01:40<32:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 160/3295 [01:40<32:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 161/3295 [01:40<32:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 161/3295 [01:40<32:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 162/3295 [01:40<32:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 162/3295 [01:40<32:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 163/3295 [01:42<32:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 163/3295 [01:42<32:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 164/3295 [01:42<32:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▍         | 164/3295 [01:42<32:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 165/3295 [01:42<32:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 165/3295 [01:42<32:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 166/3295 [01:44<32:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 166/3295 [01:44<32:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 167/3295 [01:44<32:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 167/3295 [01:44<32:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 168/3295 [01:44<32:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 168/3295 [01:44<32:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 169/3295 [01:46<32:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 169/3295 [01:46<32:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 170/3295 [01:46<32:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 170/3295 [01:46<32:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 171/3295 [01:46<32:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 171/3295 [01:46<32:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 172/3295 [01:48<32:45,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 172/3295 [01:48<32:45,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 173/3295 [01:48<32:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 173/3295 [01:48<32:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 174/3295 [01:48<32:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 174/3295 [01:48<32:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 175/3295 [01:50<32:43,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 175/3295 [01:50<32:43,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 176/3295 [01:50<32:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 176/3295 [01:50<32:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 177/3295 [01:50<32:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 177/3295 [01:50<32:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 178/3295 [01:52<32:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 178/3295 [01:52<32:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 179/3295 [01:52<32:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 179/3295 [01:52<32:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 180/3295 [01:52<32:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 180/3295 [01:52<32:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 181/3295 [01:54<32:41,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   5%|▌         | 181/3295 [01:54<32:41,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 182/3295 [01:54<32:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 182/3295 [01:54<32:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 183/3295 [01:54<32:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 183/3295 [01:54<32:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 184/3295 [01:55<32:38,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 184/3295 [01:55<32:38,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 185/3295 [01:55<32:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 185/3295 [01:55<32:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 186/3295 [01:55<32:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 186/3295 [01:55<32:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 187/3295 [01:57<32:37,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 187/3295 [01:57<32:37,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 188/3295 [01:57<32:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 188/3295 [01:57<32:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 189/3295 [01:57<32:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 189/3295 [01:57<32:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 190/3295 [01:59<32:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 190/3295 [01:59<32:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 191/3295 [01:59<32:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 191/3295 [01:59<32:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 192/3295 [01:59<32:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 192/3295 [01:59<32:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 193/3295 [02:01<32:35,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 193/3295 [02:01<32:35,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 194/3295 [02:01<32:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 194/3295 [02:01<32:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 195/3295 [02:01<32:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 195/3295 [02:01<32:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 196/3295 [02:03<32:34,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 196/3295 [02:03<32:34,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 197/3295 [02:03<32:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 197/3295 [02:03<32:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 198/3295 [02:03<32:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 198/3295 [02:03<32:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 199/3295 [02:05<32:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 199/3295 [02:05<32:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 200/3295 [02:05<32:22,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 200/3295 [02:05<32:22,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 201/3295 [02:05<32:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 201/3295 [02:05<32:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 202/3295 [02:07<32:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 202/3295 [02:07<32:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 203/3295 [02:07<32:19,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 203/3295 [02:07<32:19,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 204/3295 [02:07<32:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 204/3295 [02:07<32:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 205/3295 [02:09<32:28,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▌         | 205/3295 [02:09<32:28,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 206/3295 [02:09<32:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 206/3295 [02:09<32:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 207/3295 [02:09<32:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 207/3295 [02:09<32:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 208/3295 [02:11<32:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 208/3295 [02:11<32:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 209/3295 [02:11<32:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 209/3295 [02:11<32:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 210/3295 [02:11<32:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 210/3295 [02:11<32:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 211/3295 [02:13<32:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 211/3295 [02:13<32:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 212/3295 [02:13<32:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 212/3295 [02:13<32:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 213/3295 [02:13<32:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 213/3295 [02:13<32:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 214/3295 [02:15<32:24,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   6%|▋         | 214/3295 [02:15<32:24,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 215/3295 [02:15<32:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 215/3295 [02:15<32:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 216/3295 [02:15<32:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 216/3295 [02:15<32:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 217/3295 [02:16<32:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 217/3295 [02:16<32:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 218/3295 [02:16<32:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 218/3295 [02:16<32:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 219/3295 [02:16<32:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 219/3295 [02:16<32:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 220/3295 [02:18<32:20,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 220/3295 [02:18<32:20,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 221/3295 [02:18<32:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 221/3295 [02:18<32:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 222/3295 [02:18<32:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 222/3295 [02:18<32:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 223/3295 [02:20<32:18,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 223/3295 [02:20<32:18,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 224/3295 [02:20<32:09,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 224/3295 [02:20<32:09,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 225/3295 [02:20<32:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 225/3295 [02:20<32:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 226/3295 [02:22<32:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 226/3295 [02:22<32:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 227/3295 [02:22<32:07,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 227/3295 [02:22<32:07,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 228/3295 [02:22<31:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 228/3295 [02:22<31:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 229/3295 [02:24<32:14,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 229/3295 [02:24<32:14,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 230/3295 [02:24<32:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 230/3295 [02:24<32:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 231/3295 [02:24<31:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 231/3295 [02:24<31:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 232/3295 [02:26<32:09,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 232/3295 [02:26<32:09,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 233/3295 [02:26<32:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 233/3295 [02:26<32:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 234/3295 [02:26<31:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 234/3295 [02:26<31:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 235/3295 [02:27<32:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 235/3295 [02:27<32:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 236/3295 [02:27<31:58,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 236/3295 [02:27<31:58,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 237/3295 [02:28<31:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 237/3295 [02:28<31:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 238/3295 [02:29<32:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 238/3295 [02:29<32:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 239/3295 [02:29<31:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 239/3295 [02:29<31:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 240/3295 [02:29<31:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 240/3295 [02:29<31:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 241/3295 [02:31<32:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 241/3295 [02:31<32:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 242/3295 [02:31<31:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 242/3295 [02:31<31:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 243/3295 [02:31<31:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 243/3295 [02:31<31:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 244/3295 [02:33<32:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 244/3295 [02:33<32:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 245/3295 [02:33<31:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 245/3295 [02:33<31:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 246/3295 [02:33<31:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 246/3295 [02:33<31:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 247/3295 [02:35<31:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   7%|▋         | 247/3295 [02:35<31:57,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 248/3295 [02:35<31:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 248/3295 [02:35<31:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 249/3295 [02:35<31:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 249/3295 [02:35<31:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 250/3295 [02:37<31:56,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 250/3295 [02:37<31:56,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 251/3295 [02:37<31:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 251/3295 [02:37<31:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 252/3295 [02:37<31:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 252/3295 [02:37<31:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 253/3295 [02:39<31:55,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 253/3295 [02:39<31:55,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 254/3295 [02:39<31:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 254/3295 [02:39<31:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 255/3295 [02:39<31:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 255/3295 [02:39<31:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 256/3295 [02:41<31:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 256/3295 [02:41<31:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 257/3295 [02:41<31:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 257/3295 [02:41<31:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 258/3295 [02:41<31:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 258/3295 [02:41<31:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 259/3295 [02:42<31:50,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 259/3295 [02:42<31:50,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 260/3295 [02:42<31:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 260/3295 [02:42<31:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 261/3295 [02:43<31:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 261/3295 [02:43<31:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 262/3295 [02:44<31:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 262/3295 [02:44<31:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 263/3295 [02:44<31:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 263/3295 [02:44<31:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 264/3295 [02:44<31:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 264/3295 [02:44<31:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 265/3295 [02:47<31:50,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 265/3295 [02:47<31:50,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 266/3295 [02:47<31:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 266/3295 [02:47<31:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 267/3295 [02:47<31:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 267/3295 [02:47<31:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 268/3295 [02:49<31:49,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 268/3295 [02:49<31:49,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 269/3295 [02:49<31:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 269/3295 [02:49<31:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 270/3295 [02:49<31:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 270/3295 [02:49<31:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 271/3295 [02:51<31:48,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 271/3295 [02:51<31:48,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 272/3295 [02:51<31:40,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 272/3295 [02:51<31:40,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 273/3295 [02:51<31:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 273/3295 [02:51<31:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 274/3295 [02:52<31:46,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 274/3295 [02:52<31:46,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 275/3295 [02:52<31:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 275/3295 [02:52<31:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 276/3295 [02:52<31:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 276/3295 [02:52<31:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 277/3295 [02:54<31:46,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 277/3295 [02:54<31:46,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 278/3295 [02:54<31:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 278/3295 [02:54<31:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 279/3295 [02:55<31:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 279/3295 [02:55<31:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 280/3295 [02:56<31:43,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   8%|▊         | 280/3295 [02:56<31:43,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 281/3295 [02:56<31:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 281/3295 [02:56<31:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 282/3295 [02:56<31:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 282/3295 [02:56<31:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 283/3295 [02:58<31:43,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 283/3295 [02:58<31:43,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 284/3295 [02:58<31:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 284/3295 [02:58<31:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 285/3295 [02:58<31:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 285/3295 [02:58<31:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 286/3295 [03:00<31:41,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 286/3295 [03:00<31:41,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 287/3295 [03:00<31:34,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 287/3295 [03:00<31:34,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 288/3295 [03:00<31:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▊         | 288/3295 [03:00<31:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 289/3295 [03:02<31:37,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 289/3295 [03:02<31:37,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 290/3295 [03:02<31:30,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 290/3295 [03:02<31:30,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 291/3295 [03:02<31:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 291/3295 [03:02<31:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 292/3295 [03:04<31:34,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 292/3295 [03:04<31:34,  1.58it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 293/3295 [03:04<31:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 293/3295 [03:04<31:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 294/3295 [03:04<31:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 294/3295 [03:04<31:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 295/3295 [03:06<31:32,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 295/3295 [03:06<31:32,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 296/3295 [03:06<31:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 296/3295 [03:06<31:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 297/3295 [03:06<31:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 297/3295 [03:06<31:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 298/3295 [03:07<31:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 298/3295 [03:07<31:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 299/3295 [03:07<31:23,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 299/3295 [03:07<31:23,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 300/3295 [03:07<31:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 300/3295 [03:07<31:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 301/3295 [03:09<31:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 301/3295 [03:09<31:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 302/3295 [03:09<31:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 302/3295 [03:09<31:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 303/3295 [03:09<31:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 303/3295 [03:09<31:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 304/3295 [03:11<31:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 304/3295 [03:11<31:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 305/3295 [03:11<31:17,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 305/3295 [03:11<31:17,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 306/3295 [03:11<31:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 306/3295 [03:11<31:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 307/3295 [03:13<31:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 307/3295 [03:13<31:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 308/3295 [03:13<31:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 308/3295 [03:13<31:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 309/3295 [03:13<31:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 309/3295 [03:13<31:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 310/3295 [03:15<31:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 310/3295 [03:15<31:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 311/3295 [03:15<31:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 311/3295 [03:15<31:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 312/3295 [03:15<31:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 312/3295 [03:15<31:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 313/3295 [03:16<31:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:   9%|▉         | 313/3295 [03:16<31:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 314/3295 [03:17<31:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 314/3295 [03:17<31:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 315/3295 [03:17<31:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 315/3295 [03:17<31:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 316/3295 [03:18<31:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 316/3295 [03:18<31:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 317/3295 [03:18<31:09,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 317/3295 [03:18<31:09,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 318/3295 [03:18<31:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 318/3295 [03:18<31:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 319/3295 [03:20<31:14,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 319/3295 [03:20<31:14,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 320/3295 [03:20<31:08,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 320/3295 [03:20<31:08,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 321/3295 [03:20<31:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 321/3295 [03:20<31:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 322/3295 [03:22<31:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 322/3295 [03:22<31:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 323/3295 [03:22<31:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 323/3295 [03:22<31:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 324/3295 [03:22<31:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 324/3295 [03:22<31:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 325/3295 [03:24<31:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 325/3295 [03:24<31:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 326/3295 [03:24<31:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 326/3295 [03:24<31:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 327/3295 [03:24<30:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 327/3295 [03:24<30:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 328/3295 [03:26<31:07,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 328/3295 [03:26<31:07,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 329/3295 [03:26<31:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|▉         | 329/3295 [03:26<31:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 330/3295 [03:26<30:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 330/3295 [03:26<30:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 331/3295 [03:28<31:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 331/3295 [03:28<31:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 332/3295 [03:28<30:59,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 332/3295 [03:28<30:59,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 333/3295 [03:28<30:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 333/3295 [03:28<30:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 334/3295 [03:30<31:02,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 334/3295 [03:30<31:02,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 335/3295 [03:30<30:56,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 335/3295 [03:30<30:56,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 336/3295 [03:30<30:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 336/3295 [03:30<30:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 337/3295 [03:31<31:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 337/3295 [03:31<31:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 338/3295 [03:31<30:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 338/3295 [03:31<30:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 339/3295 [03:31<30:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 339/3295 [03:31<30:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 340/3295 [03:33<30:59,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 340/3295 [03:33<30:59,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 341/3295 [03:33<30:53,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 341/3295 [03:33<30:53,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 342/3295 [03:33<30:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 342/3295 [03:33<30:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 343/3295 [03:35<30:56,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 343/3295 [03:35<30:56,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 344/3295 [03:35<30:50,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 344/3295 [03:35<30:50,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 345/3295 [03:35<30:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  10%|█         | 345/3295 [03:35<30:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 346/3295 [03:37<30:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 346/3295 [03:37<30:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 347/3295 [03:37<30:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 347/3295 [03:37<30:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 348/3295 [03:37<30:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 348/3295 [03:37<30:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 349/3295 [03:39<30:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 349/3295 [03:39<30:51,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 350/3295 [03:39<30:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 350/3295 [03:39<30:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 351/3295 [03:39<30:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 351/3295 [03:39<30:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 352/3295 [03:41<30:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 352/3295 [03:41<30:48,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 353/3295 [03:41<30:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 353/3295 [03:41<30:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 354/3295 [03:41<30:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 354/3295 [03:41<30:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 355/3295 [03:43<30:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 355/3295 [03:43<30:47,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 356/3295 [03:43<30:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 356/3295 [03:43<30:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 357/3295 [03:43<30:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 357/3295 [03:43<30:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 358/3295 [03:44<30:45,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 358/3295 [03:44<30:45,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 359/3295 [03:44<30:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 359/3295 [03:44<30:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 360/3295 [03:45<30:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 360/3295 [03:45<30:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 361/3295 [03:46<30:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 361/3295 [03:46<30:42,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 362/3295 [03:46<30:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 362/3295 [03:46<30:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 363/3295 [03:46<30:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 363/3295 [03:46<30:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 364/3295 [03:48<30:40,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 364/3295 [03:48<30:40,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 365/3295 [03:48<30:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 365/3295 [03:48<30:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 366/3295 [03:48<30:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 366/3295 [03:48<30:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 367/3295 [03:50<30:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 367/3295 [03:50<30:39,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 368/3295 [03:50<30:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 368/3295 [03:50<30:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 369/3295 [03:50<30:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 369/3295 [03:50<30:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 370/3295 [03:52<30:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█         | 370/3295 [03:52<30:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 371/3295 [03:52<30:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 371/3295 [03:52<30:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 372/3295 [03:52<30:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 372/3295 [03:52<30:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 373/3295 [03:54<30:33,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 373/3295 [03:54<30:33,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 374/3295 [03:54<30:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 374/3295 [03:54<30:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 375/3295 [03:54<30:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 375/3295 [03:54<30:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 376/3295 [03:55<30:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 376/3295 [03:55<30:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 377/3295 [03:55<30:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 377/3295 [03:55<30:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 378/3295 [03:55<30:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  11%|█▏        | 378/3295 [03:55<30:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 379/3295 [03:57<30:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 379/3295 [03:57<30:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 380/3295 [03:57<30:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 380/3295 [03:57<30:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 381/3295 [03:57<30:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 381/3295 [03:57<30:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 382/3295 [03:59<30:26,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 382/3295 [03:59<30:26,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 383/3295 [03:59<30:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 383/3295 [03:59<30:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 384/3295 [03:59<30:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 384/3295 [03:59<30:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 385/3295 [04:01<30:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 385/3295 [04:01<30:24,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 386/3295 [04:01<30:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 386/3295 [04:01<30:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 387/3295 [04:01<30:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 387/3295 [04:01<30:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 388/3295 [04:03<30:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 388/3295 [04:03<30:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 389/3295 [04:03<30:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 389/3295 [04:03<30:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 390/3295 [04:03<30:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 390/3295 [04:03<30:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 391/3295 [04:05<30:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 391/3295 [04:05<30:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 392/3295 [04:05<30:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 392/3295 [04:05<30:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 393/3295 [04:05<30:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 393/3295 [04:05<30:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 394/3295 [04:06<30:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 394/3295 [04:06<30:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 395/3295 [04:06<30:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 395/3295 [04:06<30:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 396/3295 [04:06<30:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 396/3295 [04:06<30:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 397/3295 [04:08<30:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 397/3295 [04:08<30:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 398/3295 [04:08<30:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 398/3295 [04:08<30:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 399/3295 [04:08<30:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 399/3295 [04:08<30:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 400/3295 [04:10<30:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 400/3295 [04:10<30:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 401/3295 [04:10<30:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 401/3295 [04:10<30:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 402/3295 [04:10<30:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 402/3295 [04:10<30:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 403/3295 [04:12<30:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 403/3295 [04:12<30:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 404/3295 [04:12<30:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 404/3295 [04:12<30:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 405/3295 [04:12<30:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 405/3295 [04:12<30:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 406/3295 [04:14<30:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 406/3295 [04:14<30:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 407/3295 [04:14<30:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 407/3295 [04:14<30:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 408/3295 [04:14<29:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 408/3295 [04:14<29:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 409/3295 [04:16<30:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 409/3295 [04:16<30:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 410/3295 [04:16<30:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 410/3295 [04:16<30:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 411/3295 [04:16<29:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  12%|█▏        | 411/3295 [04:16<29:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 412/3295 [04:17<30:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 412/3295 [04:17<30:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 413/3295 [04:17<29:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 413/3295 [04:17<29:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 414/3295 [04:17<29:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 414/3295 [04:17<29:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 415/3295 [04:19<30:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 415/3295 [04:19<30:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 416/3295 [04:19<29:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 416/3295 [04:19<29:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 417/3295 [04:19<29:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 417/3295 [04:19<29:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 418/3295 [04:21<30:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 418/3295 [04:21<30:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 419/3295 [04:21<29:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 419/3295 [04:21<29:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 420/3295 [04:21<29:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 420/3295 [04:21<29:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 421/3295 [04:23<30:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 421/3295 [04:23<30:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 422/3295 [04:23<29:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 422/3295 [04:23<29:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 423/3295 [04:23<29:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 423/3295 [04:23<29:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 424/3295 [04:25<29:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 424/3295 [04:25<29:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 425/3295 [04:25<29:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 425/3295 [04:25<29:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 426/3295 [04:25<29:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 426/3295 [04:25<29:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 427/3295 [04:27<29:58,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 427/3295 [04:27<29:58,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 428/3295 [04:27<29:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 428/3295 [04:27<29:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 429/3295 [04:27<29:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 429/3295 [04:27<29:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 430/3295 [04:29<29:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 430/3295 [04:29<29:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 431/3295 [04:29<29:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 431/3295 [04:29<29:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 432/3295 [04:29<29:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 432/3295 [04:29<29:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 433/3295 [04:31<29:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 433/3295 [04:31<29:54,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 434/3295 [04:31<29:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 434/3295 [04:31<29:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 435/3295 [04:31<29:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 435/3295 [04:31<29:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 436/3295 [04:33<29:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 436/3295 [04:33<29:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 437/3295 [04:33<29:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 437/3295 [04:33<29:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 438/3295 [04:33<29:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 438/3295 [04:33<29:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 439/3295 [04:35<29:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 439/3295 [04:35<29:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 440/3295 [04:35<29:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 440/3295 [04:35<29:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 441/3295 [04:35<29:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 441/3295 [04:35<29:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 442/3295 [04:37<29:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 442/3295 [04:37<29:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 443/3295 [04:37<29:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 443/3295 [04:37<29:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 444/3295 [04:37<29:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  13%|█▎        | 444/3295 [04:37<29:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 445/3295 [04:38<29:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 445/3295 [04:38<29:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 446/3295 [04:38<29:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 446/3295 [04:38<29:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 447/3295 [04:38<29:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 447/3295 [04:38<29:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 448/3295 [04:40<29:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 448/3295 [04:40<29:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 449/3295 [04:40<29:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 449/3295 [04:40<29:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 450/3295 [04:40<29:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 450/3295 [04:40<29:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 451/3295 [04:42<29:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 451/3295 [04:42<29:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 452/3295 [04:42<29:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 452/3295 [04:42<29:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 453/3295 [04:42<29:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▎        | 453/3295 [04:42<29:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 454/3295 [04:44<29:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 454/3295 [04:44<29:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 455/3295 [04:44<29:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 455/3295 [04:44<29:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 456/3295 [04:44<29:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 456/3295 [04:44<29:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 457/3295 [04:46<29:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 457/3295 [04:46<29:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 458/3295 [04:46<29:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 458/3295 [04:46<29:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 459/3295 [04:46<29:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 459/3295 [04:46<29:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 460/3295 [04:48<29:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 460/3295 [04:48<29:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 461/3295 [04:48<29:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 461/3295 [04:48<29:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 462/3295 [04:48<29:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 462/3295 [04:48<29:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 463/3295 [04:49<29:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 463/3295 [04:49<29:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 464/3295 [04:49<29:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 464/3295 [04:49<29:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 465/3295 [04:49<29:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 465/3295 [04:49<29:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 466/3295 [04:51<29:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 466/3295 [04:51<29:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 467/3295 [04:51<29:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 467/3295 [04:51<29:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 468/3295 [04:51<29:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 468/3295 [04:51<29:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 469/3295 [04:53<29:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 469/3295 [04:53<29:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 470/3295 [04:53<29:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 470/3295 [04:53<29:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 471/3295 [04:53<29:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 471/3295 [04:53<29:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 472/3295 [04:55<29:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 472/3295 [04:55<29:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 473/3295 [04:55<29:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 473/3295 [04:55<29:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 474/3295 [04:55<29:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 474/3295 [04:55<29:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 475/3295 [04:57<29:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 475/3295 [04:57<29:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 476/3295 [04:57<29:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 476/3295 [04:57<29:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 477/3295 [04:57<29:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  14%|█▍        | 477/3295 [04:57<29:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 478/3295 [04:58<29:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 478/3295 [04:58<29:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 479/3295 [04:58<29:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 479/3295 [04:58<29:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 480/3295 [04:59<29:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 480/3295 [04:59<29:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 481/3295 [05:00<29:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 481/3295 [05:00<29:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 482/3295 [05:00<29:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 482/3295 [05:00<29:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 483/3295 [05:00<29:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 483/3295 [05:00<29:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 484/3295 [05:02<29:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 484/3295 [05:02<29:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 485/3295 [05:02<29:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 485/3295 [05:02<29:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 486/3295 [05:02<29:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 486/3295 [05:02<29:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 487/3295 [05:04<29:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 487/3295 [05:04<29:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 488/3295 [05:04<29:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 488/3295 [05:04<29:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 489/3295 [05:04<29:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 489/3295 [05:04<29:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 490/3295 [05:06<29:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 490/3295 [05:06<29:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 491/3295 [05:06<29:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 491/3295 [05:06<29:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 492/3295 [05:06<29:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 492/3295 [05:06<29:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 493/3295 [05:08<29:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 493/3295 [05:08<29:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 494/3295 [05:08<29:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▍        | 494/3295 [05:08<29:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 495/3295 [05:08<29:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 495/3295 [05:08<29:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 496/3295 [05:10<29:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 496/3295 [05:10<29:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 497/3295 [05:10<29:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 497/3295 [05:10<29:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 498/3295 [05:10<29:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 498/3295 [05:10<29:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 499/3295 [05:11<29:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 499/3295 [05:11<29:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 500/3295 [05:11<29:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 500/3295 [05:11<29:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 501/3295 [05:11<28:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 501/3295 [05:11<28:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 502/3295 [05:13<29:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 502/3295 [05:13<29:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 503/3295 [05:13<29:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 503/3295 [05:13<29:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 504/3295 [05:13<28:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 504/3295 [05:13<28:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 505/3295 [05:15<29:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 505/3295 [05:15<29:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 506/3295 [05:15<28:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 506/3295 [05:15<28:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 507/3295 [05:15<28:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 507/3295 [05:15<28:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 508/3295 [05:17<29:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 508/3295 [05:17<29:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 509/3295 [05:17<28:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 509/3295 [05:17<28:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 510/3295 [05:17<28:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  15%|█▌        | 510/3295 [05:17<28:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 511/3295 [05:19<28:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 511/3295 [05:19<28:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 512/3295 [05:19<28:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 512/3295 [05:19<28:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 513/3295 [05:19<28:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 513/3295 [05:19<28:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 514/3295 [05:20<28:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 514/3295 [05:20<28:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 515/3295 [05:20<28:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 515/3295 [05:20<28:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 516/3295 [05:21<28:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 516/3295 [05:21<28:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 517/3295 [05:22<28:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 517/3295 [05:22<28:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 518/3295 [05:22<28:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 518/3295 [05:22<28:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 519/3295 [05:22<28:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 519/3295 [05:22<28:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 520/3295 [05:24<28:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 520/3295 [05:24<28:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 521/3295 [05:24<28:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 521/3295 [05:24<28:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 522/3295 [05:24<28:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 522/3295 [05:24<28:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 523/3295 [05:26<28:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 523/3295 [05:26<28:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 524/3295 [05:26<28:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 524/3295 [05:26<28:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 525/3295 [05:26<28:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 525/3295 [05:26<28:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 526/3295 [05:28<28:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 526/3295 [05:28<28:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 527/3295 [05:28<28:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 527/3295 [05:28<28:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 528/3295 [05:28<28:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 528/3295 [05:28<28:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 529/3295 [05:29<28:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 529/3295 [05:29<28:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 530/3295 [05:30<28:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 530/3295 [05:30<28:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 531/3295 [05:30<28:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 531/3295 [05:30<28:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 532/3295 [05:31<28:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 532/3295 [05:31<28:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 533/3295 [05:31<28:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 533/3295 [05:31<28:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 534/3295 [05:31<28:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 534/3295 [05:31<28:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 535/3295 [05:33<28:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▌        | 535/3295 [05:33<28:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 536/3295 [05:33<28:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 536/3295 [05:33<28:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 537/3295 [05:33<28:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 537/3295 [05:33<28:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 538/3295 [05:35<28:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 538/3295 [05:35<28:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 539/3295 [05:35<28:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 539/3295 [05:35<28:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 540/3295 [05:35<28:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 540/3295 [05:35<28:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 541/3295 [05:37<28:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 541/3295 [05:37<28:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 542/3295 [05:37<28:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 542/3295 [05:37<28:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 543/3295 [05:37<28:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  16%|█▋        | 543/3295 [05:37<28:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 544/3295 [05:39<28:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 544/3295 [05:39<28:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 545/3295 [05:39<28:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 545/3295 [05:39<28:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 546/3295 [05:39<28:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 546/3295 [05:39<28:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 547/3295 [05:41<28:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 547/3295 [05:41<28:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 548/3295 [05:41<28:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 548/3295 [05:41<28:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 549/3295 [05:41<28:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 549/3295 [05:41<28:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 550/3295 [05:42<28:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 550/3295 [05:42<28:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 551/3295 [05:42<28:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 551/3295 [05:42<28:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 552/3295 [05:42<28:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 552/3295 [05:42<28:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 553/3295 [05:44<28:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 553/3295 [05:44<28:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 554/3295 [05:44<28:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 554/3295 [05:44<28:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 555/3295 [05:44<28:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 555/3295 [05:44<28:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 556/3295 [05:46<28:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 556/3295 [05:46<28:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 557/3295 [05:46<28:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 557/3295 [05:46<28:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 558/3295 [05:46<28:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 558/3295 [05:46<28:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 559/3295 [05:48<28:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 559/3295 [05:48<28:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 560/3295 [05:48<28:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 560/3295 [05:48<28:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 561/3295 [05:48<28:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 561/3295 [05:48<28:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 562/3295 [05:50<28:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 562/3295 [05:50<28:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 563/3295 [05:50<28:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 563/3295 [05:50<28:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 564/3295 [05:50<28:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 564/3295 [05:50<28:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 565/3295 [05:52<28:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 565/3295 [05:52<28:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 566/3295 [05:52<28:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 566/3295 [05:52<28:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 567/3295 [05:52<28:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 567/3295 [05:52<28:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 568/3295 [05:53<28:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 568/3295 [05:53<28:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 569/3295 [05:53<28:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 569/3295 [05:53<28:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 570/3295 [05:53<28:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 570/3295 [05:53<28:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 571/3295 [05:55<28:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 571/3295 [05:55<28:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 572/3295 [05:55<28:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 572/3295 [05:55<28:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 573/3295 [05:55<28:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 573/3295 [05:55<28:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 574/3295 [05:57<28:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 574/3295 [05:57<28:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 575/3295 [05:57<28:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 575/3295 [05:57<28:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 576/3295 [05:57<28:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  17%|█▋        | 576/3295 [05:57<28:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 577/3295 [05:59<28:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 577/3295 [05:59<28:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 578/3295 [05:59<28:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 578/3295 [05:59<28:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 579/3295 [05:59<28:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 579/3295 [05:59<28:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 580/3295 [06:01<28:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 580/3295 [06:01<28:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 581/3295 [06:01<28:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 581/3295 [06:01<28:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 582/3295 [06:01<28:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 582/3295 [06:01<28:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 583/3295 [06:03<28:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 583/3295 [06:03<28:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 584/3295 [06:03<28:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 584/3295 [06:03<28:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 585/3295 [06:03<28:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 585/3295 [06:03<28:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 586/3295 [06:04<28:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 586/3295 [06:04<28:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 587/3295 [06:05<28:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 587/3295 [06:05<28:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 588/3295 [06:05<28:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 588/3295 [06:05<28:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 589/3295 [06:06<28:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 589/3295 [06:06<28:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 590/3295 [06:06<28:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 590/3295 [06:06<28:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 591/3295 [06:06<27:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 591/3295 [06:06<27:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 592/3295 [06:08<28:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 592/3295 [06:08<28:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 593/3295 [06:08<27:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 593/3295 [06:08<27:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 594/3295 [06:08<27:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 594/3295 [06:08<27:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 595/3295 [06:10<28:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 595/3295 [06:10<28:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 596/3295 [06:10<27:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 596/3295 [06:10<27:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 597/3295 [06:10<27:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 597/3295 [06:10<27:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 598/3295 [06:12<27:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 598/3295 [06:12<27:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 599/3295 [06:12<27:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 599/3295 [06:12<27:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 600/3295 [06:12<27:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 600/3295 [06:12<27:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 601/3295 [06:14<27:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 601/3295 [06:14<27:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 602/3295 [06:14<27:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 602/3295 [06:14<27:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 603/3295 [06:14<27:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 603/3295 [06:14<27:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 604/3295 [06:15<27:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 604/3295 [06:15<27:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 605/3295 [06:15<27:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 605/3295 [06:15<27:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 606/3295 [06:15<27:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 606/3295 [06:15<27:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 607/3295 [06:17<27:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 607/3295 [06:17<27:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 608/3295 [06:17<27:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 608/3295 [06:17<27:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 609/3295 [06:17<27:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  18%|█▊        | 609/3295 [06:17<27:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 610/3295 [06:19<27:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 610/3295 [06:19<27:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 611/3295 [06:19<27:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 611/3295 [06:19<27:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 612/3295 [06:19<27:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 612/3295 [06:19<27:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 613/3295 [06:21<27:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 613/3295 [06:21<27:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 614/3295 [06:21<27:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 614/3295 [06:21<27:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 615/3295 [06:21<27:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 615/3295 [06:21<27:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 616/3295 [06:23<27:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 616/3295 [06:23<27:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 617/3295 [06:23<27:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▊        | 617/3295 [06:23<27:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 618/3295 [06:23<27:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 618/3295 [06:23<27:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 619/3295 [06:25<27:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 619/3295 [06:25<27:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 620/3295 [06:25<27:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 620/3295 [06:25<27:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 621/3295 [06:25<27:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 621/3295 [06:25<27:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 622/3295 [06:27<27:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 622/3295 [06:27<27:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 623/3295 [06:27<27:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 623/3295 [06:27<27:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 624/3295 [06:27<27:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 624/3295 [06:27<27:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 625/3295 [06:29<27:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 625/3295 [06:29<27:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 626/3295 [06:29<27:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 626/3295 [06:29<27:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 627/3295 [06:29<27:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 627/3295 [06:29<27:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 628/3295 [06:31<27:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 628/3295 [06:31<27:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 629/3295 [06:31<27:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 629/3295 [06:31<27:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 630/3295 [06:31<27:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 630/3295 [06:31<27:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 631/3295 [06:32<27:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 631/3295 [06:32<27:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 632/3295 [06:32<27:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 632/3295 [06:32<27:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 633/3295 [06:32<27:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 633/3295 [06:32<27:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 634/3295 [06:34<27:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 634/3295 [06:34<27:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 635/3295 [06:34<27:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 635/3295 [06:34<27:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 636/3295 [06:34<27:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 636/3295 [06:34<27:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 637/3295 [06:36<27:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 637/3295 [06:36<27:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 638/3295 [06:36<27:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 638/3295 [06:36<27:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 639/3295 [06:36<27:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 639/3295 [06:36<27:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 640/3295 [06:38<27:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 640/3295 [06:38<27:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 641/3295 [06:38<27:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 641/3295 [06:38<27:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 642/3295 [06:38<27:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  19%|█▉        | 642/3295 [06:38<27:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 643/3295 [06:40<27:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 643/3295 [06:40<27:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 644/3295 [06:40<27:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 644/3295 [06:40<27:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 645/3295 [06:40<27:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 645/3295 [06:40<27:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 646/3295 [06:42<27:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 646/3295 [06:42<27:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 647/3295 [06:42<27:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 647/3295 [06:42<27:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 648/3295 [06:42<27:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 648/3295 [06:42<27:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 649/3295 [06:43<27:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 649/3295 [06:43<27:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 650/3295 [06:43<27:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 650/3295 [06:43<27:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 651/3295 [06:43<27:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 651/3295 [06:43<27:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 652/3295 [06:45<27:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 652/3295 [06:45<27:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 653/3295 [06:45<27:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 653/3295 [06:45<27:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 654/3295 [06:45<27:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 654/3295 [06:45<27:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 655/3295 [06:47<27:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 655/3295 [06:47<27:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 656/3295 [06:47<27:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 656/3295 [06:47<27:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 657/3295 [06:47<27:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 657/3295 [06:47<27:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 658/3295 [06:49<27:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|█▉        | 658/3295 [06:49<27:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 659/3295 [06:49<27:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 659/3295 [06:49<27:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 660/3295 [06:49<27:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 660/3295 [06:49<27:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 661/3295 [06:51<27:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 661/3295 [06:51<27:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 662/3295 [06:51<27:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 662/3295 [06:51<27:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 663/3295 [06:51<27:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 663/3295 [06:51<27:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 664/3295 [06:53<27:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 664/3295 [06:53<27:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 665/3295 [06:53<27:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 665/3295 [06:53<27:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 666/3295 [06:53<27:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 666/3295 [06:53<27:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 667/3295 [06:54<27:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 667/3295 [06:54<27:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 668/3295 [06:54<27:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 668/3295 [06:54<27:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 669/3295 [06:54<27:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 669/3295 [06:54<27:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 670/3295 [06:56<27:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 670/3295 [06:56<27:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 671/3295 [06:56<27:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 671/3295 [06:56<27:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 672/3295 [06:56<27:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 672/3295 [06:56<27:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 673/3295 [06:58<27:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 673/3295 [06:58<27:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 674/3295 [06:58<27:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 674/3295 [06:58<27:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 675/3295 [06:58<27:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  20%|██        | 675/3295 [06:58<27:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 676/3295 [07:00<27:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 676/3295 [07:00<27:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 677/3295 [07:00<27:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 677/3295 [07:00<27:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 678/3295 [07:00<27:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 678/3295 [07:00<27:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 679/3295 [07:02<27:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 679/3295 [07:02<27:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 680/3295 [07:02<27:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 680/3295 [07:02<27:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 681/3295 [07:02<27:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 681/3295 [07:02<27:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 682/3295 [07:03<27:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 682/3295 [07:03<27:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 683/3295 [07:03<27:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 683/3295 [07:03<27:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 684/3295 [07:03<26:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 684/3295 [07:03<26:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 685/3295 [07:05<27:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 685/3295 [07:05<27:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 686/3295 [07:05<26:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 686/3295 [07:05<26:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 687/3295 [07:05<26:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 687/3295 [07:05<26:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 688/3295 [07:07<26:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 688/3295 [07:07<26:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 689/3295 [07:07<26:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 689/3295 [07:07<26:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 690/3295 [07:07<26:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 690/3295 [07:07<26:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 691/3295 [07:09<26:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 691/3295 [07:09<26:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 692/3295 [07:09<26:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 692/3295 [07:09<26:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 693/3295 [07:09<26:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 693/3295 [07:09<26:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 694/3295 [07:10<26:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 694/3295 [07:10<26:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 695/3295 [07:11<26:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 695/3295 [07:11<26:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 696/3295 [07:11<26:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 696/3295 [07:11<26:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 697/3295 [07:12<26:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 697/3295 [07:12<26:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 698/3295 [07:12<26:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 698/3295 [07:12<26:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 699/3295 [07:12<26:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 699/3295 [07:12<26:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 700/3295 [07:14<26:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██        | 700/3295 [07:14<26:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 701/3295 [07:14<26:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 701/3295 [07:14<26:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 702/3295 [07:14<26:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 702/3295 [07:14<26:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 703/3295 [07:16<26:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 703/3295 [07:16<26:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 704/3295 [07:16<26:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 704/3295 [07:16<26:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 705/3295 [07:16<26:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 705/3295 [07:16<26:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 706/3295 [07:18<26:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 706/3295 [07:18<26:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 707/3295 [07:18<26:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 707/3295 [07:18<26:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 708/3295 [07:18<26:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  21%|██▏       | 708/3295 [07:18<26:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 709/3295 [07:20<26:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 709/3295 [07:20<26:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 710/3295 [07:20<26:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 710/3295 [07:20<26:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 711/3295 [07:20<26:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 711/3295 [07:20<26:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 712/3295 [07:22<26:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 712/3295 [07:22<26:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 713/3295 [07:22<26:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 713/3295 [07:22<26:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 714/3295 [07:22<26:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 714/3295 [07:22<26:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 715/3295 [07:23<26:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 715/3295 [07:23<26:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 716/3295 [07:24<26:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 716/3295 [07:24<26:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 717/3295 [07:24<26:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 717/3295 [07:24<26:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 718/3295 [07:25<26:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 718/3295 [07:25<26:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 719/3295 [07:26<26:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 719/3295 [07:26<26:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 720/3295 [07:26<26:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 720/3295 [07:26<26:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 721/3295 [07:27<26:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 721/3295 [07:27<26:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 722/3295 [07:28<26:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 722/3295 [07:28<26:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 723/3295 [07:28<26:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 723/3295 [07:28<26:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 724/3295 [07:29<26:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 724/3295 [07:29<26:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 725/3295 [07:29<26:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 725/3295 [07:29<26:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 726/3295 [07:29<26:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 726/3295 [07:29<26:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 727/3295 [07:31<26:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 727/3295 [07:31<26:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 728/3295 [07:31<26:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 728/3295 [07:31<26:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 729/3295 [07:31<26:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 729/3295 [07:31<26:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 730/3295 [07:33<26:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 730/3295 [07:33<26:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 731/3295 [07:33<26:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 731/3295 [07:33<26:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 732/3295 [07:33<26:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 732/3295 [07:33<26:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 733/3295 [07:35<26:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 733/3295 [07:35<26:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 734/3295 [07:35<26:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 734/3295 [07:35<26:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 735/3295 [07:35<26:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 735/3295 [07:35<26:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 736/3295 [07:37<26:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 736/3295 [07:37<26:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 737/3295 [07:37<26:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 737/3295 [07:37<26:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 738/3295 [07:37<26:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 738/3295 [07:37<26:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 739/3295 [07:39<26:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 739/3295 [07:39<26:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 740/3295 [07:39<26:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 740/3295 [07:39<26:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 741/3295 [07:39<26:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  22%|██▏       | 741/3295 [07:39<26:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 742/3295 [07:40<26:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 742/3295 [07:40<26:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 743/3295 [07:41<26:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 743/3295 [07:41<26:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 744/3295 [07:41<26:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 744/3295 [07:41<26:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 745/3295 [07:42<26:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 745/3295 [07:42<26:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 746/3295 [07:43<26:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 746/3295 [07:43<26:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 747/3295 [07:43<26:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 747/3295 [07:43<26:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 748/3295 [07:44<26:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 748/3295 [07:44<26:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 749/3295 [07:44<26:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 749/3295 [07:44<26:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 750/3295 [07:44<26:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 750/3295 [07:44<26:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 751/3295 [07:46<26:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 751/3295 [07:46<26:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 752/3295 [07:46<26:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 752/3295 [07:46<26:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 753/3295 [07:46<26:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 753/3295 [07:46<26:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 754/3295 [07:48<26:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 754/3295 [07:48<26:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 755/3295 [07:48<26:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 755/3295 [07:48<26:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 756/3295 [07:48<26:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 756/3295 [07:48<26:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 757/3295 [07:49<26:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 757/3295 [07:49<26:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 758/3295 [07:50<26:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 758/3295 [07:50<26:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 759/3295 [07:50<26:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 759/3295 [07:50<26:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 760/3295 [07:51<26:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 760/3295 [07:51<26:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 761/3295 [07:52<26:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 761/3295 [07:52<26:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 762/3295 [07:52<26:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 762/3295 [07:52<26:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 763/3295 [07:53<26:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 763/3295 [07:53<26:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 764/3295 [07:53<26:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 764/3295 [07:53<26:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 765/3295 [07:53<26:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 765/3295 [07:53<26:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 766/3295 [07:55<26:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 766/3295 [07:55<26:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 767/3295 [07:55<26:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 767/3295 [07:55<26:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 768/3295 [07:55<26:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 768/3295 [07:55<26:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 769/3295 [07:57<26:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 769/3295 [07:57<26:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 770/3295 [07:57<26:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 770/3295 [07:57<26:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 771/3295 [07:57<26:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 771/3295 [07:57<26:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 772/3295 [07:59<26:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 772/3295 [07:59<26:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 773/3295 [07:59<26:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 773/3295 [07:59<26:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 774/3295 [07:59<26:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  23%|██▎       | 774/3295 [07:59<26:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 775/3295 [08:01<26:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 775/3295 [08:01<26:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 776/3295 [08:01<26:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 776/3295 [08:01<26:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 777/3295 [08:01<26:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 777/3295 [08:01<26:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 778/3295 [08:03<26:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 778/3295 [08:03<26:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 779/3295 [08:03<26:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 779/3295 [08:03<26:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 780/3295 [08:03<25:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 780/3295 [08:03<25:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 781/3295 [08:04<26:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 781/3295 [08:04<26:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 782/3295 [08:05<25:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▎       | 782/3295 [08:05<25:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 783/3295 [08:05<25:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 783/3295 [08:05<25:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 784/3295 [08:06<25:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 784/3295 [08:06<25:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 785/3295 [08:07<25:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 785/3295 [08:07<25:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 786/3295 [08:07<25:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 786/3295 [08:07<25:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 787/3295 [08:08<25:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 787/3295 [08:08<25:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 788/3295 [08:09<25:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 788/3295 [08:09<25:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 789/3295 [08:09<25:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 789/3295 [08:09<25:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 790/3295 [08:10<25:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 790/3295 [08:10<25:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 791/3295 [08:10<25:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 791/3295 [08:10<25:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 792/3295 [08:10<25:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 792/3295 [08:10<25:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 793/3295 [08:12<25:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 793/3295 [08:12<25:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 794/3295 [08:12<25:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 794/3295 [08:12<25:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 795/3295 [08:12<25:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 795/3295 [08:12<25:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 796/3295 [08:14<25:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 796/3295 [08:14<25:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 797/3295 [08:14<25:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 797/3295 [08:14<25:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 798/3295 [08:14<25:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 798/3295 [08:14<25:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 799/3295 [08:15<25:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 799/3295 [08:15<25:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 800/3295 [08:16<25:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 800/3295 [08:16<25:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 801/3295 [08:16<25:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 801/3295 [08:16<25:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 802/3295 [08:17<25:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 802/3295 [08:17<25:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 803/3295 [08:18<25:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 803/3295 [08:18<25:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 804/3295 [08:18<25:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 804/3295 [08:18<25:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 805/3295 [08:19<25:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 805/3295 [08:19<25:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 806/3295 [08:20<25:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 806/3295 [08:20<25:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 807/3295 [08:20<25:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  24%|██▍       | 807/3295 [08:20<25:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 808/3295 [08:21<25:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 808/3295 [08:21<25:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 809/3295 [08:22<25:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 809/3295 [08:22<25:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 810/3295 [08:22<25:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 810/3295 [08:22<25:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 811/3295 [08:22<25:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 811/3295 [08:22<25:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 812/3295 [08:24<25:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 812/3295 [08:24<25:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 813/3295 [08:24<25:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 813/3295 [08:24<25:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 814/3295 [08:24<25:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 814/3295 [08:24<25:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 815/3295 [08:26<25:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 815/3295 [08:26<25:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 816/3295 [08:26<25:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 816/3295 [08:26<25:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 817/3295 [08:26<25:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 817/3295 [08:26<25:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 818/3295 [08:27<25:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 818/3295 [08:27<25:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 819/3295 [08:27<25:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 819/3295 [08:27<25:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 820/3295 [08:28<25:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 820/3295 [08:28<25:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 821/3295 [08:29<25:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 821/3295 [08:29<25:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 822/3295 [08:29<25:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 822/3295 [08:29<25:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 823/3295 [08:30<25:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▍       | 823/3295 [08:30<25:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 824/3295 [08:31<25:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 824/3295 [08:31<25:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 825/3295 [08:31<25:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 825/3295 [08:31<25:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 826/3295 [08:32<25:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 826/3295 [08:32<25:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 827/3295 [08:33<25:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 827/3295 [08:33<25:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 828/3295 [08:33<25:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 828/3295 [08:33<25:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 829/3295 [08:33<25:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 829/3295 [08:33<25:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 830/3295 [08:35<25:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 830/3295 [08:35<25:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 831/3295 [08:35<25:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 831/3295 [08:35<25:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 832/3295 [08:35<25:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 832/3295 [08:35<25:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 833/3295 [08:37<25:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 833/3295 [08:37<25:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 834/3295 [08:37<25:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 834/3295 [08:37<25:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 835/3295 [08:37<25:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 835/3295 [08:37<25:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 836/3295 [08:39<25:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 836/3295 [08:39<25:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 837/3295 [08:39<25:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 837/3295 [08:39<25:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 838/3295 [08:39<25:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 838/3295 [08:39<25:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 839/3295 [08:41<25:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 839/3295 [08:41<25:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 840/3295 [08:41<25:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  25%|██▌       | 840/3295 [08:41<25:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 841/3295 [08:41<25:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 841/3295 [08:41<25:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 842/3295 [08:43<25:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 842/3295 [08:43<25:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 843/3295 [08:43<25:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 843/3295 [08:43<25:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 844/3295 [08:43<25:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 844/3295 [08:43<25:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 845/3295 [08:44<25:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 845/3295 [08:44<25:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 846/3295 [08:44<25:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 846/3295 [08:44<25:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 847/3295 [08:44<25:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 847/3295 [08:44<25:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 848/3295 [08:46<25:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 848/3295 [08:46<25:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 849/3295 [08:46<25:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 849/3295 [08:46<25:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 850/3295 [08:46<25:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 850/3295 [08:46<25:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 851/3295 [08:48<25:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 851/3295 [08:48<25:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 852/3295 [08:48<25:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 852/3295 [08:48<25:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 853/3295 [08:48<25:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 853/3295 [08:48<25:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 854/3295 [08:50<25:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 854/3295 [08:50<25:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 855/3295 [08:50<25:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 855/3295 [08:50<25:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 856/3295 [08:50<25:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 856/3295 [08:50<25:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 857/3295 [08:52<25:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 857/3295 [08:52<25:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 858/3295 [08:52<25:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 858/3295 [08:52<25:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 859/3295 [08:52<25:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 859/3295 [08:52<25:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 860/3295 [08:54<25:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 860/3295 [08:54<25:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 861/3295 [08:54<25:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 861/3295 [08:54<25:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 862/3295 [08:54<25:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 862/3295 [08:54<25:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 863/3295 [08:56<25:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 863/3295 [08:56<25:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 864/3295 [08:56<25:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▌       | 864/3295 [08:56<25:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 865/3295 [08:56<25:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 865/3295 [08:56<25:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 866/3295 [08:57<25:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 866/3295 [08:57<25:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 867/3295 [08:58<25:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 867/3295 [08:58<25:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 868/3295 [08:58<25:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 868/3295 [08:58<25:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 869/3295 [08:59<25:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 869/3295 [08:59<25:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 870/3295 [08:59<25:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 870/3295 [08:59<25:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 871/3295 [08:59<25:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 871/3295 [08:59<25:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 872/3295 [09:01<25:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 872/3295 [09:01<25:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 873/3295 [09:01<25:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  26%|██▋       | 873/3295 [09:01<25:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 874/3295 [09:01<25:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 874/3295 [09:01<25:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 875/3295 [09:03<25:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 875/3295 [09:03<25:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 876/3295 [09:03<25:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 876/3295 [09:03<25:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 877/3295 [09:03<24:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 877/3295 [09:03<24:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 878/3295 [09:05<25:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 878/3295 [09:05<25:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 879/3295 [09:05<24:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 879/3295 [09:05<24:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 880/3295 [09:05<24:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 880/3295 [09:05<24:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 881/3295 [09:06<24:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 881/3295 [09:06<24:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 882/3295 [09:06<24:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 882/3295 [09:06<24:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 883/3295 [09:07<24:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 883/3295 [09:07<24:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 884/3295 [09:08<24:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 884/3295 [09:08<24:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 885/3295 [09:08<24:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 885/3295 [09:08<24:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 886/3295 [09:08<24:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 886/3295 [09:08<24:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 887/3295 [09:10<24:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 887/3295 [09:10<24:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 888/3295 [09:10<24:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 888/3295 [09:10<24:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 889/3295 [09:10<24:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 889/3295 [09:10<24:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 890/3295 [09:12<24:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 890/3295 [09:12<24:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 891/3295 [09:12<24:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 891/3295 [09:12<24:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 892/3295 [09:12<24:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 892/3295 [09:12<24:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 893/3295 [09:14<24:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 893/3295 [09:14<24:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 894/3295 [09:14<24:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 894/3295 [09:14<24:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 895/3295 [09:14<24:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 895/3295 [09:14<24:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 896/3295 [09:16<24:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 896/3295 [09:16<24:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 897/3295 [09:16<24:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 897/3295 [09:16<24:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 898/3295 [09:16<24:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 898/3295 [09:16<24:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 899/3295 [09:18<24:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 899/3295 [09:18<24:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 900/3295 [09:18<24:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 900/3295 [09:18<24:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 901/3295 [09:18<24:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 901/3295 [09:18<24:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 902/3295 [09:20<24:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 902/3295 [09:20<24:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 903/3295 [09:20<24:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 903/3295 [09:20<24:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 904/3295 [09:20<24:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 904/3295 [09:20<24:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 905/3295 [09:22<24:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 905/3295 [09:22<24:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 906/3295 [09:22<24:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  27%|██▋       | 906/3295 [09:22<24:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 907/3295 [09:22<24:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 907/3295 [09:22<24:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 908/3295 [09:24<24:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 908/3295 [09:24<24:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 909/3295 [09:24<24:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 909/3295 [09:24<24:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 910/3295 [09:24<24:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 910/3295 [09:24<24:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 911/3295 [09:26<24:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 911/3295 [09:26<24:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 912/3295 [09:26<24:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 912/3295 [09:26<24:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 913/3295 [09:26<24:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 913/3295 [09:26<24:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 914/3295 [09:27<24:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 914/3295 [09:27<24:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 915/3295 [09:27<24:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 915/3295 [09:27<24:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 916/3295 [09:27<24:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 916/3295 [09:27<24:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 917/3295 [09:29<24:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 917/3295 [09:29<24:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 918/3295 [09:29<24:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 918/3295 [09:29<24:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 919/3295 [09:29<24:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 919/3295 [09:29<24:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 920/3295 [09:31<24:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 920/3295 [09:31<24:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 921/3295 [09:31<24:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 921/3295 [09:31<24:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 922/3295 [09:31<24:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 922/3295 [09:31<24:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 923/3295 [09:33<24:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 923/3295 [09:33<24:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 924/3295 [09:33<24:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 924/3295 [09:33<24:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 925/3295 [09:33<24:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 925/3295 [09:33<24:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 926/3295 [09:35<24:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 926/3295 [09:35<24:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 927/3295 [09:35<24:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 927/3295 [09:35<24:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 928/3295 [09:35<24:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 928/3295 [09:35<24:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 929/3295 [09:37<24:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 929/3295 [09:37<24:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 930/3295 [09:37<24:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 930/3295 [09:37<24:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 931/3295 [09:37<24:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 931/3295 [09:37<24:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 932/3295 [09:39<24:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 932/3295 [09:39<24:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 933/3295 [09:39<24:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 933/3295 [09:39<24:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 934/3295 [09:39<24:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 934/3295 [09:39<24:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 935/3295 [09:41<24:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 935/3295 [09:41<24:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 936/3295 [09:41<24:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 936/3295 [09:41<24:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 937/3295 [09:41<24:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 937/3295 [09:41<24:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 938/3295 [09:42<24:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 938/3295 [09:42<24:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 939/3295 [09:42<24:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  28%|██▊       | 939/3295 [09:42<24:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 940/3295 [09:42<24:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 940/3295 [09:42<24:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 941/3295 [09:44<24:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 941/3295 [09:44<24:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 942/3295 [09:44<24:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 942/3295 [09:44<24:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 943/3295 [09:44<24:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 943/3295 [09:44<24:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 944/3295 [09:46<24:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 944/3295 [09:46<24:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 945/3295 [09:46<24:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 945/3295 [09:46<24:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 946/3295 [09:46<24:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 946/3295 [09:46<24:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 947/3295 [09:48<24:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▊       | 947/3295 [09:48<24:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 948/3295 [09:48<24:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 948/3295 [09:48<24:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 949/3295 [09:48<24:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 949/3295 [09:48<24:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 950/3295 [09:50<24:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 950/3295 [09:50<24:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 951/3295 [09:50<24:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 951/3295 [09:50<24:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 952/3295 [09:50<24:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 952/3295 [09:50<24:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 953/3295 [09:52<24:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 953/3295 [09:52<24:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 954/3295 [09:52<24:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 954/3295 [09:52<24:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 955/3295 [09:52<24:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 955/3295 [09:52<24:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 956/3295 [09:53<24:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 956/3295 [09:53<24:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 957/3295 [09:54<24:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 957/3295 [09:54<24:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 958/3295 [09:54<24:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 958/3295 [09:54<24:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 959/3295 [09:55<24:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 959/3295 [09:55<24:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 960/3295 [09:55<24:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 960/3295 [09:55<24:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 961/3295 [09:55<24:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 961/3295 [09:55<24:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 962/3295 [09:57<24:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 962/3295 [09:57<24:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 963/3295 [09:57<24:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 963/3295 [09:57<24:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 964/3295 [09:57<24:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 964/3295 [09:57<24:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 965/3295 [09:59<24:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 965/3295 [09:59<24:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 966/3295 [09:59<24:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 966/3295 [09:59<24:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 967/3295 [09:59<24:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 967/3295 [09:59<24:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 968/3295 [10:01<24:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 968/3295 [10:01<24:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 969/3295 [10:01<24:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 969/3295 [10:01<24:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 970/3295 [10:01<24:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 970/3295 [10:01<24:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 971/3295 [10:03<24:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 971/3295 [10:03<24:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 972/3295 [10:03<24:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  29%|██▉       | 972/3295 [10:03<24:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 973/3295 [10:03<24:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 973/3295 [10:03<24:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 974/3295 [10:05<24:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 974/3295 [10:05<24:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 975/3295 [10:05<24:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 975/3295 [10:05<24:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 976/3295 [10:05<23:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 976/3295 [10:05<23:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 977/3295 [10:07<24:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 977/3295 [10:07<24:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 978/3295 [10:07<23:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 978/3295 [10:07<23:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 979/3295 [10:07<23:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 979/3295 [10:07<23:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 980/3295 [10:09<23:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 980/3295 [10:09<23:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 981/3295 [10:09<23:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 981/3295 [10:09<23:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 982/3295 [10:09<23:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 982/3295 [10:09<23:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 983/3295 [10:11<23:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 983/3295 [10:11<23:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 984/3295 [10:11<23:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 984/3295 [10:11<23:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 985/3295 [10:11<23:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 985/3295 [10:11<23:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 986/3295 [10:12<23:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 986/3295 [10:12<23:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 987/3295 [10:12<23:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 987/3295 [10:12<23:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 988/3295 [10:12<23:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|██▉       | 988/3295 [10:12<23:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 989/3295 [10:14<23:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 989/3295 [10:14<23:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 990/3295 [10:14<23:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 990/3295 [10:14<23:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 991/3295 [10:14<23:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 991/3295 [10:14<23:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 992/3295 [10:16<23:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 992/3295 [10:16<23:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 993/3295 [10:16<23:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 993/3295 [10:16<23:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 994/3295 [10:16<23:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 994/3295 [10:16<23:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 995/3295 [10:18<23:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 995/3295 [10:18<23:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 996/3295 [10:18<23:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 996/3295 [10:18<23:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 997/3295 [10:18<23:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 997/3295 [10:18<23:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 998/3295 [10:20<23:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 998/3295 [10:20<23:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 999/3295 [10:20<23:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 999/3295 [10:20<23:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1000/3295 [10:20<23:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1000/3295 [10:20<23:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1001/3295 [10:21<23:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1001/3295 [10:21<23:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1002/3295 [10:21<23:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1002/3295 [10:21<23:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1003/3295 [10:21<23:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1003/3295 [10:21<23:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1004/3295 [10:23<23:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  30%|███       | 1004/3295 [10:23<23:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1005/3295 [10:23<23:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1005/3295 [10:23<23:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1006/3295 [10:23<23:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1006/3295 [10:23<23:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1007/3295 [10:25<23:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1007/3295 [10:25<23:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1008/3295 [10:25<23:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1008/3295 [10:25<23:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1009/3295 [10:25<23:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1009/3295 [10:25<23:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1010/3295 [10:27<23:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1010/3295 [10:27<23:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1011/3295 [10:27<23:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1011/3295 [10:27<23:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1012/3295 [10:27<23:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1012/3295 [10:27<23:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1013/3295 [10:29<23:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1013/3295 [10:29<23:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1014/3295 [10:29<23:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1014/3295 [10:29<23:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1015/3295 [10:29<23:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1015/3295 [10:29<23:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1016/3295 [10:31<23:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1016/3295 [10:31<23:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1017/3295 [10:31<23:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1017/3295 [10:31<23:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1018/3295 [10:31<23:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1018/3295 [10:31<23:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1019/3295 [10:33<23:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1019/3295 [10:33<23:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1020/3295 [10:33<23:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1020/3295 [10:33<23:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1021/3295 [10:33<23:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1021/3295 [10:33<23:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1022/3295 [10:34<23:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1022/3295 [10:34<23:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1023/3295 [10:34<23:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1023/3295 [10:34<23:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1024/3295 [10:34<23:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1024/3295 [10:34<23:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1025/3295 [10:36<23:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1025/3295 [10:36<23:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1026/3295 [10:36<23:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1026/3295 [10:36<23:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1027/3295 [10:36<23:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1027/3295 [10:36<23:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1028/3295 [10:38<23:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1028/3295 [10:38<23:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1029/3295 [10:38<23:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███       | 1029/3295 [10:38<23:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1030/3295 [10:38<23:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1030/3295 [10:38<23:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1031/3295 [10:40<23:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1031/3295 [10:40<23:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1032/3295 [10:40<23:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1032/3295 [10:40<23:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1033/3295 [10:40<23:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1033/3295 [10:40<23:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1034/3295 [10:42<23:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1034/3295 [10:42<23:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1035/3295 [10:42<23:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1035/3295 [10:42<23:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1036/3295 [10:42<23:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1036/3295 [10:42<23:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1037/3295 [10:44<23:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  31%|███▏      | 1037/3295 [10:44<23:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1038/3295 [10:44<23:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1038/3295 [10:44<23:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1039/3295 [10:44<23:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1039/3295 [10:44<23:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1040/3295 [10:46<23:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1040/3295 [10:46<23:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1041/3295 [10:46<23:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1041/3295 [10:46<23:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1042/3295 [10:46<23:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1042/3295 [10:46<23:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1043/3295 [10:47<23:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1043/3295 [10:47<23:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1044/3295 [10:47<23:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1044/3295 [10:47<23:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1045/3295 [10:47<23:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1045/3295 [10:47<23:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1046/3295 [10:49<23:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1046/3295 [10:49<23:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1047/3295 [10:49<23:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1047/3295 [10:49<23:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1048/3295 [10:49<23:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1048/3295 [10:49<23:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1049/3295 [10:51<23:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1049/3295 [10:51<23:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1050/3295 [10:51<23:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1050/3295 [10:51<23:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1051/3295 [10:51<23:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1051/3295 [10:51<23:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1052/3295 [10:53<23:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1052/3295 [10:53<23:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1053/3295 [10:53<23:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1053/3295 [10:53<23:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1054/3295 [10:53<23:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1054/3295 [10:53<23:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1055/3295 [10:55<23:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1055/3295 [10:55<23:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1056/3295 [10:55<23:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1056/3295 [10:55<23:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1057/3295 [10:55<23:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1057/3295 [10:55<23:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1058/3295 [10:57<23:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1058/3295 [10:57<23:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1059/3295 [10:57<23:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1059/3295 [10:57<23:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1060/3295 [10:57<23:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1060/3295 [10:57<23:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1061/3295 [10:58<23:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1061/3295 [10:58<23:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1062/3295 [10:59<23:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1062/3295 [10:59<23:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1063/3295 [10:59<23:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1063/3295 [10:59<23:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1064/3295 [11:00<23:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1064/3295 [11:00<23:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1065/3295 [11:00<23:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1065/3295 [11:00<23:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1066/3295 [11:01<23:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1066/3295 [11:01<23:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1067/3295 [11:02<23:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1067/3295 [11:02<23:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1068/3295 [11:02<23:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1068/3295 [11:02<23:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1069/3295 [11:02<23:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1069/3295 [11:02<23:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1070/3295 [11:04<23:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  32%|███▏      | 1070/3295 [11:04<23:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1071/3295 [11:04<23:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1071/3295 [11:04<23:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1072/3295 [11:04<22:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1072/3295 [11:04<22:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1073/3295 [11:06<22:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1073/3295 [11:06<22:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1074/3295 [11:06<22:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1074/3295 [11:06<22:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1075/3295 [11:06<22:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1075/3295 [11:06<22:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1076/3295 [11:08<22:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1076/3295 [11:08<22:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1077/3295 [11:08<22:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1077/3295 [11:08<22:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1078/3295 [11:08<22:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1078/3295 [11:08<22:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1079/3295 [11:10<22:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1079/3295 [11:10<22:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1080/3295 [11:10<22:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1080/3295 [11:10<22:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1081/3295 [11:10<22:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1081/3295 [11:10<22:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1082/3295 [11:11<22:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1082/3295 [11:11<22:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1083/3295 [11:11<22:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1083/3295 [11:11<22:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1084/3295 [11:11<22:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1084/3295 [11:11<22:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1085/3295 [11:13<22:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1085/3295 [11:13<22:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1086/3295 [11:13<22:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1086/3295 [11:13<22:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1087/3295 [11:13<22:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1087/3295 [11:13<22:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1088/3295 [11:15<22:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1088/3295 [11:15<22:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1089/3295 [11:15<22:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1089/3295 [11:15<22:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1090/3295 [11:15<22:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1090/3295 [11:15<22:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1091/3295 [11:17<22:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1091/3295 [11:17<22:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1092/3295 [11:17<22:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1092/3295 [11:17<22:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1093/3295 [11:17<22:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1093/3295 [11:17<22:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1094/3295 [11:19<22:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1094/3295 [11:19<22:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1095/3295 [11:19<22:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1095/3295 [11:19<22:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1096/3295 [11:19<22:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1096/3295 [11:19<22:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1097/3295 [11:21<22:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1097/3295 [11:21<22:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1098/3295 [11:21<22:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1098/3295 [11:21<22:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1099/3295 [11:21<22:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1099/3295 [11:21<22:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1100/3295 [11:23<22:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1100/3295 [11:23<22:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1101/3295 [11:23<22:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1101/3295 [11:23<22:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1102/3295 [11:23<22:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1102/3295 [11:23<22:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1103/3295 [11:24<22:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  33%|███▎      | 1103/3295 [11:24<22:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1104/3295 [11:24<22:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1104/3295 [11:24<22:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1105/3295 [11:24<22:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1105/3295 [11:24<22:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1106/3295 [11:26<22:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1106/3295 [11:26<22:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1107/3295 [11:26<22:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1107/3295 [11:26<22:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1108/3295 [11:26<22:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1108/3295 [11:26<22:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1109/3295 [11:28<22:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1109/3295 [11:28<22:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1110/3295 [11:28<22:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1110/3295 [11:28<22:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1111/3295 [11:28<22:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1111/3295 [11:28<22:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1112/3295 [11:30<22:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▎      | 1112/3295 [11:30<22:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1113/3295 [11:30<22:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1113/3295 [11:30<22:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1114/3295 [11:30<22:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1114/3295 [11:30<22:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1115/3295 [11:32<22:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1115/3295 [11:32<22:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1116/3295 [11:32<22:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1116/3295 [11:32<22:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1117/3295 [11:32<22:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1117/3295 [11:32<22:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1118/3295 [11:34<22:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1118/3295 [11:34<22:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1119/3295 [11:34<22:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1119/3295 [11:34<22:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1120/3295 [11:34<22:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1120/3295 [11:34<22:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1121/3295 [11:36<22:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1121/3295 [11:36<22:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1122/3295 [11:36<22:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1122/3295 [11:36<22:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1123/3295 [11:36<22:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1123/3295 [11:36<22:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1124/3295 [11:38<22:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1124/3295 [11:38<22:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1125/3295 [11:38<22:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1125/3295 [11:38<22:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1126/3295 [11:38<22:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1126/3295 [11:38<22:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1127/3295 [11:39<22:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1127/3295 [11:39<22:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1128/3295 [11:39<22:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1128/3295 [11:39<22:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1129/3295 [11:39<22:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1129/3295 [11:39<22:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1130/3295 [11:41<22:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1130/3295 [11:41<22:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1131/3295 [11:41<22:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1131/3295 [11:41<22:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1132/3295 [11:41<22:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1132/3295 [11:41<22:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1133/3295 [11:43<22:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1133/3295 [11:43<22:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1134/3295 [11:43<22:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1134/3295 [11:43<22:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1135/3295 [11:43<22:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1135/3295 [11:43<22:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1136/3295 [11:45<22:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  34%|███▍      | 1136/3295 [11:45<22:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1137/3295 [11:45<22:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1137/3295 [11:45<22:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1138/3295 [11:45<22:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1138/3295 [11:45<22:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1139/3295 [11:47<22:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1139/3295 [11:47<22:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1140/3295 [11:47<22:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1140/3295 [11:47<22:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1141/3295 [11:47<22:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1141/3295 [11:47<22:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1142/3295 [11:49<22:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1142/3295 [11:49<22:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1143/3295 [11:49<22:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1143/3295 [11:49<22:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1144/3295 [11:49<22:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1144/3295 [11:49<22:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1145/3295 [11:50<22:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1145/3295 [11:50<22:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1146/3295 [11:50<22:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1146/3295 [11:50<22:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1147/3295 [11:50<22:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1147/3295 [11:50<22:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1148/3295 [11:52<22:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1148/3295 [11:52<22:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1149/3295 [11:52<22:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1149/3295 [11:52<22:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1150/3295 [11:52<22:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1150/3295 [11:52<22:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1151/3295 [11:54<22:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1151/3295 [11:54<22:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1152/3295 [11:54<22:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1152/3295 [11:54<22:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1153/3295 [11:54<22:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▍      | 1153/3295 [11:54<22:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1154/3295 [11:56<22:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1154/3295 [11:56<22:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1155/3295 [11:56<22:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1155/3295 [11:56<22:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1156/3295 [11:56<22:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1156/3295 [11:56<22:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1157/3295 [11:57<22:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1157/3295 [11:57<22:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1158/3295 [11:57<22:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1158/3295 [11:57<22:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1159/3295 [11:57<22:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1159/3295 [11:57<22:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1160/3295 [11:59<22:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1160/3295 [11:59<22:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1161/3295 [11:59<22:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1161/3295 [11:59<22:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1162/3295 [11:59<22:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1162/3295 [11:59<22:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1163/3295 [12:01<22:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1163/3295 [12:01<22:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1164/3295 [12:01<22:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1164/3295 [12:01<22:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1165/3295 [12:01<21:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1165/3295 [12:01<21:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1166/3295 [12:03<22:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1166/3295 [12:03<22:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1167/3295 [12:03<21:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1167/3295 [12:03<21:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1168/3295 [12:03<21:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1168/3295 [12:03<21:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1169/3295 [12:05<21:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  35%|███▌      | 1169/3295 [12:05<21:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1170/3295 [12:05<21:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1170/3295 [12:05<21:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1171/3295 [12:05<21:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1171/3295 [12:05<21:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1172/3295 [12:06<21:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1172/3295 [12:06<21:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1173/3295 [12:07<21:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1173/3295 [12:07<21:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1174/3295 [12:07<21:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1174/3295 [12:07<21:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1175/3295 [12:08<21:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1175/3295 [12:08<21:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1176/3295 [12:08<21:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1176/3295 [12:08<21:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1177/3295 [12:08<21:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1177/3295 [12:08<21:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1178/3295 [12:10<21:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1178/3295 [12:10<21:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1179/3295 [12:10<21:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1179/3295 [12:10<21:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1180/3295 [12:10<21:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1180/3295 [12:10<21:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1181/3295 [12:12<21:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1181/3295 [12:12<21:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1182/3295 [12:12<21:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1182/3295 [12:12<21:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1183/3295 [12:12<21:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1183/3295 [12:12<21:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1184/3295 [12:14<21:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1184/3295 [12:14<21:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1185/3295 [12:14<21:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1185/3295 [12:14<21:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1186/3295 [12:14<21:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1186/3295 [12:14<21:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1187/3295 [12:16<21:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1187/3295 [12:16<21:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1188/3295 [12:16<21:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1188/3295 [12:16<21:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1189/3295 [12:16<21:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1189/3295 [12:16<21:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1190/3295 [12:18<21:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1190/3295 [12:18<21:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1191/3295 [12:18<21:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1191/3295 [12:18<21:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1192/3295 [12:18<21:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1192/3295 [12:18<21:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1193/3295 [12:20<21:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1193/3295 [12:20<21:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1194/3295 [12:20<21:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▌      | 1194/3295 [12:20<21:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1195/3295 [12:20<21:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1195/3295 [12:20<21:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1196/3295 [12:22<21:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1196/3295 [12:22<21:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1197/3295 [12:22<21:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1197/3295 [12:22<21:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1198/3295 [12:22<21:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1198/3295 [12:22<21:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1199/3295 [12:24<21:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1199/3295 [12:24<21:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1200/3295 [12:24<21:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1200/3295 [12:24<21:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1201/3295 [12:24<21:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1201/3295 [12:24<21:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1202/3295 [12:26<21:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  36%|███▋      | 1202/3295 [12:26<21:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1203/3295 [12:26<21:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1203/3295 [12:26<21:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1204/3295 [12:26<21:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1204/3295 [12:26<21:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1205/3295 [12:28<21:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1205/3295 [12:28<21:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1206/3295 [12:28<21:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1206/3295 [12:28<21:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1207/3295 [12:28<21:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1207/3295 [12:28<21:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1208/3295 [12:29<21:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1208/3295 [12:29<21:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1209/3295 [12:29<21:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1209/3295 [12:29<21:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1210/3295 [12:29<21:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1210/3295 [12:29<21:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1211/3295 [12:31<21:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1211/3295 [12:31<21:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1212/3295 [12:31<21:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1212/3295 [12:31<21:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1213/3295 [12:31<21:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1213/3295 [12:31<21:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1214/3295 [12:33<21:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1214/3295 [12:33<21:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1215/3295 [12:33<21:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1215/3295 [12:33<21:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1216/3295 [12:33<21:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1216/3295 [12:33<21:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1217/3295 [12:35<21:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1217/3295 [12:35<21:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1218/3295 [12:35<21:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1218/3295 [12:35<21:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1219/3295 [12:35<21:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1219/3295 [12:35<21:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1220/3295 [12:37<21:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1220/3295 [12:37<21:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1221/3295 [12:37<21:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1221/3295 [12:37<21:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1222/3295 [12:37<21:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1222/3295 [12:37<21:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1223/3295 [12:39<21:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1223/3295 [12:39<21:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1224/3295 [12:39<21:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1224/3295 [12:39<21:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1225/3295 [12:39<21:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1225/3295 [12:39<21:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1226/3295 [12:40<21:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1226/3295 [12:40<21:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1227/3295 [12:41<21:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1227/3295 [12:41<21:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1228/3295 [12:41<21:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1228/3295 [12:41<21:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1229/3295 [12:42<21:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1229/3295 [12:42<21:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1230/3295 [12:42<21:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1230/3295 [12:42<21:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1231/3295 [12:42<21:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1231/3295 [12:42<21:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1232/3295 [12:44<21:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1232/3295 [12:44<21:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1233/3295 [12:44<21:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1233/3295 [12:44<21:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1234/3295 [12:44<21:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1234/3295 [12:44<21:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1235/3295 [12:46<21:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  37%|███▋      | 1235/3295 [12:46<21:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1236/3295 [12:46<21:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1236/3295 [12:46<21:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1237/3295 [12:46<21:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1237/3295 [12:46<21:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1238/3295 [12:48<21:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1238/3295 [12:48<21:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1239/3295 [12:48<21:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1239/3295 [12:48<21:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1240/3295 [12:48<21:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1240/3295 [12:48<21:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1241/3295 [12:50<21:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1241/3295 [12:50<21:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1242/3295 [12:50<21:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1242/3295 [12:50<21:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1243/3295 [12:50<21:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1243/3295 [12:50<21:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1244/3295 [12:51<21:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1244/3295 [12:51<21:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1245/3295 [12:51<21:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1245/3295 [12:51<21:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1246/3295 [12:51<21:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1246/3295 [12:51<21:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1247/3295 [12:53<21:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1247/3295 [12:53<21:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1248/3295 [12:53<21:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1248/3295 [12:53<21:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1249/3295 [12:53<21:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1249/3295 [12:53<21:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1250/3295 [12:55<21:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1250/3295 [12:55<21:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1251/3295 [12:55<21:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1251/3295 [12:55<21:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1252/3295 [12:55<21:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1252/3295 [12:55<21:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1253/3295 [12:57<21:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1253/3295 [12:57<21:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1254/3295 [12:57<21:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1254/3295 [12:57<21:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1255/3295 [12:57<21:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1255/3295 [12:57<21:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1256/3295 [12:59<21:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1256/3295 [12:59<21:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1257/3295 [12:59<21:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1257/3295 [12:59<21:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1258/3295 [12:59<21:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1258/3295 [12:59<21:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1259/3295 [13:01<21:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1259/3295 [13:01<21:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1260/3295 [13:01<21:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1260/3295 [13:01<21:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1261/3295 [13:01<21:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1261/3295 [13:01<21:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1262/3295 [13:02<21:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1262/3295 [13:02<21:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1263/3295 [13:03<20:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1263/3295 [13:03<20:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1264/3295 [13:03<20:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1264/3295 [13:03<20:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1265/3295 [13:04<20:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1265/3295 [13:04<20:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1266/3295 [13:04<20:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1266/3295 [13:04<20:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1267/3295 [13:04<20:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1267/3295 [13:04<20:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1268/3295 [13:06<20:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  38%|███▊      | 1268/3295 [13:06<20:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1269/3295 [13:06<20:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1269/3295 [13:06<20:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1270/3295 [13:06<20:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1270/3295 [13:06<20:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1271/3295 [13:08<20:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1271/3295 [13:08<20:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1272/3295 [13:08<20:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1272/3295 [13:08<20:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1273/3295 [13:08<20:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1273/3295 [13:08<20:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1274/3295 [13:10<20:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1274/3295 [13:10<20:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1275/3295 [13:10<20:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1275/3295 [13:10<20:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1276/3295 [13:10<20:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▊      | 1276/3295 [13:10<20:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1277/3295 [13:12<20:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1277/3295 [13:12<20:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1278/3295 [13:12<20:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1278/3295 [13:12<20:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1279/3295 [13:12<20:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1279/3295 [13:12<20:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1280/3295 [13:13<20:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1280/3295 [13:13<20:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1281/3295 [13:13<20:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1281/3295 [13:13<20:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1282/3295 [13:13<20:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1282/3295 [13:13<20:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1283/3295 [13:15<20:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1283/3295 [13:15<20:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1284/3295 [13:15<20:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1284/3295 [13:15<20:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1285/3295 [13:15<20:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1285/3295 [13:15<20:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1286/3295 [13:17<20:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1286/3295 [13:17<20:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1287/3295 [13:17<20:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1287/3295 [13:17<20:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1288/3295 [13:17<20:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1288/3295 [13:17<20:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1289/3295 [13:19<20:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1289/3295 [13:19<20:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1290/3295 [13:19<20:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1290/3295 [13:19<20:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1291/3295 [13:19<20:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1291/3295 [13:19<20:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1292/3295 [13:21<20:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1292/3295 [13:21<20:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1293/3295 [13:21<20:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1293/3295 [13:21<20:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1294/3295 [13:21<20:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1294/3295 [13:21<20:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1295/3295 [13:23<20:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1295/3295 [13:23<20:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1296/3295 [13:23<20:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1296/3295 [13:23<20:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1297/3295 [13:23<20:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1297/3295 [13:23<20:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1298/3295 [13:25<20:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1298/3295 [13:25<20:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1299/3295 [13:25<20:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1299/3295 [13:25<20:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1300/3295 [13:25<20:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1300/3295 [13:25<20:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1301/3295 [13:26<20:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  39%|███▉      | 1301/3295 [13:26<20:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1302/3295 [13:26<20:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1302/3295 [13:26<20:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1303/3295 [13:26<20:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1303/3295 [13:26<20:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1304/3295 [13:28<20:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1304/3295 [13:28<20:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1305/3295 [13:28<20:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1305/3295 [13:28<20:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1306/3295 [13:28<20:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1306/3295 [13:28<20:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1307/3295 [13:30<20:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1307/3295 [13:30<20:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1308/3295 [13:30<20:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1308/3295 [13:30<20:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1309/3295 [13:30<20:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1309/3295 [13:30<20:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1310/3295 [13:32<20:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1310/3295 [13:32<20:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1311/3295 [13:32<20:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1311/3295 [13:32<20:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1312/3295 [13:32<20:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1312/3295 [13:32<20:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1313/3295 [13:34<20:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1313/3295 [13:34<20:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1314/3295 [13:34<20:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1314/3295 [13:34<20:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1315/3295 [13:34<20:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1315/3295 [13:34<20:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1316/3295 [13:36<20:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1316/3295 [13:36<20:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1317/3295 [13:36<20:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|███▉      | 1317/3295 [13:36<20:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1318/3295 [13:36<20:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1318/3295 [13:36<20:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1319/3295 [13:38<20:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1319/3295 [13:38<20:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1320/3295 [13:38<20:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1320/3295 [13:38<20:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1321/3295 [13:38<20:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1321/3295 [13:38<20:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1322/3295 [13:40<20:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1322/3295 [13:40<20:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1323/3295 [13:40<20:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1323/3295 [13:40<20:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1324/3295 [13:40<20:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1324/3295 [13:40<20:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1325/3295 [13:41<20:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1325/3295 [13:41<20:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1326/3295 [13:41<20:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1326/3295 [13:41<20:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1327/3295 [13:42<20:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1327/3295 [13:42<20:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1328/3295 [13:43<20:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1328/3295 [13:43<20:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1329/3295 [13:43<20:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1329/3295 [13:43<20:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1330/3295 [13:43<20:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1330/3295 [13:43<20:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1331/3295 [13:45<20:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1331/3295 [13:45<20:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1332/3295 [13:45<20:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1332/3295 [13:45<20:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1333/3295 [13:45<20:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1333/3295 [13:45<20:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1334/3295 [13:47<20:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  40%|████      | 1334/3295 [13:47<20:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1335/3295 [13:47<20:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1335/3295 [13:47<20:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1336/3295 [13:47<20:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1336/3295 [13:47<20:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1337/3295 [13:49<20:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1337/3295 [13:49<20:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1338/3295 [13:49<20:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1338/3295 [13:49<20:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1339/3295 [13:49<20:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1339/3295 [13:49<20:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1340/3295 [13:51<20:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1340/3295 [13:51<20:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1341/3295 [13:51<20:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1341/3295 [13:51<20:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1342/3295 [13:51<20:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1342/3295 [13:51<20:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1343/3295 [13:53<20:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1343/3295 [13:53<20:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1344/3295 [13:53<20:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1344/3295 [13:53<20:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1345/3295 [13:53<20:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1345/3295 [13:53<20:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1346/3295 [13:55<20:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1346/3295 [13:55<20:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1347/3295 [13:55<20:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1347/3295 [13:55<20:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1348/3295 [13:55<20:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1348/3295 [13:55<20:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1349/3295 [13:56<20:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1349/3295 [13:56<20:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1350/3295 [13:56<20:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1350/3295 [13:56<20:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1351/3295 [13:56<20:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1351/3295 [13:56<20:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1352/3295 [13:58<20:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1352/3295 [13:58<20:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1353/3295 [13:58<20:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1353/3295 [13:58<20:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1354/3295 [13:58<20:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1354/3295 [13:58<20:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1355/3295 [14:00<20:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1355/3295 [14:00<20:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1356/3295 [14:00<20:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1356/3295 [14:00<20:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1357/3295 [14:00<20:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1357/3295 [14:00<20:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1358/3295 [14:02<20:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1358/3295 [14:02<20:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1359/3295 [14:02<19:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████      | 1359/3295 [14:02<19:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1360/3295 [14:02<19:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1360/3295 [14:02<19:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1361/3295 [14:04<19:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1361/3295 [14:04<19:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1362/3295 [14:04<19:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1362/3295 [14:04<19:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1363/3295 [14:04<19:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1363/3295 [14:04<19:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1364/3295 [14:05<19:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1364/3295 [14:05<19:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1365/3295 [14:05<19:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1365/3295 [14:05<19:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1366/3295 [14:05<19:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1366/3295 [14:05<19:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1367/3295 [14:07<19:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  41%|████▏     | 1367/3295 [14:07<19:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1368/3295 [14:07<19:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1368/3295 [14:07<19:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1369/3295 [14:07<19:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1369/3295 [14:07<19:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1370/3295 [14:09<19:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1370/3295 [14:09<19:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1371/3295 [14:09<19:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1371/3295 [14:09<19:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1372/3295 [14:09<19:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1372/3295 [14:09<19:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1373/3295 [14:11<19:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1373/3295 [14:11<19:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1374/3295 [14:11<19:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1374/3295 [14:11<19:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1375/3295 [14:11<19:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1375/3295 [14:11<19:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1376/3295 [14:13<19:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1376/3295 [14:13<19:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1377/3295 [14:13<19:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1377/3295 [14:13<19:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1378/3295 [14:13<19:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1378/3295 [14:13<19:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1379/3295 [14:15<19:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1379/3295 [14:15<19:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1380/3295 [14:15<19:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1380/3295 [14:15<19:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1381/3295 [14:15<19:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1381/3295 [14:15<19:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1382/3295 [14:17<19:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1382/3295 [14:17<19:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1383/3295 [14:17<19:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1383/3295 [14:17<19:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1384/3295 [14:17<19:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1384/3295 [14:17<19:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1385/3295 [14:19<19:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1385/3295 [14:19<19:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1386/3295 [14:19<19:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1386/3295 [14:19<19:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1387/3295 [14:19<19:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1387/3295 [14:19<19:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1388/3295 [14:20<19:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1388/3295 [14:20<19:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1389/3295 [14:20<19:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1389/3295 [14:20<19:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1390/3295 [14:20<19:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1390/3295 [14:20<19:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1391/3295 [14:22<19:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1391/3295 [14:22<19:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1392/3295 [14:22<19:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1392/3295 [14:22<19:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1393/3295 [14:22<19:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1393/3295 [14:22<19:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1394/3295 [14:24<19:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1394/3295 [14:24<19:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1395/3295 [14:24<19:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1395/3295 [14:24<19:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1396/3295 [14:24<19:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1396/3295 [14:24<19:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1397/3295 [14:26<19:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1397/3295 [14:26<19:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1398/3295 [14:26<19:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1398/3295 [14:26<19:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1399/3295 [14:26<19:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1399/3295 [14:26<19:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1400/3295 [14:28<19:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  42%|████▏     | 1400/3295 [14:28<19:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1401/3295 [14:28<19:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1401/3295 [14:28<19:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1402/3295 [14:28<19:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1402/3295 [14:28<19:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1403/3295 [14:30<19:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1403/3295 [14:30<19:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1404/3295 [14:30<19:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1404/3295 [14:30<19:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1405/3295 [14:30<19:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1405/3295 [14:30<19:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1406/3295 [14:32<19:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1406/3295 [14:32<19:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1407/3295 [14:32<19:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1407/3295 [14:32<19:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1408/3295 [14:32<19:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1408/3295 [14:32<19:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1409/3295 [14:34<19:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1409/3295 [14:34<19:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1410/3295 [14:34<19:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1410/3295 [14:34<19:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1411/3295 [14:34<19:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1411/3295 [14:34<19:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1412/3295 [14:35<19:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1412/3295 [14:35<19:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1413/3295 [14:36<19:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1413/3295 [14:36<19:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1414/3295 [14:36<19:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1414/3295 [14:36<19:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1415/3295 [14:37<19:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1415/3295 [14:37<19:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1416/3295 [14:38<19:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1416/3295 [14:38<19:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1417/3295 [14:38<19:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1417/3295 [14:38<19:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1418/3295 [14:39<19:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1418/3295 [14:39<19:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1419/3295 [14:39<19:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1419/3295 [14:39<19:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1420/3295 [14:39<19:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1420/3295 [14:39<19:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1421/3295 [14:41<19:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1421/3295 [14:41<19:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1422/3295 [14:41<19:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1422/3295 [14:41<19:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1423/3295 [14:41<19:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1423/3295 [14:41<19:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1424/3295 [14:43<19:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1424/3295 [14:43<19:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1425/3295 [14:43<19:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1425/3295 [14:43<19:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1426/3295 [14:43<19:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1426/3295 [14:43<19:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1427/3295 [14:45<19:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1427/3295 [14:45<19:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1428/3295 [14:45<19:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1428/3295 [14:45<19:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1429/3295 [14:45<19:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1429/3295 [14:45<19:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1430/3295 [14:47<19:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1430/3295 [14:47<19:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1431/3295 [14:47<19:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1431/3295 [14:47<19:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1432/3295 [14:47<19:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1432/3295 [14:47<19:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1433/3295 [14:49<19:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  43%|████▎     | 1433/3295 [14:49<19:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1434/3295 [14:49<19:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1434/3295 [14:49<19:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1435/3295 [14:49<19:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1435/3295 [14:49<19:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1436/3295 [14:50<19:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1436/3295 [14:50<19:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1437/3295 [14:50<19:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1437/3295 [14:50<19:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1438/3295 [14:51<19:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1438/3295 [14:51<19:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1439/3295 [14:52<19:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1439/3295 [14:52<19:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1440/3295 [14:52<19:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1440/3295 [14:52<19:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1441/3295 [14:52<19:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▎     | 1441/3295 [14:52<19:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1442/3295 [14:54<19:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1442/3295 [14:54<19:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1443/3295 [14:54<19:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1443/3295 [14:54<19:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1444/3295 [14:54<19:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1444/3295 [14:54<19:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1445/3295 [14:56<19:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1445/3295 [14:56<19:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1446/3295 [14:56<19:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1446/3295 [14:56<19:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1447/3295 [14:56<19:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1447/3295 [14:56<19:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1448/3295 [14:58<19:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1448/3295 [14:58<19:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1449/3295 [14:58<19:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1449/3295 [14:58<19:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1450/3295 [14:58<19:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1450/3295 [14:58<19:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1451/3295 [15:00<19:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1451/3295 [15:00<19:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1452/3295 [15:00<19:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1452/3295 [15:00<19:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1453/3295 [15:00<19:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1453/3295 [15:00<19:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1454/3295 [15:02<19:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1454/3295 [15:02<19:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1455/3295 [15:02<19:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1455/3295 [15:02<19:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1456/3295 [15:02<18:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1456/3295 [15:02<18:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1457/3295 [15:04<19:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1457/3295 [15:04<19:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1458/3295 [15:04<18:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1458/3295 [15:04<18:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1459/3295 [15:04<18:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1459/3295 [15:04<18:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1460/3295 [15:06<18:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1460/3295 [15:06<18:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1461/3295 [15:06<18:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1461/3295 [15:06<18:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1462/3295 [15:06<18:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1462/3295 [15:06<18:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1463/3295 [15:07<18:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1463/3295 [15:07<18:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1464/3295 [15:07<18:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1464/3295 [15:07<18:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1465/3295 [15:07<18:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1465/3295 [15:07<18:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1466/3295 [15:09<18:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  44%|████▍     | 1466/3295 [15:09<18:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1467/3295 [15:09<18:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1467/3295 [15:09<18:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1468/3295 [15:09<18:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1468/3295 [15:09<18:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1469/3295 [15:11<18:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1469/3295 [15:11<18:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1470/3295 [15:11<18:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1470/3295 [15:11<18:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1471/3295 [15:11<18:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1471/3295 [15:11<18:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1472/3295 [15:13<18:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1472/3295 [15:13<18:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1473/3295 [15:13<18:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1473/3295 [15:13<18:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1474/3295 [15:13<18:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1474/3295 [15:13<18:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1475/3295 [15:15<18:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1475/3295 [15:15<18:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1476/3295 [15:15<18:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1476/3295 [15:15<18:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1477/3295 [15:15<18:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1477/3295 [15:15<18:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1478/3295 [15:16<18:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1478/3295 [15:16<18:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1479/3295 [15:16<18:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1479/3295 [15:16<18:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1480/3295 [15:16<18:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1480/3295 [15:16<18:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1481/3295 [15:18<18:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1481/3295 [15:18<18:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1482/3295 [15:18<18:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▍     | 1482/3295 [15:18<18:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1483/3295 [15:18<18:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1483/3295 [15:18<18:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1484/3295 [15:20<18:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1484/3295 [15:20<18:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1485/3295 [15:20<18:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1485/3295 [15:20<18:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1486/3295 [15:20<18:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1486/3295 [15:20<18:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1487/3295 [15:22<18:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1487/3295 [15:22<18:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1488/3295 [15:22<18:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1488/3295 [15:22<18:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1489/3295 [15:22<18:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1489/3295 [15:22<18:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1490/3295 [15:23<18:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1490/3295 [15:23<18:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1491/3295 [15:23<18:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1491/3295 [15:23<18:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1492/3295 [15:23<18:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1492/3295 [15:23<18:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1493/3295 [15:25<18:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1493/3295 [15:25<18:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1494/3295 [15:25<18:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1494/3295 [15:25<18:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1495/3295 [15:25<18:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1495/3295 [15:25<18:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1496/3295 [15:27<18:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1496/3295 [15:27<18:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1497/3295 [15:27<18:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1497/3295 [15:27<18:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1498/3295 [15:27<18:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1498/3295 [15:27<18:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1499/3295 [15:29<18:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  45%|████▌     | 1499/3295 [15:29<18:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1500/3295 [15:29<18:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1500/3295 [15:29<18:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1501/3295 [15:29<18:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1501/3295 [15:29<18:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1502/3295 [15:31<18:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1502/3295 [15:31<18:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1503/3295 [15:31<18:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1503/3295 [15:31<18:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1504/3295 [15:31<18:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1504/3295 [15:31<18:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1505/3295 [15:33<18:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1505/3295 [15:33<18:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1506/3295 [15:33<18:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1506/3295 [15:33<18:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1507/3295 [15:33<18:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1507/3295 [15:33<18:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1508/3295 [15:35<18:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1508/3295 [15:35<18:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1509/3295 [15:35<18:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1509/3295 [15:35<18:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1510/3295 [15:35<18:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1510/3295 [15:35<18:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1511/3295 [15:36<18:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1511/3295 [15:36<18:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1512/3295 [15:36<18:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1512/3295 [15:36<18:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1513/3295 [15:36<18:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1513/3295 [15:36<18:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1514/3295 [15:38<18:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1514/3295 [15:38<18:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1515/3295 [15:38<18:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1515/3295 [15:38<18:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1516/3295 [15:38<18:21,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1516/3295 [15:38<18:21,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1517/3295 [15:40<18:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1517/3295 [15:40<18:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1518/3295 [15:40<18:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1518/3295 [15:40<18:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1519/3295 [15:40<18:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1519/3295 [15:40<18:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1520/3295 [15:42<18:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1520/3295 [15:42<18:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1521/3295 [15:42<18:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1521/3295 [15:42<18:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1522/3295 [15:42<18:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1522/3295 [15:42<18:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1523/3295 [15:44<18:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▌     | 1523/3295 [15:44<18:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1524/3295 [15:44<18:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1524/3295 [15:44<18:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1525/3295 [15:44<18:15,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1525/3295 [15:44<18:15,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1526/3295 [15:45<18:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1526/3295 [15:45<18:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1527/3295 [15:45<18:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1527/3295 [15:45<18:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1528/3295 [15:45<18:13,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1528/3295 [15:45<18:13,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1529/3295 [15:47<18:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1529/3295 [15:47<18:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1530/3295 [15:47<18:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1530/3295 [15:47<18:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1531/3295 [15:47<18:11,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1531/3295 [15:47<18:12,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1532/3295 [15:49<18:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  46%|████▋     | 1532/3295 [15:49<18:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1533/3295 [15:49<18:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1533/3295 [15:49<18:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1534/3295 [15:49<18:10,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1534/3295 [15:49<18:10,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1535/3295 [15:51<18:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1535/3295 [15:51<18:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1536/3295 [15:51<18:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1536/3295 [15:51<18:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1537/3295 [15:51<18:08,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1537/3295 [15:51<18:08,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1538/3295 [15:53<18:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1538/3295 [15:53<18:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1539/3295 [15:53<18:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1539/3295 [15:53<18:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1540/3295 [15:53<18:06,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1540/3295 [15:53<18:06,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1541/3295 [15:55<18:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1541/3295 [15:55<18:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1542/3295 [15:55<18:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1542/3295 [15:55<18:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1543/3295 [15:55<18:04,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1543/3295 [15:55<18:04,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1544/3295 [15:57<18:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1544/3295 [15:57<18:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1545/3295 [15:57<18:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1545/3295 [15:57<18:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1546/3295 [15:57<18:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1546/3295 [15:57<18:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1547/3295 [15:59<18:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1547/3295 [15:59<18:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1548/3295 [15:59<18:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1548/3295 [15:59<18:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1549/3295 [15:59<18:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1549/3295 [15:59<18:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1550/3295 [16:00<18:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1550/3295 [16:00<18:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1551/3295 [16:01<18:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1551/3295 [16:01<18:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1552/3295 [16:01<17:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1552/3295 [16:01<17:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1553/3295 [16:02<17:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1553/3295 [16:02<17:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1554/3295 [16:02<17:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1554/3295 [16:02<17:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1555/3295 [16:02<17:57,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1555/3295 [16:02<17:57,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1556/3295 [16:04<17:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1556/3295 [16:04<17:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1557/3295 [16:04<17:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1557/3295 [16:04<17:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1558/3295 [16:04<17:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1558/3295 [16:04<17:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1559/3295 [16:06<17:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1559/3295 [16:06<17:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1560/3295 [16:06<17:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1560/3295 [16:06<17:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1561/3295 [16:06<17:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1561/3295 [16:06<17:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1562/3295 [16:08<17:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1562/3295 [16:08<17:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1563/3295 [16:08<17:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1563/3295 [16:08<17:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1564/3295 [16:08<17:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1564/3295 [16:08<17:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1565/3295 [16:10<17:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  47%|████▋     | 1565/3295 [16:10<17:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1566/3295 [16:10<17:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1566/3295 [16:10<17:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1567/3295 [16:10<17:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1567/3295 [16:10<17:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1568/3295 [16:12<17:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1568/3295 [16:12<17:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1569/3295 [16:12<17:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1569/3295 [16:12<17:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1570/3295 [16:12<17:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1570/3295 [16:12<17:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1571/3295 [16:14<17:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1571/3295 [16:14<17:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1572/3295 [16:14<17:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1572/3295 [16:14<17:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1573/3295 [16:14<17:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1573/3295 [16:14<17:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1574/3295 [16:15<17:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1574/3295 [16:15<17:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1575/3295 [16:15<17:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1575/3295 [16:15<17:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1576/3295 [16:15<17:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1576/3295 [16:15<17:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1577/3295 [16:17<17:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1577/3295 [16:17<17:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1578/3295 [16:17<17:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1578/3295 [16:17<17:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1579/3295 [16:17<17:42,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1579/3295 [16:17<17:42,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1580/3295 [16:19<17:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1580/3295 [16:19<17:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1581/3295 [16:19<17:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1581/3295 [16:19<17:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1582/3295 [16:19<17:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1582/3295 [16:19<17:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1583/3295 [16:21<17:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1583/3295 [16:21<17:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1584/3295 [16:21<17:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1584/3295 [16:21<17:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1585/3295 [16:21<17:38,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1585/3295 [16:21<17:38,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1586/3295 [16:23<17:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1586/3295 [16:23<17:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1587/3295 [16:23<17:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1587/3295 [16:23<17:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1588/3295 [16:23<17:36,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1588/3295 [16:23<17:36,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1589/3295 [16:25<17:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1589/3295 [16:25<17:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1590/3295 [16:25<17:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1590/3295 [16:25<17:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1591/3295 [16:25<17:35,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1591/3295 [16:25<17:35,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1592/3295 [16:26<17:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1592/3295 [16:26<17:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1593/3295 [16:27<17:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1593/3295 [16:27<17:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1594/3295 [16:27<17:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1594/3295 [16:27<17:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1595/3295 [16:28<17:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1595/3295 [16:28<17:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1596/3295 [16:28<17:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1596/3295 [16:28<17:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1597/3295 [16:28<17:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1597/3295 [16:28<17:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1598/3295 [16:30<17:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  48%|████▊     | 1598/3295 [16:30<17:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1599/3295 [16:30<17:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1599/3295 [16:30<17:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1600/3295 [16:30<17:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1600/3295 [16:30<17:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1601/3295 [16:32<17:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1601/3295 [16:32<17:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1602/3295 [16:32<17:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1602/3295 [16:32<17:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1603/3295 [16:32<17:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1603/3295 [16:32<17:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1604/3295 [16:34<17:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1604/3295 [16:34<17:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1605/3295 [16:34<17:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1605/3295 [16:34<17:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1606/3295 [16:34<17:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▊     | 1606/3295 [16:34<17:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1607/3295 [16:36<17:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1607/3295 [16:36<17:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1608/3295 [16:36<17:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1608/3295 [16:36<17:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1609/3295 [16:36<17:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1609/3295 [16:36<17:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1610/3295 [16:38<17:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1610/3295 [16:38<17:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1611/3295 [16:38<17:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1611/3295 [16:38<17:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1612/3295 [16:38<17:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1612/3295 [16:38<17:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1613/3295 [16:40<17:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1613/3295 [16:40<17:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1614/3295 [16:40<17:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1614/3295 [16:40<17:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1615/3295 [16:40<17:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1615/3295 [16:40<17:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1616/3295 [16:42<17:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1616/3295 [16:42<17:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1617/3295 [16:42<17:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1617/3295 [16:42<17:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1618/3295 [16:42<17:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1618/3295 [16:42<17:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1619/3295 [16:43<17:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1619/3295 [16:43<17:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1620/3295 [16:43<17:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1620/3295 [16:43<17:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1621/3295 [16:43<17:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1621/3295 [16:43<17:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1622/3295 [16:45<17:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1622/3295 [16:45<17:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1623/3295 [16:45<17:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1623/3295 [16:45<17:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1624/3295 [16:45<17:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1624/3295 [16:45<17:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1625/3295 [16:47<17:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1625/3295 [16:47<17:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1626/3295 [16:47<17:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1626/3295 [16:47<17:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1627/3295 [16:48<17:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1627/3295 [16:48<17:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1628/3295 [16:49<17:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1628/3295 [16:49<17:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1629/3295 [16:49<17:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1629/3295 [16:49<17:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1630/3295 [16:49<17:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1630/3295 [16:49<17:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1631/3295 [16:51<17:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  49%|████▉     | 1631/3295 [16:51<17:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1632/3295 [16:51<17:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1632/3295 [16:51<17:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1633/3295 [16:51<17:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1633/3295 [16:51<17:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1634/3295 [16:53<17:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1634/3295 [16:53<17:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1635/3295 [16:53<17:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1635/3295 [16:53<17:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1636/3295 [16:53<17:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1636/3295 [16:53<17:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1637/3295 [16:55<17:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1637/3295 [16:55<17:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1638/3295 [16:55<17:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1638/3295 [16:55<17:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1639/3295 [16:55<17:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1639/3295 [16:55<17:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1640/3295 [16:57<17:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1640/3295 [16:57<17:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1641/3295 [16:57<17:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1641/3295 [16:57<17:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1642/3295 [16:57<17:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1642/3295 [16:57<17:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1643/3295 [16:58<17:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1643/3295 [16:58<17:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1644/3295 [16:58<17:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1644/3295 [16:58<17:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1645/3295 [16:58<17:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1645/3295 [16:58<17:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1646/3295 [17:00<17:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1646/3295 [17:00<17:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1647/3295 [17:00<17:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|████▉     | 1647/3295 [17:00<17:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1648/3295 [17:00<17:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1648/3295 [17:00<17:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1649/3295 [17:02<17:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1649/3295 [17:02<17:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1650/3295 [17:02<16:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1650/3295 [17:02<16:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1651/3295 [17:02<16:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1651/3295 [17:02<16:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1652/3295 [17:04<16:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1652/3295 [17:04<16:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1653/3295 [17:04<16:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1653/3295 [17:04<16:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1654/3295 [17:04<16:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1654/3295 [17:04<16:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1655/3295 [17:05<16:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1655/3295 [17:05<16:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1656/3295 [17:06<16:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1656/3295 [17:06<16:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1657/3295 [17:06<16:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1657/3295 [17:06<16:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1658/3295 [17:07<16:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1658/3295 [17:07<16:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1659/3295 [17:07<16:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1659/3295 [17:07<16:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1660/3295 [17:07<16:52,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1660/3295 [17:07<16:52,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1661/3295 [17:09<16:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1661/3295 [17:09<16:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1662/3295 [17:09<16:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1662/3295 [17:09<16:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1663/3295 [17:09<16:50,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  50%|█████     | 1663/3295 [17:09<16:50,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1664/3295 [17:11<16:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1664/3295 [17:11<16:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1665/3295 [17:11<16:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1665/3295 [17:11<16:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1666/3295 [17:11<16:48,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1666/3295 [17:11<16:48,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1667/3295 [17:13<16:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1667/3295 [17:13<16:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1668/3295 [17:13<16:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1668/3295 [17:13<16:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1669/3295 [17:13<16:46,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1669/3295 [17:13<16:46,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1670/3295 [17:15<16:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1670/3295 [17:15<16:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1671/3295 [17:15<16:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1671/3295 [17:15<16:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1672/3295 [17:15<16:44,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1672/3295 [17:15<16:44,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1673/3295 [17:16<16:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1673/3295 [17:16<16:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1674/3295 [17:16<16:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1674/3295 [17:16<16:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1675/3295 [17:16<16:42,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1675/3295 [17:16<16:42,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1676/3295 [17:18<16:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1676/3295 [17:18<16:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1677/3295 [17:18<16:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1677/3295 [17:18<16:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1678/3295 [17:18<16:41,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1678/3295 [17:18<16:41,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1679/3295 [17:20<16:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1679/3295 [17:20<16:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1680/3295 [17:20<16:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1680/3295 [17:20<16:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1681/3295 [17:20<16:39,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1681/3295 [17:20<16:39,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1682/3295 [17:22<16:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1682/3295 [17:22<16:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1683/3295 [17:22<16:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1683/3295 [17:22<16:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1684/3295 [17:22<16:37,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1684/3295 [17:22<16:37,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1685/3295 [17:24<16:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1685/3295 [17:24<16:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1686/3295 [17:24<16:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1686/3295 [17:24<16:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1687/3295 [17:24<16:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1687/3295 [17:24<16:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1688/3295 [17:26<16:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████     | 1688/3295 [17:26<16:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1689/3295 [17:26<16:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1689/3295 [17:26<16:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1690/3295 [17:26<16:33,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1690/3295 [17:26<16:33,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1691/3295 [17:28<16:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1691/3295 [17:28<16:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1692/3295 [17:28<16:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1692/3295 [17:28<16:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1693/3295 [17:28<16:31,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1693/3295 [17:28<16:31,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1694/3295 [17:29<16:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1694/3295 [17:29<16:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1695/3295 [17:29<16:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1695/3295 [17:29<16:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1696/3295 [17:29<16:29,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  51%|█████▏    | 1696/3295 [17:29<16:29,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1697/3295 [17:31<16:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1697/3295 [17:31<16:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1698/3295 [17:31<16:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1698/3295 [17:31<16:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1699/3295 [17:31<16:28,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1699/3295 [17:31<16:28,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1700/3295 [17:33<16:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1700/3295 [17:33<16:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1701/3295 [17:33<16:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1701/3295 [17:33<16:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1702/3295 [17:33<16:26,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1702/3295 [17:33<16:26,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1703/3295 [17:35<16:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1703/3295 [17:35<16:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1704/3295 [17:35<16:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1704/3295 [17:35<16:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1705/3295 [17:35<16:24,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1705/3295 [17:35<16:24,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1706/3295 [17:37<16:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1706/3295 [17:37<16:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1707/3295 [17:37<16:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1707/3295 [17:37<16:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1708/3295 [17:37<16:22,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1708/3295 [17:37<16:22,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1709/3295 [17:39<16:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1709/3295 [17:39<16:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1710/3295 [17:39<16:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1710/3295 [17:39<16:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1711/3295 [17:39<16:20,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1711/3295 [17:39<16:20,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1712/3295 [17:41<16:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1712/3295 [17:41<16:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1713/3295 [17:41<16:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1713/3295 [17:41<16:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1714/3295 [17:41<16:18,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1714/3295 [17:41<16:18,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1715/3295 [17:42<16:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1715/3295 [17:42<16:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1716/3295 [17:42<16:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1716/3295 [17:42<16:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1717/3295 [17:42<16:16,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1717/3295 [17:42<16:16,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1718/3295 [17:44<16:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1718/3295 [17:44<16:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1719/3295 [17:44<16:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1719/3295 [17:44<16:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1720/3295 [17:44<16:15,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1720/3295 [17:44<16:15,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1721/3295 [17:46<16:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1721/3295 [17:46<16:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1722/3295 [17:46<16:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1722/3295 [17:46<16:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1723/3295 [17:46<16:13,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1723/3295 [17:46<16:13,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1724/3295 [17:48<16:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1724/3295 [17:48<16:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1725/3295 [17:48<16:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1725/3295 [17:48<16:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1726/3295 [17:48<16:11,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1726/3295 [17:48<16:11,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1727/3295 [17:50<16:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1727/3295 [17:50<16:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1728/3295 [17:50<16:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1728/3295 [17:50<16:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1729/3295 [17:50<16:09,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  52%|█████▏    | 1729/3295 [17:50<16:09,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1730/3295 [17:52<16:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1730/3295 [17:52<16:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1731/3295 [17:52<16:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1731/3295 [17:52<16:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1732/3295 [17:52<16:07,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1732/3295 [17:52<16:07,  1.62it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1733/3295 [17:54<16:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1733/3295 [17:54<16:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1734/3295 [17:54<16:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1734/3295 [17:54<16:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1735/3295 [17:54<16:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1735/3295 [17:54<16:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1736/3295 [17:56<16:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1736/3295 [17:56<16:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1737/3295 [17:56<16:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1737/3295 [17:56<16:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1738/3295 [17:56<16:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1738/3295 [17:56<16:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1739/3295 [17:58<16:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1739/3295 [17:58<16:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1740/3295 [17:58<16:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1740/3295 [17:58<16:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1741/3295 [17:58<16:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1741/3295 [17:58<16:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1742/3295 [18:00<16:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1742/3295 [18:00<16:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1743/3295 [18:00<16:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1743/3295 [18:00<16:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1744/3295 [18:00<16:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1744/3295 [18:00<16:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1745/3295 [18:02<16:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1745/3295 [18:02<16:01,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1746/3295 [18:02<16:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1746/3295 [18:02<16:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1747/3295 [18:02<15:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1747/3295 [18:02<15:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1748/3295 [18:04<15:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1748/3295 [18:04<15:59,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1749/3295 [18:04<15:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1749/3295 [18:04<15:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1750/3295 [18:04<15:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1750/3295 [18:04<15:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1751/3295 [18:07<15:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1751/3295 [18:07<15:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1752/3295 [18:07<15:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1752/3295 [18:07<15:57,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1753/3295 [18:07<15:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1753/3295 [18:07<15:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1754/3295 [18:09<15:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1754/3295 [18:09<15:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1755/3295 [18:09<15:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1755/3295 [18:09<15:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1756/3295 [18:09<15:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1756/3295 [18:09<15:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1757/3295 [18:10<15:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1757/3295 [18:10<15:54,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1758/3295 [18:10<15:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1758/3295 [18:10<15:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1759/3295 [18:11<15:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1759/3295 [18:11<15:52,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1760/3295 [18:12<15:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1760/3295 [18:12<15:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1761/3295 [18:12<15:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1761/3295 [18:12<15:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1762/3295 [18:12<15:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  53%|█████▎    | 1762/3295 [18:12<15:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1763/3295 [18:14<15:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1763/3295 [18:14<15:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1764/3295 [18:14<15:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1764/3295 [18:14<15:50,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1765/3295 [18:14<15:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1765/3295 [18:14<15:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1766/3295 [18:16<15:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1766/3295 [18:16<15:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1767/3295 [18:16<15:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1767/3295 [18:16<15:48,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1768/3295 [18:16<15:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1768/3295 [18:16<15:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1769/3295 [18:18<15:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1769/3295 [18:18<15:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1770/3295 [18:18<15:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1770/3295 [18:18<15:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1771/3295 [18:18<15:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▎    | 1771/3295 [18:18<15:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1772/3295 [18:20<15:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1772/3295 [18:20<15:46,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1773/3295 [18:21<15:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1773/3295 [18:21<15:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1774/3295 [18:21<15:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1774/3295 [18:21<15:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1775/3295 [18:22<15:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1775/3295 [18:22<15:44,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1776/3295 [18:22<15:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1776/3295 [18:22<15:43,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1777/3295 [18:22<15:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1777/3295 [18:22<15:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1778/3295 [18:24<15:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1778/3295 [18:24<15:42,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1779/3295 [18:24<15:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1779/3295 [18:24<15:41,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1780/3295 [18:24<15:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1780/3295 [18:24<15:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1781/3295 [18:26<15:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1781/3295 [18:26<15:40,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1782/3295 [18:26<15:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1782/3295 [18:26<15:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1783/3295 [18:26<15:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1783/3295 [18:26<15:38,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1784/3295 [18:28<15:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1784/3295 [18:28<15:39,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1785/3295 [18:28<15:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1785/3295 [18:28<15:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1786/3295 [18:28<15:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1786/3295 [18:28<15:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1787/3295 [18:31<15:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1787/3295 [18:31<15:37,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1788/3295 [18:31<15:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1788/3295 [18:31<15:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1789/3295 [18:31<15:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1789/3295 [18:31<15:35,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1790/3295 [18:33<15:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1790/3295 [18:33<15:36,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1791/3295 [18:33<15:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1791/3295 [18:33<15:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1792/3295 [18:33<15:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1792/3295 [18:33<15:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1793/3295 [18:35<15:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1793/3295 [18:35<15:34,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1794/3295 [18:35<15:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1794/3295 [18:35<15:33,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1795/3295 [18:35<15:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  54%|█████▍    | 1795/3295 [18:35<15:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1796/3295 [18:37<15:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1796/3295 [18:37<15:32,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1797/3295 [18:37<15:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1797/3295 [18:37<15:31,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1798/3295 [18:37<15:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1798/3295 [18:37<15:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1799/3295 [18:39<15:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1799/3295 [18:39<15:30,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1800/3295 [18:39<15:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1800/3295 [18:39<15:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1801/3295 [18:39<15:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1801/3295 [18:39<15:28,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1802/3295 [18:41<15:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1802/3295 [18:41<15:29,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1803/3295 [18:41<15:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1803/3295 [18:41<15:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1804/3295 [18:41<15:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1804/3295 [18:41<15:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1805/3295 [18:43<15:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1805/3295 [18:43<15:27,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1806/3295 [18:43<15:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1806/3295 [18:43<15:26,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1807/3295 [18:43<15:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1807/3295 [18:43<15:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1808/3295 [18:45<15:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1808/3295 [18:45<15:25,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1809/3295 [18:45<15:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1809/3295 [18:45<15:24,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1810/3295 [18:45<15:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1810/3295 [18:45<15:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1811/3295 [18:47<15:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1811/3295 [18:47<15:23,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1812/3295 [18:47<15:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▍    | 1812/3295 [18:47<15:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1813/3295 [18:47<15:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1813/3295 [18:47<15:21,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1814/3295 [18:49<15:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1814/3295 [18:49<15:22,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1815/3295 [18:49<15:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1815/3295 [18:49<15:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1816/3295 [18:49<15:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1816/3295 [18:49<15:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1817/3295 [18:51<15:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1817/3295 [18:51<15:20,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1818/3295 [18:51<15:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1818/3295 [18:51<15:19,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1819/3295 [18:51<15:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1819/3295 [18:51<15:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1820/3295 [18:53<15:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1820/3295 [18:53<15:18,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1821/3295 [18:53<15:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1821/3295 [18:53<15:17,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1822/3295 [18:53<15:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1822/3295 [18:53<15:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1823/3295 [18:55<15:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1823/3295 [18:55<15:16,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1824/3295 [18:55<15:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1824/3295 [18:55<15:15,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1825/3295 [18:55<15:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1825/3295 [18:55<15:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1826/3295 [18:57<15:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1826/3295 [18:57<15:14,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1827/3295 [18:57<15:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1827/3295 [18:57<15:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1828/3295 [18:57<15:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  55%|█████▌    | 1828/3295 [18:57<15:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1829/3295 [18:59<15:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1829/3295 [18:59<15:13,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1830/3295 [18:59<15:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1830/3295 [18:59<15:12,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1831/3295 [18:59<15:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1831/3295 [18:59<15:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1832/3295 [19:01<15:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1832/3295 [19:01<15:11,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1833/3295 [19:01<15:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1833/3295 [19:01<15:10,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1834/3295 [19:01<15:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1834/3295 [19:01<15:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1835/3295 [19:03<15:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1835/3295 [19:03<15:09,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1836/3295 [19:03<15:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1836/3295 [19:03<15:08,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1837/3295 [19:03<15:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1837/3295 [19:03<15:07,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1838/3295 [19:05<15:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1838/3295 [19:05<15:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1839/3295 [19:05<15:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1839/3295 [19:05<15:06,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1840/3295 [19:05<15:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1840/3295 [19:05<15:05,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1841/3295 [19:07<15:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1841/3295 [19:07<15:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1842/3295 [19:07<15:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1842/3295 [19:07<15:04,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1843/3295 [19:07<15:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1843/3295 [19:07<15:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1844/3295 [19:09<15:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1844/3295 [19:09<15:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1845/3295 [19:09<15:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1845/3295 [19:09<15:03,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1846/3295 [19:09<15:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1846/3295 [19:09<15:02,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1847/3295 [19:11<15:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1847/3295 [19:11<15:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1848/3295 [19:11<15:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1848/3295 [19:11<15:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1849/3295 [19:11<15:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1849/3295 [19:11<15:00,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1850/3295 [19:13<15:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1850/3295 [19:13<15:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1851/3295 [19:13<14:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1851/3295 [19:13<14:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1852/3295 [19:13<14:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1852/3295 [19:13<14:58,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1853/3295 [19:15<14:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▌    | 1853/3295 [19:15<14:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1854/3295 [19:15<14:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1854/3295 [19:15<14:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1855/3295 [19:15<14:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1855/3295 [19:15<14:56,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1856/3295 [19:17<14:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1856/3295 [19:17<14:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1857/3295 [19:17<14:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1857/3295 [19:17<14:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1858/3295 [19:17<14:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1858/3295 [19:17<14:55,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1859/3295 [19:19<14:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1859/3295 [19:19<14:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1860/3295 [19:19<14:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1860/3295 [19:19<14:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1861/3295 [19:19<14:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  56%|█████▋    | 1861/3295 [19:19<14:53,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1862/3295 [19:21<14:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1862/3295 [19:21<14:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1863/3295 [19:21<14:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1863/3295 [19:21<14:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1864/3295 [19:21<14:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1864/3295 [19:21<14:51,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1865/3295 [19:23<14:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1865/3295 [19:23<14:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1866/3295 [19:23<14:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1866/3295 [19:23<14:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1867/3295 [19:23<14:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1867/3295 [19:23<14:49,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1868/3295 [19:24<14:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1868/3295 [19:24<14:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1869/3295 [19:24<14:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1869/3295 [19:24<14:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1870/3295 [19:24<14:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1870/3295 [19:24<14:47,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1871/3295 [19:26<14:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1871/3295 [19:26<14:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1872/3295 [19:26<14:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1872/3295 [19:26<14:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1873/3295 [19:26<14:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1873/3295 [19:26<14:45,  1.61it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1874/3295 [19:28<14:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1874/3295 [19:28<14:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1875/3295 [19:28<14:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1875/3295 [19:28<14:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1876/3295 [19:28<14:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1876/3295 [19:28<14:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1877/3295 [19:31<14:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1877/3295 [19:31<14:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1878/3295 [19:31<14:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1878/3295 [19:31<14:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1879/3295 [19:31<14:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1879/3295 [19:31<14:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1880/3295 [19:33<14:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1880/3295 [19:33<14:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1881/3295 [19:33<14:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1881/3295 [19:33<14:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1882/3295 [19:33<14:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1882/3295 [19:33<14:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1883/3295 [19:34<14:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1883/3295 [19:34<14:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1884/3295 [19:34<14:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1884/3295 [19:34<14:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1885/3295 [19:34<14:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1885/3295 [19:34<14:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1886/3295 [19:36<14:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1886/3295 [19:36<14:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1887/3295 [19:36<14:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1887/3295 [19:36<14:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1888/3295 [19:36<14:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1888/3295 [19:36<14:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1889/3295 [19:38<14:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1889/3295 [19:38<14:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1890/3295 [19:38<14:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1890/3295 [19:38<14:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1891/3295 [19:38<14:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1891/3295 [19:38<14:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1892/3295 [19:40<14:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1892/3295 [19:40<14:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1893/3295 [19:40<14:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1893/3295 [19:40<14:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1894/3295 [19:40<14:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  57%|█████▋    | 1894/3295 [19:40<14:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1895/3295 [19:42<14:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1895/3295 [19:42<14:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1896/3295 [19:42<14:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1896/3295 [19:42<14:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1897/3295 [19:42<14:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1897/3295 [19:42<14:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1898/3295 [19:44<14:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1898/3295 [19:44<14:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1899/3295 [19:44<14:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1899/3295 [19:44<14:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1900/3295 [19:44<14:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1900/3295 [19:44<14:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1901/3295 [19:46<14:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1901/3295 [19:46<14:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1902/3295 [19:46<14:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1902/3295 [19:46<14:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1903/3295 [19:46<14:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1903/3295 [19:46<14:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1904/3295 [19:48<14:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1904/3295 [19:48<14:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1905/3295 [19:48<14:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1905/3295 [19:48<14:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1906/3295 [19:48<14:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1906/3295 [19:48<14:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1907/3295 [19:50<14:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1907/3295 [19:50<14:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1908/3295 [19:50<14:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1908/3295 [19:50<14:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1909/3295 [19:50<14:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1909/3295 [19:50<14:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1910/3295 [19:52<14:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1910/3295 [19:52<14:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1911/3295 [19:52<14:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1911/3295 [19:52<14:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1912/3295 [19:52<14:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1912/3295 [19:52<14:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1913/3295 [19:54<14:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1913/3295 [19:54<14:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1914/3295 [19:54<14:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1914/3295 [19:54<14:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1915/3295 [19:54<14:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1915/3295 [19:54<14:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1916/3295 [19:56<14:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1916/3295 [19:56<14:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1917/3295 [19:56<14:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1917/3295 [19:56<14:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1918/3295 [19:56<14:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1918/3295 [19:56<14:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1919/3295 [19:58<14:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1919/3295 [19:58<14:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1920/3295 [19:58<14:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1920/3295 [19:58<14:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1921/3295 [19:58<14:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1921/3295 [19:58<14:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1922/3295 [19:59<14:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1922/3295 [19:59<14:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1923/3295 [19:59<14:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1923/3295 [19:59<14:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1924/3295 [19:59<14:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1924/3295 [19:59<14:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1925/3295 [20:01<14:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1925/3295 [20:01<14:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1926/3295 [20:01<14:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1926/3295 [20:01<14:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1927/3295 [20:01<14:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  58%|█████▊    | 1927/3295 [20:01<14:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1928/3295 [20:03<14:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1928/3295 [20:03<14:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1929/3295 [20:03<14:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1929/3295 [20:03<14:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1930/3295 [20:03<14:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1930/3295 [20:03<14:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1931/3295 [20:05<14:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1931/3295 [20:05<14:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1932/3295 [20:05<14:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1932/3295 [20:05<14:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1933/3295 [20:05<14:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1933/3295 [20:05<14:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1934/3295 [20:07<14:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1934/3295 [20:07<14:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1935/3295 [20:07<14:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▊    | 1935/3295 [20:07<14:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1936/3295 [20:07<14:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1936/3295 [20:07<14:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1937/3295 [20:09<14:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1937/3295 [20:09<14:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1938/3295 [20:09<14:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1938/3295 [20:09<14:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1939/3295 [20:10<14:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1939/3295 [20:10<14:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1940/3295 [20:11<14:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1940/3295 [20:11<14:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1941/3295 [20:11<14:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1941/3295 [20:11<14:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1942/3295 [20:11<14:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1942/3295 [20:11<14:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1943/3295 [20:13<14:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1943/3295 [20:13<14:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1944/3295 [20:13<14:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1944/3295 [20:13<14:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1945/3295 [20:13<14:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1945/3295 [20:13<14:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1946/3295 [20:15<14:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1946/3295 [20:15<14:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1947/3295 [20:15<14:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1947/3295 [20:15<14:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1948/3295 [20:15<14:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1948/3295 [20:15<14:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1949/3295 [20:17<14:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1949/3295 [20:17<14:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1950/3295 [20:17<13:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1950/3295 [20:17<13:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1951/3295 [20:17<13:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1951/3295 [20:17<13:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1952/3295 [20:19<13:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1952/3295 [20:19<13:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1953/3295 [20:19<13:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1953/3295 [20:19<13:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1954/3295 [20:19<13:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1954/3295 [20:19<13:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1955/3295 [20:21<13:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1955/3295 [20:21<13:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1956/3295 [20:21<13:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1956/3295 [20:21<13:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1957/3295 [20:21<13:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1957/3295 [20:21<13:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1958/3295 [20:23<13:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1958/3295 [20:23<13:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1959/3295 [20:23<13:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1959/3295 [20:23<13:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1960/3295 [20:23<13:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  59%|█████▉    | 1960/3295 [20:23<13:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1961/3295 [20:25<13:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1961/3295 [20:25<13:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1962/3295 [20:25<13:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1962/3295 [20:25<13:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1963/3295 [20:25<13:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1963/3295 [20:25<13:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1964/3295 [20:26<13:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1964/3295 [20:26<13:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1965/3295 [20:26<13:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1965/3295 [20:26<13:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1966/3295 [20:26<13:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1966/3295 [20:26<13:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1967/3295 [20:28<13:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1967/3295 [20:28<13:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1968/3295 [20:28<13:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1968/3295 [20:28<13:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1969/3295 [20:28<13:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1969/3295 [20:28<13:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1970/3295 [20:30<13:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1970/3295 [20:30<13:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1971/3295 [20:30<13:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1971/3295 [20:30<13:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1972/3295 [20:30<13:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1972/3295 [20:30<13:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1973/3295 [20:32<13:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1973/3295 [20:32<13:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1974/3295 [20:32<13:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1974/3295 [20:32<13:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1975/3295 [20:32<13:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1975/3295 [20:32<13:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1976/3295 [20:34<13:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|█████▉    | 1976/3295 [20:34<13:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1977/3295 [20:34<13:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1977/3295 [20:34<13:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1978/3295 [20:34<13:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1978/3295 [20:34<13:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1979/3295 [20:36<13:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1979/3295 [20:36<13:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1980/3295 [20:36<13:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1980/3295 [20:36<13:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1981/3295 [20:36<13:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1981/3295 [20:36<13:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1982/3295 [20:38<13:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1982/3295 [20:38<13:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1983/3295 [20:38<13:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1983/3295 [20:38<13:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1984/3295 [20:38<13:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1984/3295 [20:38<13:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1985/3295 [20:40<13:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1985/3295 [20:40<13:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1986/3295 [20:40<13:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1986/3295 [20:40<13:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1987/3295 [20:40<13:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1987/3295 [20:40<13:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1988/3295 [20:42<13:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1988/3295 [20:42<13:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1989/3295 [20:42<13:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1989/3295 [20:42<13:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1990/3295 [20:42<13:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1990/3295 [20:42<13:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1991/3295 [20:44<13:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1991/3295 [20:44<13:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1992/3295 [20:44<13:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1992/3295 [20:44<13:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1993/3295 [20:44<13:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  60%|██████    | 1993/3295 [20:44<13:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1994/3295 [20:46<13:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1994/3295 [20:46<13:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1995/3295 [20:46<13:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1995/3295 [20:46<13:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1996/3295 [20:46<13:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1996/3295 [20:46<13:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1997/3295 [20:48<13:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1997/3295 [20:48<13:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1998/3295 [20:48<13:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1998/3295 [20:48<13:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1999/3295 [20:48<13:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 1999/3295 [20:48<13:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2000/3295 [20:50<13:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2000/3295 [20:50<13:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2001/3295 [20:50<13:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2001/3295 [20:50<13:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2002/3295 [20:50<13:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2002/3295 [20:50<13:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2003/3295 [20:52<13:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2003/3295 [20:52<13:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2004/3295 [20:52<13:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2004/3295 [20:52<13:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2005/3295 [20:52<13:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2005/3295 [20:52<13:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2006/3295 [20:54<13:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2006/3295 [20:54<13:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2007/3295 [20:54<13:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2007/3295 [20:54<13:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2008/3295 [20:54<13:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2008/3295 [20:54<13:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2009/3295 [20:56<13:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2009/3295 [20:56<13:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2010/3295 [20:56<13:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2010/3295 [20:56<13:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2011/3295 [20:56<13:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2011/3295 [20:56<13:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2012/3295 [20:58<13:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2012/3295 [20:58<13:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2013/3295 [20:58<13:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2013/3295 [20:58<13:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2014/3295 [20:58<13:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2014/3295 [20:58<13:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2015/3295 [21:00<13:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2015/3295 [21:00<13:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2016/3295 [21:00<13:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2016/3295 [21:00<13:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2017/3295 [21:00<13:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2017/3295 [21:00<13:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2018/3295 [21:02<13:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████    | 2018/3295 [21:02<13:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2019/3295 [21:02<13:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2019/3295 [21:02<13:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2020/3295 [21:02<13:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2020/3295 [21:02<13:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2021/3295 [21:04<13:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2021/3295 [21:04<13:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2022/3295 [21:04<13:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2022/3295 [21:04<13:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2023/3295 [21:04<13:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2023/3295 [21:04<13:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2024/3295 [21:05<13:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2024/3295 [21:05<13:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2025/3295 [21:06<13:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2025/3295 [21:06<13:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2026/3295 [21:06<13:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  61%|██████▏   | 2026/3295 [21:06<13:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2027/3295 [21:07<13:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2027/3295 [21:07<13:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2028/3295 [21:07<13:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2028/3295 [21:07<13:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2029/3295 [21:07<13:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2029/3295 [21:07<13:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2030/3295 [21:09<13:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2030/3295 [21:09<13:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2031/3295 [21:09<13:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2031/3295 [21:09<13:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2032/3295 [21:09<13:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2032/3295 [21:09<13:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2033/3295 [21:11<13:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2033/3295 [21:11<13:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2034/3295 [21:11<13:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2034/3295 [21:11<13:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2035/3295 [21:11<13:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2035/3295 [21:11<13:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2036/3295 [21:13<13:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2036/3295 [21:13<13:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2037/3295 [21:13<13:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2037/3295 [21:13<13:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2038/3295 [21:13<13:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2038/3295 [21:13<13:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2039/3295 [21:15<13:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2039/3295 [21:15<13:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2040/3295 [21:15<13:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2040/3295 [21:15<13:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2041/3295 [21:15<13:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2041/3295 [21:15<13:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2042/3295 [21:17<13:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2042/3295 [21:17<13:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2043/3295 [21:17<13:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2043/3295 [21:17<13:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2044/3295 [21:17<13:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2044/3295 [21:17<13:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2045/3295 [21:19<13:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2045/3295 [21:19<13:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2046/3295 [21:19<13:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2046/3295 [21:19<13:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2047/3295 [21:19<13:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2047/3295 [21:19<13:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2048/3295 [21:21<13:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2048/3295 [21:21<13:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2049/3295 [21:21<12:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2049/3295 [21:21<12:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2050/3295 [21:21<12:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2050/3295 [21:21<12:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2051/3295 [21:23<12:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2051/3295 [21:23<12:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2052/3295 [21:23<12:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2052/3295 [21:23<12:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2053/3295 [21:23<12:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2053/3295 [21:23<12:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2054/3295 [21:25<12:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2054/3295 [21:25<12:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2055/3295 [21:25<12:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2055/3295 [21:25<12:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2056/3295 [21:25<12:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2056/3295 [21:25<12:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2057/3295 [21:27<12:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2057/3295 [21:27<12:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2058/3295 [21:27<12:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2058/3295 [21:27<12:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2059/3295 [21:27<12:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  62%|██████▏   | 2059/3295 [21:27<12:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2060/3295 [21:28<12:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2060/3295 [21:28<12:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2061/3295 [21:28<12:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2061/3295 [21:28<12:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2062/3295 [21:28<12:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2062/3295 [21:28<12:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2063/3295 [21:30<12:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2063/3295 [21:30<12:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2064/3295 [21:30<12:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2064/3295 [21:30<12:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2065/3295 [21:30<12:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2065/3295 [21:30<12:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2066/3295 [21:32<12:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2066/3295 [21:32<12:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2067/3295 [21:32<12:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2067/3295 [21:32<12:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2068/3295 [21:32<12:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2068/3295 [21:32<12:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2069/3295 [21:34<12:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2069/3295 [21:34<12:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2070/3295 [21:34<12:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2070/3295 [21:34<12:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2071/3295 [21:34<12:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2071/3295 [21:34<12:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2072/3295 [21:36<12:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2072/3295 [21:36<12:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2073/3295 [21:36<12:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2073/3295 [21:36<12:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2074/3295 [21:36<12:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2074/3295 [21:36<12:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2075/3295 [21:38<12:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2075/3295 [21:38<12:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2076/3295 [21:38<12:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2076/3295 [21:38<12:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2077/3295 [21:38<12:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2077/3295 [21:38<12:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2078/3295 [21:40<12:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2078/3295 [21:40<12:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2079/3295 [21:40<12:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2079/3295 [21:40<12:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2080/3295 [21:40<12:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2080/3295 [21:40<12:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2081/3295 [21:42<12:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2081/3295 [21:42<12:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2082/3295 [21:42<12:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2082/3295 [21:42<12:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2083/3295 [21:42<12:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2083/3295 [21:42<12:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2084/3295 [21:44<12:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2084/3295 [21:44<12:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2085/3295 [21:44<12:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2085/3295 [21:44<12:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2086/3295 [21:44<12:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2086/3295 [21:44<12:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2087/3295 [21:46<12:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2087/3295 [21:46<12:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2088/3295 [21:46<12:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2088/3295 [21:46<12:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2089/3295 [21:46<12:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2089/3295 [21:46<12:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2090/3295 [21:48<12:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2090/3295 [21:48<12:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2091/3295 [21:48<12:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2091/3295 [21:48<12:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2092/3295 [21:48<12:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  63%|██████▎   | 2092/3295 [21:48<12:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2093/3295 [21:50<12:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2093/3295 [21:50<12:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2094/3295 [21:50<12:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2094/3295 [21:50<12:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2095/3295 [21:50<12:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2095/3295 [21:50<12:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2096/3295 [21:51<12:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2096/3295 [21:51<12:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2097/3295 [21:51<12:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2097/3295 [21:51<12:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2098/3295 [21:51<12:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2098/3295 [21:51<12:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2099/3295 [21:53<12:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2099/3295 [21:53<12:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2100/3295 [21:53<12:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▎   | 2100/3295 [21:53<12:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2101/3295 [21:53<12:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2101/3295 [21:53<12:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2102/3295 [21:55<12:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2102/3295 [21:55<12:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2103/3295 [21:55<12:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2103/3295 [21:55<12:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2104/3295 [21:55<12:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2104/3295 [21:55<12:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2105/3295 [21:57<12:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2105/3295 [21:57<12:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2106/3295 [21:57<12:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2106/3295 [21:57<12:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2107/3295 [21:57<12:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2107/3295 [21:57<12:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2108/3295 [21:59<12:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2108/3295 [21:59<12:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2109/3295 [21:59<12:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2109/3295 [21:59<12:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2110/3295 [21:59<12:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2110/3295 [21:59<12:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2111/3295 [22:01<12:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2111/3295 [22:01<12:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2112/3295 [22:01<12:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2112/3295 [22:01<12:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2113/3295 [22:01<12:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2113/3295 [22:01<12:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2114/3295 [22:03<12:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2114/3295 [22:03<12:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2115/3295 [22:03<12:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2115/3295 [22:03<12:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2116/3295 [22:03<12:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2116/3295 [22:03<12:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2117/3295 [22:05<12:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2117/3295 [22:05<12:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2118/3295 [22:05<12:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2118/3295 [22:05<12:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2119/3295 [22:05<12:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2119/3295 [22:05<12:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2120/3295 [22:06<12:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2120/3295 [22:06<12:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2121/3295 [22:07<12:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2121/3295 [22:07<12:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2122/3295 [22:07<12:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2122/3295 [22:07<12:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2123/3295 [22:08<12:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2123/3295 [22:08<12:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2124/3295 [22:08<12:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2124/3295 [22:08<12:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2125/3295 [22:08<12:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  64%|██████▍   | 2125/3295 [22:08<12:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2126/3295 [22:10<12:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2126/3295 [22:10<12:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2127/3295 [22:10<12:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2127/3295 [22:10<12:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2128/3295 [22:10<12:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2128/3295 [22:10<12:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2129/3295 [22:12<12:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2129/3295 [22:12<12:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2130/3295 [22:12<12:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2130/3295 [22:12<12:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2131/3295 [22:12<12:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2131/3295 [22:12<12:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2132/3295 [22:14<12:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2132/3295 [22:14<12:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2133/3295 [22:14<12:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2133/3295 [22:14<12:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2134/3295 [22:14<12:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2134/3295 [22:14<12:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2135/3295 [22:16<12:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2135/3295 [22:16<12:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2136/3295 [22:16<12:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2136/3295 [22:16<12:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2137/3295 [22:16<12:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2137/3295 [22:16<12:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2138/3295 [22:18<12:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2138/3295 [22:18<12:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2139/3295 [22:18<12:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2139/3295 [22:18<12:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2140/3295 [22:18<12:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2140/3295 [22:18<12:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2141/3295 [22:20<12:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▍   | 2141/3295 [22:20<12:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2142/3295 [22:20<12:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2142/3295 [22:20<12:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2143/3295 [22:20<12:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2143/3295 [22:20<12:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2144/3295 [22:21<12:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2144/3295 [22:21<12:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2145/3295 [22:21<11:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2145/3295 [22:21<11:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2146/3295 [22:22<11:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2146/3295 [22:22<11:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2147/3295 [22:23<11:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2147/3295 [22:23<11:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2148/3295 [22:23<11:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2148/3295 [22:23<11:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2149/3295 [22:23<11:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2149/3295 [22:23<11:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2150/3295 [22:25<11:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2150/3295 [22:25<11:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2151/3295 [22:25<11:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2151/3295 [22:25<11:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2152/3295 [22:25<11:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2152/3295 [22:25<11:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2153/3295 [22:27<11:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2153/3295 [22:27<11:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2154/3295 [22:27<11:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2154/3295 [22:27<11:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2155/3295 [22:27<11:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2155/3295 [22:27<11:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2156/3295 [22:29<11:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2156/3295 [22:29<11:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2157/3295 [22:29<11:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2157/3295 [22:29<11:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2158/3295 [22:29<11:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  65%|██████▌   | 2158/3295 [22:29<11:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2159/3295 [22:31<11:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2159/3295 [22:31<11:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2160/3295 [22:31<11:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2160/3295 [22:31<11:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2161/3295 [22:31<11:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2161/3295 [22:31<11:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2162/3295 [22:33<11:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2162/3295 [22:33<11:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2163/3295 [22:33<11:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2163/3295 [22:33<11:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2164/3295 [22:33<11:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2164/3295 [22:33<11:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2165/3295 [22:35<11:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2165/3295 [22:35<11:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2166/3295 [22:35<11:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2166/3295 [22:35<11:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2167/3295 [22:35<11:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2167/3295 [22:35<11:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2168/3295 [22:37<11:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2168/3295 [22:37<11:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2169/3295 [22:37<11:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2169/3295 [22:37<11:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2170/3295 [22:37<11:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2170/3295 [22:37<11:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2171/3295 [22:39<11:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2171/3295 [22:39<11:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2172/3295 [22:39<11:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2172/3295 [22:39<11:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2173/3295 [22:39<11:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2173/3295 [22:39<11:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2174/3295 [22:41<11:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2174/3295 [22:41<11:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2175/3295 [22:41<11:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2175/3295 [22:41<11:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2176/3295 [22:41<11:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2176/3295 [22:41<11:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2177/3295 [22:43<11:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2177/3295 [22:43<11:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2178/3295 [22:43<11:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2178/3295 [22:43<11:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2179/3295 [22:43<11:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2179/3295 [22:43<11:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2180/3295 [22:45<11:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2180/3295 [22:45<11:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2181/3295 [22:45<11:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2181/3295 [22:45<11:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2182/3295 [22:45<11:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▌   | 2182/3295 [22:45<11:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2183/3295 [22:47<11:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2183/3295 [22:47<11:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2184/3295 [22:47<11:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2184/3295 [22:47<11:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2185/3295 [22:47<11:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2185/3295 [22:47<11:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2186/3295 [22:49<11:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2186/3295 [22:49<11:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2187/3295 [22:49<11:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2187/3295 [22:49<11:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2188/3295 [22:49<11:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2188/3295 [22:49<11:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2189/3295 [22:50<11:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2189/3295 [22:50<11:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2190/3295 [22:50<11:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2190/3295 [22:50<11:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2191/3295 [22:50<11:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  66%|██████▋   | 2191/3295 [22:50<11:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2192/3295 [22:52<11:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2192/3295 [22:52<11:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2193/3295 [22:52<11:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2193/3295 [22:52<11:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2194/3295 [22:52<11:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2194/3295 [22:52<11:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2195/3295 [22:54<11:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2195/3295 [22:54<11:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2196/3295 [22:54<11:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2196/3295 [22:54<11:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2197/3295 [22:54<11:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2197/3295 [22:54<11:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2198/3295 [22:56<11:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2198/3295 [22:56<11:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2199/3295 [22:56<11:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2199/3295 [22:56<11:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2200/3295 [22:56<11:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2200/3295 [22:56<11:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2201/3295 [22:58<11:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2201/3295 [22:58<11:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2202/3295 [22:58<11:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2202/3295 [22:58<11:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2203/3295 [22:58<11:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2203/3295 [22:58<11:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2204/3295 [23:00<11:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2204/3295 [23:00<11:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2205/3295 [23:00<11:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2205/3295 [23:00<11:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2206/3295 [23:00<11:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2206/3295 [23:00<11:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2207/3295 [23:02<11:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2207/3295 [23:02<11:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2208/3295 [23:02<11:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2208/3295 [23:02<11:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2209/3295 [23:02<11:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2209/3295 [23:02<11:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2210/3295 [23:04<11:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2210/3295 [23:04<11:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2211/3295 [23:04<11:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2211/3295 [23:04<11:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2212/3295 [23:04<11:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2212/3295 [23:04<11:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2213/3295 [23:06<11:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2213/3295 [23:06<11:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2214/3295 [23:06<11:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2214/3295 [23:06<11:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2215/3295 [23:06<11:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2215/3295 [23:06<11:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2216/3295 [23:08<11:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2216/3295 [23:08<11:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2217/3295 [23:08<11:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2217/3295 [23:08<11:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2218/3295 [23:08<11:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2218/3295 [23:08<11:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2219/3295 [23:10<11:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2219/3295 [23:10<11:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2220/3295 [23:10<11:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2220/3295 [23:10<11:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2221/3295 [23:10<11:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2221/3295 [23:10<11:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2222/3295 [23:11<11:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2222/3295 [23:11<11:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2223/3295 [23:12<11:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2223/3295 [23:12<11:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2224/3295 [23:12<11:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  67%|██████▋   | 2224/3295 [23:12<11:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2225/3295 [23:13<11:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2225/3295 [23:13<11:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2226/3295 [23:13<11:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2226/3295 [23:13<11:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2227/3295 [23:13<11:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2227/3295 [23:13<11:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2228/3295 [23:15<11:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2228/3295 [23:15<11:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2229/3295 [23:15<11:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2229/3295 [23:15<11:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2230/3295 [23:15<11:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2230/3295 [23:15<11:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2231/3295 [23:17<11:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2231/3295 [23:17<11:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2232/3295 [23:17<11:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2232/3295 [23:17<11:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2233/3295 [23:17<11:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2233/3295 [23:17<11:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2234/3295 [23:19<11:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2234/3295 [23:19<11:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2235/3295 [23:19<11:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2235/3295 [23:19<11:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2236/3295 [23:19<11:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2236/3295 [23:19<11:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2237/3295 [23:21<11:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2237/3295 [23:21<11:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2238/3295 [23:21<11:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2238/3295 [23:21<11:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2239/3295 [23:21<11:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2239/3295 [23:21<11:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2240/3295 [23:23<11:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2240/3295 [23:23<11:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2241/3295 [23:23<10:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2241/3295 [23:23<10:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2242/3295 [23:23<10:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2242/3295 [23:23<10:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2243/3295 [23:25<10:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2243/3295 [23:25<10:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2244/3295 [23:25<10:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2244/3295 [23:25<10:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2245/3295 [23:25<10:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2245/3295 [23:25<10:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2246/3295 [23:27<10:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2246/3295 [23:27<10:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2247/3295 [23:27<10:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2247/3295 [23:27<10:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2248/3295 [23:27<10:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2248/3295 [23:27<10:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2249/3295 [23:28<10:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2249/3295 [23:28<10:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2250/3295 [23:28<10:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2250/3295 [23:28<10:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2251/3295 [23:29<10:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2251/3295 [23:29<10:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2252/3295 [23:30<10:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2252/3295 [23:30<10:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2253/3295 [23:30<10:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2253/3295 [23:30<10:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2254/3295 [23:30<10:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2254/3295 [23:30<10:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2255/3295 [23:32<10:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2255/3295 [23:32<10:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2256/3295 [23:32<10:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2256/3295 [23:32<10:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2257/3295 [23:32<10:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  68%|██████▊   | 2257/3295 [23:32<10:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2258/3295 [23:34<10:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2258/3295 [23:34<10:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2259/3295 [23:34<10:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2259/3295 [23:34<10:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2260/3295 [23:34<10:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2260/3295 [23:34<10:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2261/3295 [23:36<10:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2261/3295 [23:36<10:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2262/3295 [23:36<10:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2262/3295 [23:36<10:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2263/3295 [23:36<10:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2263/3295 [23:36<10:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2264/3295 [23:38<10:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2264/3295 [23:38<10:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2265/3295 [23:38<10:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▊   | 2265/3295 [23:38<10:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2266/3295 [23:38<10:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2266/3295 [23:38<10:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2267/3295 [23:40<10:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2267/3295 [23:40<10:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2268/3295 [23:40<10:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2268/3295 [23:40<10:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2269/3295 [23:40<10:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2269/3295 [23:40<10:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2270/3295 [23:42<10:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2270/3295 [23:42<10:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2271/3295 [23:42<10:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2271/3295 [23:42<10:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2272/3295 [23:42<10:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2272/3295 [23:42<10:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2273/3295 [23:43<10:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2273/3295 [23:43<10:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2274/3295 [23:44<10:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2274/3295 [23:44<10:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2275/3295 [23:44<10:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2275/3295 [23:44<10:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2276/3295 [23:45<10:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2276/3295 [23:45<10:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2277/3295 [23:45<10:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2277/3295 [23:45<10:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2278/3295 [23:45<10:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2278/3295 [23:45<10:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2279/3295 [23:47<10:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2279/3295 [23:47<10:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2280/3295 [23:47<10:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2280/3295 [23:47<10:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2281/3295 [23:47<10:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2281/3295 [23:47<10:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2282/3295 [23:49<10:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2282/3295 [23:49<10:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2283/3295 [23:49<10:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2283/3295 [23:49<10:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2284/3295 [23:49<10:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2284/3295 [23:49<10:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2285/3295 [23:51<10:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2285/3295 [23:51<10:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2286/3295 [23:51<10:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2286/3295 [23:51<10:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2287/3295 [23:51<10:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2287/3295 [23:51<10:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2288/3295 [23:53<10:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2288/3295 [23:53<10:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2289/3295 [23:53<10:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2289/3295 [23:53<10:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2290/3295 [23:53<10:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  69%|██████▉   | 2290/3295 [23:53<10:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2291/3295 [23:55<10:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2291/3295 [23:55<10:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2292/3295 [23:55<10:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2292/3295 [23:55<10:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2293/3295 [23:55<10:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2293/3295 [23:55<10:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2294/3295 [23:57<10:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2294/3295 [23:57<10:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2295/3295 [23:57<10:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2295/3295 [23:57<10:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2296/3295 [23:57<10:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2296/3295 [23:57<10:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2297/3295 [23:59<10:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2297/3295 [23:59<10:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2298/3295 [23:59<10:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2298/3295 [23:59<10:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2299/3295 [23:59<10:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2299/3295 [23:59<10:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2300/3295 [24:00<10:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2300/3295 [24:00<10:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2301/3295 [24:00<10:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2301/3295 [24:00<10:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2302/3295 [24:00<10:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2302/3295 [24:00<10:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2303/3295 [24:02<10:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2303/3295 [24:02<10:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2304/3295 [24:02<10:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2304/3295 [24:02<10:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2305/3295 [24:02<10:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2305/3295 [24:02<10:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2306/3295 [24:04<10:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|██████▉   | 2306/3295 [24:04<10:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2307/3295 [24:04<10:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2307/3295 [24:04<10:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2308/3295 [24:04<10:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2308/3295 [24:04<10:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2309/3295 [24:06<10:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2309/3295 [24:06<10:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2310/3295 [24:06<10:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2310/3295 [24:06<10:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2311/3295 [24:06<10:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2311/3295 [24:06<10:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2312/3295 [24:08<10:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2312/3295 [24:08<10:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2313/3295 [24:08<10:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2313/3295 [24:08<10:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2314/3295 [24:08<10:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2314/3295 [24:08<10:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2315/3295 [24:10<10:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2315/3295 [24:10<10:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2316/3295 [24:10<10:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2316/3295 [24:10<10:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2317/3295 [24:10<10:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2317/3295 [24:10<10:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2318/3295 [24:12<10:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2318/3295 [24:12<10:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2319/3295 [24:12<10:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2319/3295 [24:12<10:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2320/3295 [24:12<10:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2320/3295 [24:12<10:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2321/3295 [24:14<10:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2321/3295 [24:14<10:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2322/3295 [24:14<10:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  70%|███████   | 2322/3295 [24:14<10:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2323/3295 [24:14<10:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2323/3295 [24:14<10:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2324/3295 [24:16<10:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2324/3295 [24:16<10:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2325/3295 [24:16<10:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2325/3295 [24:16<10:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2326/3295 [24:16<10:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2326/3295 [24:16<10:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2327/3295 [24:17<10:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2327/3295 [24:17<10:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2328/3295 [24:18<10:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2328/3295 [24:18<10:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2329/3295 [24:18<10:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2329/3295 [24:18<10:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2330/3295 [24:19<10:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2330/3295 [24:19<10:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2331/3295 [24:19<10:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2331/3295 [24:19<10:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2332/3295 [24:19<10:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2332/3295 [24:19<10:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2333/3295 [24:21<10:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2333/3295 [24:21<10:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2334/3295 [24:21<10:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2334/3295 [24:21<10:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2335/3295 [24:21<10:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2335/3295 [24:21<10:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2336/3295 [24:23<10:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2336/3295 [24:23<10:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2337/3295 [24:23<09:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2337/3295 [24:23<09:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2338/3295 [24:23<09:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2338/3295 [24:23<09:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2339/3295 [24:25<09:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2339/3295 [24:25<09:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2340/3295 [24:25<09:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2340/3295 [24:25<09:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2341/3295 [24:25<09:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2341/3295 [24:25<09:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2342/3295 [24:27<09:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2342/3295 [24:27<09:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2343/3295 [24:27<09:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2343/3295 [24:27<09:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2344/3295 [24:27<09:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2344/3295 [24:27<09:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2345/3295 [24:29<09:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2345/3295 [24:29<09:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2346/3295 [24:29<09:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2346/3295 [24:29<09:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2347/3295 [24:29<09:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████   | 2347/3295 [24:29<09:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2348/3295 [24:30<09:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2348/3295 [24:30<09:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2349/3295 [24:31<09:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2349/3295 [24:31<09:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2350/3295 [24:31<09:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2350/3295 [24:31<09:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2351/3295 [24:32<09:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2351/3295 [24:32<09:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2352/3295 [24:32<09:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2352/3295 [24:32<09:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2353/3295 [24:32<09:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2353/3295 [24:32<09:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2354/3295 [24:34<09:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2354/3295 [24:34<09:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2355/3295 [24:34<09:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  71%|███████▏  | 2355/3295 [24:34<09:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2356/3295 [24:34<09:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2356/3295 [24:34<09:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2357/3295 [24:36<09:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2357/3295 [24:36<09:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2358/3295 [24:36<09:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2358/3295 [24:36<09:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2359/3295 [24:36<09:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2359/3295 [24:36<09:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2360/3295 [24:38<09:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2360/3295 [24:38<09:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2361/3295 [24:38<09:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2361/3295 [24:38<09:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2362/3295 [24:38<09:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2362/3295 [24:38<09:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2363/3295 [24:40<09:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2363/3295 [24:40<09:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2364/3295 [24:40<09:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2364/3295 [24:40<09:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2365/3295 [24:40<09:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2365/3295 [24:40<09:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2366/3295 [24:41<09:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2366/3295 [24:41<09:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2367/3295 [24:41<09:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2367/3295 [24:41<09:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2368/3295 [24:41<09:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2368/3295 [24:41<09:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2369/3295 [24:43<09:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2369/3295 [24:43<09:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2370/3295 [24:43<09:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2370/3295 [24:43<09:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2371/3295 [24:43<09:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2371/3295 [24:43<09:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2372/3295 [24:45<09:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2372/3295 [24:45<09:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2373/3295 [24:45<09:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2373/3295 [24:45<09:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2374/3295 [24:45<09:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2374/3295 [24:45<09:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2375/3295 [24:47<09:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2375/3295 [24:47<09:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2376/3295 [24:47<09:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2376/3295 [24:47<09:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2377/3295 [24:47<09:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2377/3295 [24:47<09:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2378/3295 [24:49<09:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2378/3295 [24:49<09:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2379/3295 [24:49<09:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2379/3295 [24:49<09:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2380/3295 [24:49<09:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2380/3295 [24:49<09:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2381/3295 [24:51<09:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2381/3295 [24:51<09:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2382/3295 [24:51<09:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2382/3295 [24:51<09:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2383/3295 [24:51<09:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2383/3295 [24:51<09:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2384/3295 [24:52<09:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2384/3295 [24:52<09:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2385/3295 [24:52<09:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2385/3295 [24:52<09:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2386/3295 [24:53<09:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2386/3295 [24:53<09:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2387/3295 [24:54<09:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2387/3295 [24:54<09:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2388/3295 [24:54<09:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  72%|███████▏  | 2388/3295 [24:54<09:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2389/3295 [24:54<09:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2389/3295 [24:54<09:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2390/3295 [24:56<09:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2390/3295 [24:56<09:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2391/3295 [24:56<09:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2391/3295 [24:56<09:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2392/3295 [24:56<09:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2392/3295 [24:56<09:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2393/3295 [24:58<09:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2393/3295 [24:58<09:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2394/3295 [24:58<09:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2394/3295 [24:58<09:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2395/3295 [24:58<09:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2395/3295 [24:58<09:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2396/3295 [25:00<09:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2396/3295 [25:00<09:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2397/3295 [25:00<09:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2397/3295 [25:00<09:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2398/3295 [25:00<09:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2398/3295 [25:00<09:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2399/3295 [25:02<09:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2399/3295 [25:02<09:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2400/3295 [25:02<09:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2400/3295 [25:02<09:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2401/3295 [25:02<09:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2401/3295 [25:02<09:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2402/3295 [25:03<09:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2402/3295 [25:03<09:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2403/3295 [25:03<09:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2403/3295 [25:03<09:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2404/3295 [25:03<09:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2404/3295 [25:03<09:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2405/3295 [25:05<09:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2405/3295 [25:05<09:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2406/3295 [25:05<09:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2406/3295 [25:05<09:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2407/3295 [25:05<09:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2407/3295 [25:05<09:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2408/3295 [25:07<09:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2408/3295 [25:07<09:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2409/3295 [25:07<09:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2409/3295 [25:07<09:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2410/3295 [25:07<09:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2410/3295 [25:07<09:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2411/3295 [25:09<09:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2411/3295 [25:09<09:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2412/3295 [25:09<09:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2412/3295 [25:09<09:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2413/3295 [25:09<09:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2413/3295 [25:09<09:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2414/3295 [25:11<09:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2414/3295 [25:11<09:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2415/3295 [25:11<09:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2415/3295 [25:11<09:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2416/3295 [25:11<09:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2416/3295 [25:11<09:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2417/3295 [25:13<09:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2417/3295 [25:13<09:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2418/3295 [25:13<09:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2418/3295 [25:13<09:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2419/3295 [25:13<09:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2419/3295 [25:13<09:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2420/3295 [25:15<09:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2420/3295 [25:15<09:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2421/3295 [25:15<09:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  73%|███████▎  | 2421/3295 [25:15<09:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2422/3295 [25:15<09:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2422/3295 [25:15<09:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2423/3295 [25:16<09:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2423/3295 [25:16<09:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2424/3295 [25:16<09:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2424/3295 [25:16<09:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2425/3295 [25:16<09:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2425/3295 [25:16<09:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2426/3295 [25:18<09:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2426/3295 [25:18<09:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2427/3295 [25:18<09:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2427/3295 [25:18<09:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2428/3295 [25:18<09:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2428/3295 [25:18<09:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2429/3295 [25:20<09:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2429/3295 [25:20<09:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2430/3295 [25:20<09:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▎  | 2430/3295 [25:20<09:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2431/3295 [25:20<09:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2431/3295 [25:20<09:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2432/3295 [25:22<09:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2432/3295 [25:22<09:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2433/3295 [25:22<08:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2433/3295 [25:22<08:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2434/3295 [25:22<08:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2434/3295 [25:22<08:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2435/3295 [25:24<08:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2435/3295 [25:24<08:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2436/3295 [25:24<08:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2436/3295 [25:24<08:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2437/3295 [25:24<08:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2437/3295 [25:24<08:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2438/3295 [25:26<08:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2438/3295 [25:26<08:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2439/3295 [25:26<08:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2439/3295 [25:26<08:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2440/3295 [25:26<08:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2440/3295 [25:26<08:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2441/3295 [25:28<08:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2441/3295 [25:28<08:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2442/3295 [25:28<08:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2442/3295 [25:28<08:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2443/3295 [25:28<08:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2443/3295 [25:28<08:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2444/3295 [25:29<08:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2444/3295 [25:29<08:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2445/3295 [25:29<08:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2445/3295 [25:29<08:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2446/3295 [25:29<08:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2446/3295 [25:29<08:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2447/3295 [25:31<08:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2447/3295 [25:31<08:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2448/3295 [25:31<08:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2448/3295 [25:31<08:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2449/3295 [25:31<08:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2449/3295 [25:31<08:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2450/3295 [25:33<08:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2450/3295 [25:33<08:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2451/3295 [25:33<08:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2451/3295 [25:33<08:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2452/3295 [25:33<08:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2452/3295 [25:33<08:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2453/3295 [25:35<08:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2453/3295 [25:35<08:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2454/3295 [25:35<08:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  74%|███████▍  | 2454/3295 [25:35<08:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2455/3295 [25:35<08:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2455/3295 [25:35<08:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2456/3295 [25:37<08:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2456/3295 [25:37<08:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2457/3295 [25:37<08:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2457/3295 [25:37<08:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2458/3295 [25:37<08:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2458/3295 [25:37<08:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2459/3295 [25:39<08:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2459/3295 [25:39<08:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2460/3295 [25:39<08:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2460/3295 [25:39<08:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2461/3295 [25:39<08:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2461/3295 [25:39<08:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2462/3295 [25:41<08:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2462/3295 [25:41<08:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2463/3295 [25:41<08:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2463/3295 [25:41<08:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2464/3295 [25:41<08:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2464/3295 [25:41<08:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2465/3295 [25:43<08:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2465/3295 [25:43<08:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2466/3295 [25:43<08:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2466/3295 [25:43<08:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2467/3295 [25:43<08:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2467/3295 [25:43<08:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2468/3295 [25:45<08:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2468/3295 [25:45<08:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2469/3295 [25:45<08:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2469/3295 [25:45<08:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2470/3295 [25:45<08:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2470/3295 [25:45<08:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2471/3295 [25:46<08:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▍  | 2471/3295 [25:46<08:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2472/3295 [25:46<08:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2472/3295 [25:46<08:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2473/3295 [25:46<08:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2473/3295 [25:46<08:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2474/3295 [25:48<08:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2474/3295 [25:48<08:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2475/3295 [25:48<08:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2475/3295 [25:48<08:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2476/3295 [25:48<08:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2476/3295 [25:48<08:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2477/3295 [25:50<08:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2477/3295 [25:50<08:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2478/3295 [25:50<08:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2478/3295 [25:50<08:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2479/3295 [25:50<08:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2479/3295 [25:50<08:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2480/3295 [25:52<08:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2480/3295 [25:52<08:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2481/3295 [25:52<08:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2481/3295 [25:52<08:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2482/3295 [25:52<08:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2482/3295 [25:52<08:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2483/3295 [25:54<08:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2483/3295 [25:54<08:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2484/3295 [25:54<08:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2484/3295 [25:54<08:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2485/3295 [25:54<08:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2485/3295 [25:54<08:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2486/3295 [25:55<08:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2486/3295 [25:55<08:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2487/3295 [25:55<08:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  75%|███████▌  | 2487/3295 [25:55<08:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2488/3295 [25:55<08:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2488/3295 [25:55<08:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2489/3295 [25:57<08:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2489/3295 [25:57<08:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2490/3295 [25:57<08:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2490/3295 [25:57<08:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2491/3295 [25:57<08:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2491/3295 [25:57<08:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2492/3295 [25:59<08:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2492/3295 [25:59<08:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2493/3295 [25:59<08:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2493/3295 [25:59<08:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2494/3295 [25:59<08:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2494/3295 [25:59<08:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2495/3295 [26:01<08:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2495/3295 [26:01<08:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2496/3295 [26:01<08:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2496/3295 [26:01<08:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2497/3295 [26:01<08:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2497/3295 [26:01<08:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2498/3295 [26:03<08:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2498/3295 [26:03<08:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2499/3295 [26:03<08:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2499/3295 [26:03<08:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2500/3295 [26:03<08:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2500/3295 [26:03<08:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2501/3295 [26:05<08:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2501/3295 [26:05<08:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2502/3295 [26:05<08:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2502/3295 [26:05<08:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2503/3295 [26:05<08:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2503/3295 [26:05<08:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2504/3295 [26:07<08:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2504/3295 [26:07<08:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2505/3295 [26:07<08:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2505/3295 [26:07<08:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2506/3295 [26:07<08:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2506/3295 [26:07<08:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2507/3295 [26:09<08:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2507/3295 [26:09<08:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2508/3295 [26:09<08:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2508/3295 [26:09<08:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2509/3295 [26:09<08:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2509/3295 [26:09<08:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2510/3295 [26:10<08:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2510/3295 [26:10<08:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2511/3295 [26:10<08:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2511/3295 [26:10<08:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2512/3295 [26:10<08:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▌  | 2512/3295 [26:10<08:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2513/3295 [26:12<08:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2513/3295 [26:12<08:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2514/3295 [26:12<08:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2514/3295 [26:12<08:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2515/3295 [26:12<08:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2515/3295 [26:12<08:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2516/3295 [26:14<08:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2516/3295 [26:14<08:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2517/3295 [26:14<08:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2517/3295 [26:14<08:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2518/3295 [26:14<08:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2518/3295 [26:14<08:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2519/3295 [26:16<08:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2519/3295 [26:16<08:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2520/3295 [26:16<08:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  76%|███████▋  | 2520/3295 [26:16<08:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2521/3295 [26:16<08:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2521/3295 [26:16<08:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2522/3295 [26:18<08:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2522/3295 [26:18<08:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2523/3295 [26:18<08:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2523/3295 [26:18<08:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2524/3295 [26:18<08:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2524/3295 [26:18<08:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2525/3295 [26:19<08:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2525/3295 [26:19<08:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2526/3295 [26:19<08:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2526/3295 [26:19<08:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2527/3295 [26:19<08:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2527/3295 [26:19<08:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2528/3295 [26:21<07:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2528/3295 [26:21<07:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2529/3295 [26:21<07:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2529/3295 [26:21<07:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2530/3295 [26:21<07:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2530/3295 [26:21<07:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2531/3295 [26:23<07:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2531/3295 [26:23<07:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2532/3295 [26:23<07:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2532/3295 [26:23<07:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2533/3295 [26:23<07:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2533/3295 [26:23<07:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2534/3295 [26:25<07:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2534/3295 [26:25<07:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2535/3295 [26:25<07:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2535/3295 [26:25<07:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2536/3295 [26:25<07:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2536/3295 [26:25<07:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2537/3295 [26:27<07:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2537/3295 [26:27<07:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2538/3295 [26:27<07:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2538/3295 [26:27<07:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2539/3295 [26:27<07:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2539/3295 [26:27<07:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2540/3295 [26:29<07:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2540/3295 [26:29<07:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2541/3295 [26:29<07:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2541/3295 [26:29<07:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2542/3295 [26:29<07:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2542/3295 [26:29<07:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2543/3295 [26:31<07:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2543/3295 [26:31<07:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2544/3295 [26:31<07:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2544/3295 [26:31<07:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2545/3295 [26:31<07:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2545/3295 [26:31<07:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2546/3295 [26:32<07:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2546/3295 [26:32<07:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2547/3295 [26:32<07:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2547/3295 [26:32<07:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2548/3295 [26:32<07:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2548/3295 [26:32<07:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2549/3295 [26:34<07:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2549/3295 [26:34<07:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2550/3295 [26:34<07:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2550/3295 [26:34<07:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2551/3295 [26:34<07:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2551/3295 [26:34<07:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2552/3295 [26:36<07:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2552/3295 [26:36<07:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2553/3295 [26:36<07:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  77%|███████▋  | 2553/3295 [26:36<07:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2554/3295 [26:36<07:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2554/3295 [26:36<07:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2555/3295 [26:38<07:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2555/3295 [26:38<07:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2556/3295 [26:38<07:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2556/3295 [26:38<07:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2557/3295 [26:38<07:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2557/3295 [26:38<07:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2558/3295 [26:40<07:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2558/3295 [26:40<07:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2559/3295 [26:40<07:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2559/3295 [26:40<07:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2560/3295 [26:40<07:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2560/3295 [26:40<07:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2561/3295 [26:42<07:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2561/3295 [26:42<07:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2562/3295 [26:42<07:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2562/3295 [26:42<07:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2563/3295 [26:42<07:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2563/3295 [26:42<07:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2564/3295 [26:44<07:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2564/3295 [26:44<07:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2565/3295 [26:44<07:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2565/3295 [26:44<07:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2566/3295 [26:44<07:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2566/3295 [26:44<07:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2567/3295 [26:45<07:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2567/3295 [26:45<07:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2568/3295 [26:46<07:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2568/3295 [26:46<07:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2569/3295 [26:46<07:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2569/3295 [26:46<07:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2570/3295 [26:47<07:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2570/3295 [26:47<07:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2571/3295 [26:47<07:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2571/3295 [26:47<07:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2572/3295 [26:47<07:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2572/3295 [26:47<07:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2573/3295 [26:49<07:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2573/3295 [26:49<07:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2574/3295 [26:49<07:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2574/3295 [26:49<07:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2575/3295 [26:49<07:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2575/3295 [26:49<07:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2576/3295 [26:51<07:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2576/3295 [26:51<07:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2577/3295 [26:51<07:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2577/3295 [26:51<07:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2578/3295 [26:51<07:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2578/3295 [26:51<07:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2579/3295 [26:53<07:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2579/3295 [26:53<07:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2580/3295 [26:53<07:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2580/3295 [26:53<07:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2581/3295 [26:53<07:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2581/3295 [26:53<07:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2582/3295 [26:55<07:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2582/3295 [26:55<07:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2583/3295 [26:55<07:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2583/3295 [26:55<07:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2584/3295 [26:55<07:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2584/3295 [26:55<07:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2585/3295 [26:57<07:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2585/3295 [26:57<07:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2586/3295 [26:57<07:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  78%|███████▊  | 2586/3295 [26:57<07:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2587/3295 [26:57<07:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2587/3295 [26:57<07:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2588/3295 [26:58<07:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2588/3295 [26:58<07:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2589/3295 [26:58<07:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2589/3295 [26:58<07:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2590/3295 [26:58<07:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2590/3295 [26:58<07:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2591/3295 [27:00<07:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2591/3295 [27:00<07:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2592/3295 [27:00<07:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2592/3295 [27:00<07:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2593/3295 [27:00<07:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2593/3295 [27:00<07:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2594/3295 [27:02<07:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▊  | 2594/3295 [27:02<07:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2595/3295 [27:02<07:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2595/3295 [27:02<07:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2596/3295 [27:02<07:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2596/3295 [27:02<07:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2597/3295 [27:04<07:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2597/3295 [27:04<07:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2598/3295 [27:04<07:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2598/3295 [27:04<07:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2599/3295 [27:04<07:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2599/3295 [27:04<07:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2600/3295 [27:05<07:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2600/3295 [27:05<07:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2601/3295 [27:06<07:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2601/3295 [27:06<07:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2602/3295 [27:06<07:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2602/3295 [27:06<07:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2603/3295 [27:07<07:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2603/3295 [27:07<07:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2604/3295 [27:08<07:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2604/3295 [27:08<07:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2605/3295 [27:08<07:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2605/3295 [27:08<07:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2606/3295 [27:09<07:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2606/3295 [27:09<07:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2607/3295 [27:09<07:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2607/3295 [27:09<07:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2608/3295 [27:09<07:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2608/3295 [27:09<07:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2609/3295 [27:11<07:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2609/3295 [27:11<07:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2610/3295 [27:11<07:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2610/3295 [27:11<07:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2611/3295 [27:11<07:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2611/3295 [27:11<07:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2612/3295 [27:13<07:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2612/3295 [27:13<07:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2613/3295 [27:13<07:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2613/3295 [27:13<07:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2614/3295 [27:13<07:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2614/3295 [27:13<07:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2615/3295 [27:15<07:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2615/3295 [27:15<07:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2616/3295 [27:15<07:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2616/3295 [27:15<07:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2617/3295 [27:15<07:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2617/3295 [27:15<07:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2618/3295 [27:16<07:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2618/3295 [27:16<07:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2619/3295 [27:17<07:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  79%|███████▉  | 2619/3295 [27:17<07:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2620/3295 [27:17<07:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2620/3295 [27:17<07:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2621/3295 [27:18<07:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2621/3295 [27:18<07:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2622/3295 [27:18<07:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2622/3295 [27:18<07:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2623/3295 [27:18<06:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2623/3295 [27:18<06:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2624/3295 [27:20<06:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2624/3295 [27:20<06:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2625/3295 [27:20<06:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2625/3295 [27:20<06:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2626/3295 [27:20<06:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2626/3295 [27:20<06:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2627/3295 [27:22<06:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2627/3295 [27:22<06:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2628/3295 [27:22<06:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2628/3295 [27:22<06:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2629/3295 [27:22<06:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2629/3295 [27:22<06:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2630/3295 [27:24<06:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2630/3295 [27:24<06:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2631/3295 [27:24<06:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2631/3295 [27:24<06:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2632/3295 [27:24<06:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2632/3295 [27:24<06:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2633/3295 [27:26<06:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2633/3295 [27:26<06:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2634/3295 [27:26<06:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2634/3295 [27:26<06:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2635/3295 [27:26<06:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|███████▉  | 2635/3295 [27:26<06:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2636/3295 [27:28<06:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2636/3295 [27:28<06:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2637/3295 [27:28<06:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2637/3295 [27:28<06:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2638/3295 [27:28<06:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2638/3295 [27:28<06:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2639/3295 [27:30<06:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2639/3295 [27:30<06:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2640/3295 [27:30<06:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2640/3295 [27:30<06:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2641/3295 [27:30<06:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2641/3295 [27:30<06:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2642/3295 [27:31<06:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2642/3295 [27:31<06:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2643/3295 [27:31<06:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2643/3295 [27:31<06:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2644/3295 [27:31<06:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2644/3295 [27:31<06:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2645/3295 [27:33<06:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2645/3295 [27:33<06:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2646/3295 [27:33<06:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2646/3295 [27:33<06:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2647/3295 [27:33<06:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2647/3295 [27:33<06:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2648/3295 [27:35<06:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2648/3295 [27:35<06:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2649/3295 [27:35<06:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2649/3295 [27:35<06:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2650/3295 [27:35<06:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2650/3295 [27:35<06:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2651/3295 [27:37<06:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2651/3295 [27:37<06:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2652/3295 [27:37<06:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  80%|████████  | 2652/3295 [27:37<06:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2653/3295 [27:37<06:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2653/3295 [27:37<06:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2654/3295 [27:39<06:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2654/3295 [27:39<06:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2655/3295 [27:39<06:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2655/3295 [27:39<06:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2656/3295 [27:39<06:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2656/3295 [27:39<06:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2657/3295 [27:41<06:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2657/3295 [27:41<06:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2658/3295 [27:41<06:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2658/3295 [27:41<06:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2659/3295 [27:41<06:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2659/3295 [27:41<06:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2660/3295 [27:43<06:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2660/3295 [27:43<06:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2661/3295 [27:43<06:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2661/3295 [27:43<06:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2662/3295 [27:43<06:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2662/3295 [27:43<06:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2663/3295 [27:45<06:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2663/3295 [27:45<06:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2664/3295 [27:45<06:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2664/3295 [27:45<06:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2665/3295 [27:45<06:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2665/3295 [27:45<06:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2666/3295 [27:46<06:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2666/3295 [27:46<06:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2667/3295 [27:46<06:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2667/3295 [27:46<06:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2668/3295 [27:46<06:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2668/3295 [27:46<06:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2669/3295 [27:48<06:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2669/3295 [27:48<06:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2670/3295 [27:48<06:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2670/3295 [27:48<06:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2671/3295 [27:48<06:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2671/3295 [27:48<06:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2672/3295 [27:50<06:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2672/3295 [27:50<06:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2673/3295 [27:50<06:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2673/3295 [27:50<06:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2674/3295 [27:50<06:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2674/3295 [27:50<06:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2675/3295 [27:52<06:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2675/3295 [27:52<06:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2676/3295 [27:52<06:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2676/3295 [27:52<06:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2677/3295 [27:52<06:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████  | 2677/3295 [27:52<06:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2678/3295 [27:54<06:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2678/3295 [27:54<06:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2679/3295 [27:54<06:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2679/3295 [27:54<06:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2680/3295 [27:54<06:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2680/3295 [27:54<06:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2681/3295 [27:56<06:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2681/3295 [27:56<06:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2682/3295 [27:56<06:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2682/3295 [27:56<06:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2683/3295 [27:56<06:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2683/3295 [27:56<06:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2684/3295 [27:57<06:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2684/3295 [27:57<06:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2685/3295 [27:58<06:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  81%|████████▏ | 2685/3295 [27:58<06:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2686/3295 [27:58<06:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2686/3295 [27:58<06:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2687/3295 [27:59<06:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2687/3295 [27:59<06:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2688/3295 [28:00<06:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2688/3295 [28:00<06:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2689/3295 [28:00<06:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2689/3295 [28:00<06:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2690/3295 [28:01<06:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2690/3295 [28:01<06:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2691/3295 [28:01<06:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2691/3295 [28:01<06:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2692/3295 [28:01<06:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2692/3295 [28:01<06:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2693/3295 [28:03<06:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2693/3295 [28:03<06:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2694/3295 [28:03<06:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2694/3295 [28:03<06:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2695/3295 [28:03<06:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2695/3295 [28:03<06:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2696/3295 [28:05<06:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2696/3295 [28:05<06:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2697/3295 [28:05<06:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2697/3295 [28:05<06:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2698/3295 [28:05<06:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2698/3295 [28:05<06:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2699/3295 [28:07<06:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2699/3295 [28:07<06:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2700/3295 [28:07<06:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2700/3295 [28:07<06:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2701/3295 [28:07<06:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2701/3295 [28:07<06:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2702/3295 [28:09<06:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2702/3295 [28:09<06:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2703/3295 [28:09<06:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2703/3295 [28:09<06:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2704/3295 [28:09<06:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2704/3295 [28:09<06:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2705/3295 [28:11<06:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2705/3295 [28:11<06:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2706/3295 [28:11<06:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2706/3295 [28:11<06:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2707/3295 [28:11<06:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2707/3295 [28:11<06:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2708/3295 [28:12<06:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2708/3295 [28:12<06:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2709/3295 [28:13<06:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2709/3295 [28:13<06:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2710/3295 [28:13<06:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2710/3295 [28:13<06:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2711/3295 [28:14<06:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2711/3295 [28:14<06:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2712/3295 [28:15<06:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2712/3295 [28:15<06:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2713/3295 [28:15<06:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2713/3295 [28:15<06:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2714/3295 [28:16<06:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2714/3295 [28:16<06:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2715/3295 [28:16<06:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2715/3295 [28:16<06:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2716/3295 [28:16<06:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2716/3295 [28:16<06:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2717/3295 [28:18<06:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2717/3295 [28:18<06:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2718/3295 [28:18<06:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  82%|████████▏ | 2718/3295 [28:18<06:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2719/3295 [28:18<05:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2719/3295 [28:18<05:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2720/3295 [28:20<05:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2720/3295 [28:20<05:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2721/3295 [28:20<05:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2721/3295 [28:20<05:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2722/3295 [28:20<05:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2722/3295 [28:20<05:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2723/3295 [28:21<05:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2723/3295 [28:21<05:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2724/3295 [28:22<05:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2724/3295 [28:22<05:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2725/3295 [28:22<05:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2725/3295 [28:22<05:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2726/3295 [28:23<05:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2726/3295 [28:23<05:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2727/3295 [28:24<05:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2727/3295 [28:24<05:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2728/3295 [28:24<05:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2728/3295 [28:24<05:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2729/3295 [28:25<05:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2729/3295 [28:25<05:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2730/3295 [28:25<05:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2730/3295 [28:25<05:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2731/3295 [28:25<05:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2731/3295 [28:25<05:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2732/3295 [28:27<05:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2732/3295 [28:27<05:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2733/3295 [28:27<05:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2733/3295 [28:27<05:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2734/3295 [28:27<05:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2734/3295 [28:27<05:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2735/3295 [28:29<05:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2735/3295 [28:29<05:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2736/3295 [28:29<05:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2736/3295 [28:29<05:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2737/3295 [28:29<05:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2737/3295 [28:29<05:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2738/3295 [28:31<05:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2738/3295 [28:31<05:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2739/3295 [28:31<05:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2739/3295 [28:31<05:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2740/3295 [28:31<05:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2740/3295 [28:31<05:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2741/3295 [28:33<05:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2741/3295 [28:33<05:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2742/3295 [28:33<05:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2742/3295 [28:33<05:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2743/3295 [28:33<05:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2743/3295 [28:33<05:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2744/3295 [28:35<05:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2744/3295 [28:35<05:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2745/3295 [28:35<05:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2745/3295 [28:35<05:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2746/3295 [28:35<05:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2746/3295 [28:35<05:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2747/3295 [28:37<05:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2747/3295 [28:37<05:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2748/3295 [28:37<05:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2748/3295 [28:37<05:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2749/3295 [28:37<05:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2749/3295 [28:37<05:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2750/3295 [28:39<05:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2750/3295 [28:39<05:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2751/3295 [28:39<05:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  83%|████████▎ | 2751/3295 [28:39<05:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2752/3295 [28:39<05:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2752/3295 [28:39<05:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2753/3295 [28:40<05:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2753/3295 [28:40<05:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2754/3295 [28:40<05:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2754/3295 [28:40<05:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2755/3295 [28:40<05:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2755/3295 [28:40<05:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2756/3295 [28:42<05:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2756/3295 [28:42<05:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2757/3295 [28:42<05:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2757/3295 [28:42<05:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2758/3295 [28:42<05:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2758/3295 [28:42<05:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2759/3295 [28:44<05:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▎ | 2759/3295 [28:44<05:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2760/3295 [28:44<05:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2760/3295 [28:44<05:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2761/3295 [28:44<05:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2761/3295 [28:44<05:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2762/3295 [28:46<05:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2762/3295 [28:46<05:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2763/3295 [28:46<05:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2763/3295 [28:46<05:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2764/3295 [28:46<05:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2764/3295 [28:46<05:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2765/3295 [28:48<05:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2765/3295 [28:48<05:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2766/3295 [28:48<05:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2766/3295 [28:48<05:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2767/3295 [28:48<05:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2767/3295 [28:48<05:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2768/3295 [28:50<05:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2768/3295 [28:50<05:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2769/3295 [28:50<05:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2769/3295 [28:50<05:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2770/3295 [28:50<05:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2770/3295 [28:50<05:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2771/3295 [28:51<05:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2771/3295 [28:51<05:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2772/3295 [28:51<05:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2772/3295 [28:51<05:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2773/3295 [28:51<05:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2773/3295 [28:51<05:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2774/3295 [28:53<05:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2774/3295 [28:53<05:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2775/3295 [28:53<05:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2775/3295 [28:53<05:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2776/3295 [28:53<05:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2776/3295 [28:53<05:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2777/3295 [28:55<05:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2777/3295 [28:55<05:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2778/3295 [28:55<05:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2778/3295 [28:55<05:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2779/3295 [28:55<05:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2779/3295 [28:55<05:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2780/3295 [28:57<05:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2780/3295 [28:57<05:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2781/3295 [28:57<05:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2781/3295 [28:57<05:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2782/3295 [28:57<05:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2782/3295 [28:57<05:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2783/3295 [28:59<05:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2783/3295 [28:59<05:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2784/3295 [28:59<05:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  84%|████████▍ | 2784/3295 [28:59<05:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2785/3295 [28:59<05:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2785/3295 [28:59<05:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2786/3295 [29:01<05:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2786/3295 [29:01<05:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2787/3295 [29:01<05:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2787/3295 [29:01<05:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2788/3295 [29:01<05:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2788/3295 [29:01<05:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2789/3295 [29:03<05:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2789/3295 [29:03<05:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2790/3295 [29:03<05:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2790/3295 [29:03<05:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2791/3295 [29:03<05:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2791/3295 [29:03<05:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2792/3295 [29:04<05:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2792/3295 [29:04<05:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2793/3295 [29:04<05:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2793/3295 [29:04<05:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2794/3295 [29:04<05:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2794/3295 [29:04<05:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2795/3295 [29:06<05:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2795/3295 [29:06<05:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2796/3295 [29:06<05:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2796/3295 [29:06<05:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2797/3295 [29:06<05:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2797/3295 [29:06<05:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2798/3295 [29:08<05:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2798/3295 [29:08<05:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2799/3295 [29:08<05:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2799/3295 [29:08<05:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2800/3295 [29:08<05:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▍ | 2800/3295 [29:08<05:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2801/3295 [29:10<05:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2801/3295 [29:10<05:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2802/3295 [29:10<05:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2802/3295 [29:10<05:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2803/3295 [29:10<05:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2803/3295 [29:10<05:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2804/3295 [29:12<05:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2804/3295 [29:12<05:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2805/3295 [29:12<05:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2805/3295 [29:12<05:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2806/3295 [29:12<05:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2806/3295 [29:12<05:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2807/3295 [29:14<05:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2807/3295 [29:14<05:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2808/3295 [29:14<05:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2808/3295 [29:14<05:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2809/3295 [29:14<05:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2809/3295 [29:14<05:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2810/3295 [29:15<05:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2810/3295 [29:15<05:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2811/3295 [29:15<05:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2811/3295 [29:15<05:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2812/3295 [29:15<05:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2812/3295 [29:15<05:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2813/3295 [29:17<05:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2813/3295 [29:17<05:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2814/3295 [29:17<05:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2814/3295 [29:17<05:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2815/3295 [29:17<04:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2815/3295 [29:17<04:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2816/3295 [29:19<04:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2816/3295 [29:19<04:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2817/3295 [29:19<04:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  85%|████████▌ | 2817/3295 [29:19<04:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2818/3295 [29:19<04:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2818/3295 [29:19<04:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2819/3295 [29:21<04:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2819/3295 [29:21<04:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2820/3295 [29:21<04:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2820/3295 [29:21<04:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2821/3295 [29:21<04:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2821/3295 [29:21<04:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2822/3295 [29:23<04:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2822/3295 [29:23<04:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2823/3295 [29:23<04:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2823/3295 [29:23<04:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2824/3295 [29:23<04:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2824/3295 [29:23<04:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2825/3295 [29:25<04:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2825/3295 [29:25<04:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2826/3295 [29:25<04:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2826/3295 [29:25<04:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2827/3295 [29:25<04:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2827/3295 [29:25<04:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2828/3295 [29:26<04:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2828/3295 [29:26<04:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2829/3295 [29:27<04:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2829/3295 [29:27<04:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2830/3295 [29:27<04:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2830/3295 [29:27<04:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2831/3295 [29:28<04:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2831/3295 [29:28<04:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2832/3295 [29:28<04:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2832/3295 [29:28<04:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2833/3295 [29:28<04:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2833/3295 [29:28<04:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2834/3295 [29:30<04:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2834/3295 [29:30<04:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2835/3295 [29:30<04:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2835/3295 [29:30<04:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2836/3295 [29:30<04:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2836/3295 [29:30<04:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2837/3295 [29:32<04:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2837/3295 [29:32<04:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2838/3295 [29:32<04:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2838/3295 [29:32<04:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2839/3295 [29:32<04:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2839/3295 [29:32<04:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2840/3295 [29:34<04:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2840/3295 [29:34<04:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2841/3295 [29:34<04:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▌ | 2841/3295 [29:34<04:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2842/3295 [29:34<04:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2842/3295 [29:34<04:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2843/3295 [29:36<04:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2843/3295 [29:36<04:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2844/3295 [29:36<04:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2844/3295 [29:36<04:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2845/3295 [29:36<04:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2845/3295 [29:36<04:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2846/3295 [29:38<04:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2846/3295 [29:38<04:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2847/3295 [29:38<04:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2847/3295 [29:38<04:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2848/3295 [29:38<04:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2848/3295 [29:38<04:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2849/3295 [29:40<04:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2849/3295 [29:40<04:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2850/3295 [29:40<04:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  86%|████████▋ | 2850/3295 [29:40<04:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2851/3295 [29:40<04:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2851/3295 [29:40<04:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2852/3295 [29:42<04:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2852/3295 [29:42<04:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2853/3295 [29:42<04:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2853/3295 [29:42<04:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2854/3295 [29:42<04:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2854/3295 [29:42<04:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2855/3295 [29:44<04:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2855/3295 [29:44<04:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2856/3295 [29:44<04:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2856/3295 [29:44<04:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2857/3295 [29:44<04:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2857/3295 [29:44<04:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2858/3295 [29:45<04:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2858/3295 [29:45<04:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2859/3295 [29:45<04:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2859/3295 [29:45<04:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2860/3295 [29:45<04:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2860/3295 [29:45<04:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2861/3295 [29:47<04:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2861/3295 [29:47<04:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2862/3295 [29:47<04:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2862/3295 [29:47<04:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2863/3295 [29:47<04:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2863/3295 [29:47<04:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2864/3295 [29:49<04:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2864/3295 [29:49<04:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2865/3295 [29:49<04:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2865/3295 [29:49<04:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2866/3295 [29:49<04:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2866/3295 [29:49<04:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2867/3295 [29:51<04:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2867/3295 [29:51<04:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2868/3295 [29:51<04:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2868/3295 [29:51<04:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2869/3295 [29:51<04:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2869/3295 [29:51<04:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2870/3295 [29:53<04:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2870/3295 [29:53<04:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2871/3295 [29:53<04:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2871/3295 [29:53<04:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2872/3295 [29:53<04:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2872/3295 [29:53<04:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2873/3295 [29:55<04:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2873/3295 [29:55<04:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2874/3295 [29:55<04:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2874/3295 [29:55<04:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2875/3295 [29:55<04:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2875/3295 [29:55<04:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2876/3295 [29:56<04:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2876/3295 [29:56<04:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2877/3295 [29:56<04:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2877/3295 [29:56<04:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2878/3295 [29:56<04:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2878/3295 [29:56<04:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2879/3295 [29:58<04:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2879/3295 [29:58<04:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2880/3295 [29:58<04:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2880/3295 [29:58<04:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2881/3295 [29:58<04:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2881/3295 [29:58<04:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2882/3295 [30:00<04:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2882/3295 [30:00<04:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2883/3295 [30:00<04:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  87%|████████▋ | 2883/3295 [30:00<04:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2884/3295 [30:00<04:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2884/3295 [30:00<04:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2885/3295 [30:02<04:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2885/3295 [30:02<04:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2886/3295 [30:02<04:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2886/3295 [30:02<04:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2887/3295 [30:02<04:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2887/3295 [30:02<04:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2888/3295 [30:04<04:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2888/3295 [30:04<04:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2889/3295 [30:04<04:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2889/3295 [30:04<04:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2890/3295 [30:04<04:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2890/3295 [30:04<04:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2891/3295 [30:06<04:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2891/3295 [30:06<04:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2892/3295 [30:06<04:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2892/3295 [30:06<04:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2893/3295 [30:06<04:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2893/3295 [30:06<04:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2894/3295 [30:08<04:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2894/3295 [30:08<04:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2895/3295 [30:08<04:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2895/3295 [30:08<04:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2896/3295 [30:08<04:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2896/3295 [30:08<04:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2897/3295 [30:09<04:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2897/3295 [30:09<04:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2898/3295 [30:10<04:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2898/3295 [30:10<04:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2899/3295 [30:10<04:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2899/3295 [30:10<04:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2900/3295 [30:11<04:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2900/3295 [30:11<04:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2901/3295 [30:11<04:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2901/3295 [30:11<04:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2902/3295 [30:11<04:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2902/3295 [30:11<04:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2903/3295 [30:13<04:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2903/3295 [30:13<04:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2904/3295 [30:13<04:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2904/3295 [30:13<04:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2905/3295 [30:13<04:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2905/3295 [30:13<04:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2906/3295 [30:15<04:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2906/3295 [30:15<04:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2907/3295 [30:15<04:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2907/3295 [30:15<04:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2908/3295 [30:15<04:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2908/3295 [30:15<04:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2909/3295 [30:17<04:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2909/3295 [30:17<04:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2910/3295 [30:17<04:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2910/3295 [30:17<04:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2911/3295 [30:17<03:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2911/3295 [30:17<03:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2912/3295 [30:19<03:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2912/3295 [30:19<03:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2913/3295 [30:19<03:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2913/3295 [30:19<03:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2914/3295 [30:19<03:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2914/3295 [30:19<03:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2915/3295 [30:21<03:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2915/3295 [30:21<03:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2916/3295 [30:21<03:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  88%|████████▊ | 2916/3295 [30:21<03:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2917/3295 [30:21<03:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2917/3295 [30:21<03:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2918/3295 [30:22<03:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2918/3295 [30:22<03:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2919/3295 [30:22<03:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2919/3295 [30:22<03:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2920/3295 [30:22<03:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2920/3295 [30:22<03:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2921/3295 [30:24<03:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2921/3295 [30:24<03:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2922/3295 [30:24<03:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2922/3295 [30:24<03:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2923/3295 [30:24<03:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2923/3295 [30:24<03:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2924/3295 [30:26<03:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▊ | 2924/3295 [30:26<03:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2925/3295 [30:26<03:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2925/3295 [30:26<03:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2926/3295 [30:26<03:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2926/3295 [30:26<03:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2927/3295 [30:28<03:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2927/3295 [30:28<03:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2928/3295 [30:28<03:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2928/3295 [30:28<03:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2929/3295 [30:28<03:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2929/3295 [30:28<03:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2930/3295 [30:30<03:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2930/3295 [30:30<03:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2931/3295 [30:30<03:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2931/3295 [30:30<03:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2932/3295 [30:30<03:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2932/3295 [30:30<03:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2933/3295 [30:32<03:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2933/3295 [30:32<03:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2934/3295 [30:32<03:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2934/3295 [30:32<03:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2935/3295 [30:32<03:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2935/3295 [30:32<03:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2936/3295 [30:34<03:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2936/3295 [30:34<03:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2937/3295 [30:34<03:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2937/3295 [30:34<03:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2938/3295 [30:34<03:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2938/3295 [30:34<03:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2939/3295 [30:35<03:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2939/3295 [30:35<03:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2940/3295 [30:35<03:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2940/3295 [30:35<03:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2941/3295 [30:35<03:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2941/3295 [30:35<03:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2942/3295 [30:37<03:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2942/3295 [30:37<03:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2943/3295 [30:37<03:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2943/3295 [30:37<03:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2944/3295 [30:37<03:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2944/3295 [30:37<03:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2945/3295 [30:39<03:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2945/3295 [30:39<03:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2946/3295 [30:39<03:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2946/3295 [30:39<03:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2947/3295 [30:39<03:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2947/3295 [30:39<03:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2948/3295 [30:41<03:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2948/3295 [30:41<03:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2949/3295 [30:41<03:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  89%|████████▉ | 2949/3295 [30:41<03:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2950/3295 [30:41<03:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2950/3295 [30:41<03:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2951/3295 [30:43<03:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2951/3295 [30:43<03:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2952/3295 [30:43<03:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2952/3295 [30:43<03:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2953/3295 [30:43<03:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2953/3295 [30:43<03:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2954/3295 [30:45<03:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2954/3295 [30:45<03:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2955/3295 [30:45<03:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2955/3295 [30:45<03:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2956/3295 [30:45<03:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2956/3295 [30:45<03:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2957/3295 [30:46<03:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2957/3295 [30:46<03:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2958/3295 [30:46<03:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2958/3295 [30:46<03:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2959/3295 [30:46<03:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2959/3295 [30:46<03:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2960/3295 [30:48<03:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2960/3295 [30:48<03:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2961/3295 [30:48<03:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2961/3295 [30:48<03:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2962/3295 [30:48<03:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2962/3295 [30:48<03:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2963/3295 [30:50<03:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2963/3295 [30:50<03:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2964/3295 [30:50<03:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2964/3295 [30:50<03:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2965/3295 [30:50<03:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|████████▉ | 2965/3295 [30:50<03:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2966/3295 [30:52<03:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2966/3295 [30:52<03:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2967/3295 [30:52<03:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2967/3295 [30:52<03:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2968/3295 [30:52<03:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2968/3295 [30:52<03:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2969/3295 [30:54<03:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2969/3295 [30:54<03:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2970/3295 [30:54<03:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2970/3295 [30:54<03:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2971/3295 [30:54<03:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2971/3295 [30:54<03:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2972/3295 [30:56<03:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2972/3295 [30:56<03:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2973/3295 [30:56<03:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2973/3295 [30:56<03:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2974/3295 [30:56<03:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2974/3295 [30:56<03:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2975/3295 [30:57<03:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2975/3295 [30:57<03:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2976/3295 [30:57<03:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2976/3295 [30:57<03:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2977/3295 [30:57<03:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2977/3295 [30:57<03:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2978/3295 [30:59<03:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2978/3295 [30:59<03:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2979/3295 [30:59<03:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2979/3295 [30:59<03:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2980/3295 [30:59<03:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2980/3295 [30:59<03:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2981/3295 [31:01<03:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  90%|█████████ | 2981/3295 [31:01<03:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2982/3295 [31:01<03:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2982/3295 [31:01<03:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2983/3295 [31:01<03:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2983/3295 [31:01<03:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2984/3295 [31:03<03:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2984/3295 [31:03<03:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2985/3295 [31:03<03:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2985/3295 [31:03<03:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2986/3295 [31:03<03:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2986/3295 [31:03<03:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2987/3295 [31:05<03:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2987/3295 [31:05<03:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2988/3295 [31:05<03:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2988/3295 [31:05<03:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2989/3295 [31:05<03:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2989/3295 [31:05<03:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2990/3295 [31:07<03:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2990/3295 [31:07<03:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2991/3295 [31:07<03:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2991/3295 [31:07<03:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2992/3295 [31:07<03:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2992/3295 [31:07<03:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2993/3295 [31:09<03:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2993/3295 [31:09<03:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2994/3295 [31:09<03:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2994/3295 [31:09<03:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2995/3295 [31:09<03:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2995/3295 [31:09<03:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2996/3295 [31:11<03:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2996/3295 [31:11<03:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2997/3295 [31:11<03:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2997/3295 [31:11<03:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2998/3295 [31:11<03:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2998/3295 [31:11<03:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2999/3295 [31:12<03:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 2999/3295 [31:12<03:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3000/3295 [31:12<03:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3000/3295 [31:12<03:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3001/3295 [31:13<03:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3001/3295 [31:13<03:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3002/3295 [31:14<03:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3002/3295 [31:14<03:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3003/3295 [31:14<03:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3003/3295 [31:14<03:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3004/3295 [31:14<03:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3004/3295 [31:14<03:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3005/3295 [31:16<03:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3005/3295 [31:16<03:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3006/3295 [31:16<03:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████ | 3006/3295 [31:16<03:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3007/3295 [31:16<02:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3007/3295 [31:16<02:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3008/3295 [31:18<02:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3008/3295 [31:18<02:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3009/3295 [31:18<02:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3009/3295 [31:18<02:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3010/3295 [31:18<02:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3010/3295 [31:18<02:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3011/3295 [31:20<02:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3011/3295 [31:20<02:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3012/3295 [31:20<02:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3012/3295 [31:20<02:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3013/3295 [31:20<02:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3013/3295 [31:20<02:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3014/3295 [31:22<02:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  91%|█████████▏| 3014/3295 [31:22<02:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3015/3295 [31:22<02:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3015/3295 [31:22<02:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3016/3295 [31:22<02:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3016/3295 [31:22<02:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3017/3295 [31:24<02:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3017/3295 [31:24<02:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3018/3295 [31:24<02:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3018/3295 [31:24<02:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3019/3295 [31:24<02:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3019/3295 [31:24<02:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3020/3295 [31:26<02:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3020/3295 [31:26<02:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3021/3295 [31:26<02:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3021/3295 [31:26<02:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3022/3295 [31:26<02:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3022/3295 [31:26<02:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3023/3295 [31:28<02:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3023/3295 [31:28<02:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3024/3295 [31:28<02:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3024/3295 [31:28<02:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3025/3295 [31:28<02:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3025/3295 [31:28<02:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3026/3295 [31:29<02:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3026/3295 [31:29<02:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3027/3295 [31:29<02:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3027/3295 [31:29<02:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3028/3295 [31:29<02:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3028/3295 [31:29<02:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3029/3295 [31:31<02:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3029/3295 [31:31<02:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3030/3295 [31:31<02:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3030/3295 [31:31<02:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3031/3295 [31:31<02:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3031/3295 [31:31<02:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3032/3295 [31:33<02:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3032/3295 [31:33<02:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3033/3295 [31:33<02:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3033/3295 [31:33<02:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3034/3295 [31:33<02:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3034/3295 [31:33<02:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3035/3295 [31:35<02:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3035/3295 [31:35<02:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3036/3295 [31:35<02:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3036/3295 [31:35<02:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3037/3295 [31:35<02:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3037/3295 [31:35<02:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3038/3295 [31:38<02:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3038/3295 [31:38<02:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3039/3295 [31:38<02:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3039/3295 [31:38<02:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3040/3295 [31:38<02:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3040/3295 [31:38<02:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3041/3295 [31:40<02:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3041/3295 [31:40<02:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3042/3295 [31:40<02:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3042/3295 [31:40<02:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3043/3295 [31:40<02:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3043/3295 [31:40<02:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3044/3295 [31:42<02:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3044/3295 [31:42<02:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3045/3295 [31:42<02:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3045/3295 [31:42<02:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3046/3295 [31:42<02:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3046/3295 [31:42<02:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3047/3295 [31:44<02:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  92%|█████████▏| 3047/3295 [31:44<02:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3048/3295 [31:44<02:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3048/3295 [31:44<02:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3049/3295 [31:44<02:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3049/3295 [31:44<02:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3050/3295 [31:45<02:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3050/3295 [31:45<02:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3051/3295 [31:45<02:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3051/3295 [31:45<02:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3052/3295 [31:45<02:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3052/3295 [31:45<02:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3053/3295 [31:47<02:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3053/3295 [31:47<02:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3054/3295 [31:47<02:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3054/3295 [31:47<02:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3055/3295 [31:47<02:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3055/3295 [31:47<02:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3056/3295 [31:49<02:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3056/3295 [31:49<02:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3057/3295 [31:49<02:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3057/3295 [31:49<02:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3058/3295 [31:49<02:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3058/3295 [31:49<02:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3059/3295 [31:51<02:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3059/3295 [31:51<02:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3060/3295 [31:51<02:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3060/3295 [31:51<02:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3061/3295 [31:51<02:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3061/3295 [31:51<02:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3062/3295 [31:53<02:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3062/3295 [31:53<02:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3063/3295 [31:53<02:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3063/3295 [31:53<02:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3064/3295 [31:53<02:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3064/3295 [31:53<02:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3065/3295 [31:55<02:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3065/3295 [31:55<02:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3066/3295 [31:55<02:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3066/3295 [31:55<02:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3067/3295 [31:55<02:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3067/3295 [31:55<02:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3068/3295 [31:57<02:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3068/3295 [31:57<02:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3069/3295 [31:57<02:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3069/3295 [31:57<02:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3070/3295 [31:57<02:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3070/3295 [31:57<02:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3071/3295 [31:59<02:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3071/3295 [31:59<02:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3072/3295 [31:59<02:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3072/3295 [31:59<02:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3073/3295 [31:59<02:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3073/3295 [31:59<02:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3074/3295 [32:02<02:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3074/3295 [32:02<02:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3075/3295 [32:02<02:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3075/3295 [32:02<02:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3076/3295 [32:02<02:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3076/3295 [32:02<02:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3077/3295 [32:03<02:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3077/3295 [32:03<02:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3078/3295 [32:03<02:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3078/3295 [32:03<02:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3079/3295 [32:03<02:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3079/3295 [32:03<02:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3080/3295 [32:05<02:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  93%|█████████▎| 3080/3295 [32:05<02:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3081/3295 [32:05<02:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3081/3295 [32:05<02:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3082/3295 [32:05<02:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3082/3295 [32:05<02:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3083/3295 [32:07<02:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3083/3295 [32:07<02:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3084/3295 [32:07<02:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3084/3295 [32:07<02:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3085/3295 [32:07<02:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3085/3295 [32:07<02:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3086/3295 [32:09<02:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3086/3295 [32:09<02:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3087/3295 [32:09<02:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3087/3295 [32:09<02:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3088/3295 [32:09<02:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3088/3295 [32:09<02:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3089/3295 [32:11<02:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▎| 3089/3295 [32:11<02:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3090/3295 [32:11<02:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3090/3295 [32:11<02:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3091/3295 [32:11<02:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3091/3295 [32:11<02:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3092/3295 [32:13<02:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3092/3295 [32:13<02:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3093/3295 [32:13<02:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3093/3295 [32:13<02:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3094/3295 [32:13<02:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3094/3295 [32:13<02:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3095/3295 [32:15<02:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3095/3295 [32:15<02:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3096/3295 [32:15<02:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3096/3295 [32:15<02:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3097/3295 [32:15<02:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3097/3295 [32:15<02:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3098/3295 [32:17<02:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3098/3295 [32:17<02:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3099/3295 [32:17<02:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3099/3295 [32:17<02:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3100/3295 [32:17<02:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3100/3295 [32:17<02:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3101/3295 [32:19<02:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3101/3295 [32:19<02:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3102/3295 [32:19<02:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3102/3295 [32:19<02:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3103/3295 [32:19<02:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3103/3295 [32:19<02:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3104/3295 [32:21<01:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3104/3295 [32:21<01:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3105/3295 [32:21<01:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3105/3295 [32:21<01:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3106/3295 [32:21<01:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3106/3295 [32:21<01:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3107/3295 [32:23<01:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3107/3295 [32:23<01:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3108/3295 [32:23<01:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3108/3295 [32:23<01:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3109/3295 [32:23<01:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3109/3295 [32:23<01:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3110/3295 [32:25<01:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3110/3295 [32:25<01:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3111/3295 [32:25<01:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3111/3295 [32:25<01:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3112/3295 [32:25<01:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3112/3295 [32:25<01:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3113/3295 [32:26<01:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  94%|█████████▍| 3113/3295 [32:26<01:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3114/3295 [32:26<01:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3114/3295 [32:26<01:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3115/3295 [32:26<01:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3115/3295 [32:26<01:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3116/3295 [32:28<01:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3116/3295 [32:28<01:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3117/3295 [32:28<01:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3117/3295 [32:28<01:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3118/3295 [32:28<01:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3118/3295 [32:28<01:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3119/3295 [32:30<01:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3119/3295 [32:30<01:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3120/3295 [32:30<01:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3120/3295 [32:30<01:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3121/3295 [32:30<01:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3121/3295 [32:30<01:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3122/3295 [32:32<01:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3122/3295 [32:32<01:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3123/3295 [32:32<01:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3123/3295 [32:32<01:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3124/3295 [32:32<01:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3124/3295 [32:32<01:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3125/3295 [32:34<01:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3125/3295 [32:34<01:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3126/3295 [32:34<01:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3126/3295 [32:34<01:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3127/3295 [32:34<01:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3127/3295 [32:34<01:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3128/3295 [32:36<01:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3128/3295 [32:36<01:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3129/3295 [32:36<01:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3129/3295 [32:36<01:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3130/3295 [32:36<01:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▍| 3130/3295 [32:36<01:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3131/3295 [32:38<01:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3131/3295 [32:38<01:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3132/3295 [32:38<01:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3132/3295 [32:38<01:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3133/3295 [32:38<01:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3133/3295 [32:38<01:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3134/3295 [32:40<01:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3134/3295 [32:40<01:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3135/3295 [32:40<01:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3135/3295 [32:40<01:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3136/3295 [32:40<01:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3136/3295 [32:40<01:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3137/3295 [32:42<01:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3137/3295 [32:42<01:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3138/3295 [32:42<01:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3138/3295 [32:42<01:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3139/3295 [32:42<01:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3139/3295 [32:42<01:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3140/3295 [32:44<01:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3140/3295 [32:44<01:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3141/3295 [32:44<01:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3141/3295 [32:44<01:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3142/3295 [32:44<01:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3142/3295 [32:44<01:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3143/3295 [32:46<01:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3143/3295 [32:46<01:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3144/3295 [32:46<01:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3144/3295 [32:46<01:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3145/3295 [32:46<01:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3145/3295 [32:46<01:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3146/3295 [32:48<01:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  95%|█████████▌| 3146/3295 [32:48<01:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3147/3295 [32:48<01:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3147/3295 [32:48<01:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3148/3295 [32:48<01:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3148/3295 [32:48<01:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3149/3295 [32:49<01:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3149/3295 [32:49<01:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3150/3295 [32:49<01:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3150/3295 [32:49<01:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3151/3295 [32:49<01:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3151/3295 [32:49<01:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3152/3295 [32:51<01:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3152/3295 [32:51<01:29,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3153/3295 [32:51<01:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3153/3295 [32:51<01:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3154/3295 [32:51<01:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3154/3295 [32:51<01:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3155/3295 [32:53<01:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3155/3295 [32:53<01:27,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3156/3295 [32:53<01:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3156/3295 [32:53<01:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3157/3295 [32:53<01:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3157/3295 [32:53<01:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3158/3295 [32:55<01:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3158/3295 [32:55<01:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3159/3295 [32:55<01:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3159/3295 [32:55<01:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3160/3295 [32:55<01:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3160/3295 [32:55<01:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3161/3295 [32:57<01:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3161/3295 [32:57<01:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3162/3295 [32:57<01:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3162/3295 [32:57<01:23,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3163/3295 [32:57<01:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3163/3295 [32:57<01:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3164/3295 [32:59<01:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3164/3295 [32:59<01:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3165/3295 [32:59<01:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3165/3295 [32:59<01:21,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3166/3295 [32:59<01:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3166/3295 [32:59<01:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3167/3295 [33:01<01:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3167/3295 [33:01<01:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3168/3295 [33:01<01:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3168/3295 [33:01<01:19,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3169/3295 [33:01<01:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3169/3295 [33:01<01:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3170/3295 [33:03<01:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3170/3295 [33:03<01:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3171/3295 [33:03<01:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▌| 3171/3295 [33:03<01:17,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3172/3295 [33:03<01:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3172/3295 [33:03<01:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3173/3295 [33:05<01:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3173/3295 [33:05<01:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3174/3295 [33:05<01:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3174/3295 [33:05<01:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3175/3295 [33:05<01:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3175/3295 [33:05<01:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3176/3295 [33:07<01:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3176/3295 [33:07<01:14,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3177/3295 [33:07<01:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3177/3295 [33:07<01:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3178/3295 [33:07<01:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3178/3295 [33:07<01:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3179/3295 [33:08<01:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  96%|█████████▋| 3179/3295 [33:08<01:12,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3180/3295 [33:08<01:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3180/3295 [33:08<01:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3181/3295 [33:08<01:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3181/3295 [33:08<01:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3182/3295 [33:10<01:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3182/3295 [33:10<01:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3183/3295 [33:10<01:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3183/3295 [33:10<01:10,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3184/3295 [33:10<01:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3184/3295 [33:10<01:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3185/3295 [33:12<01:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3185/3295 [33:12<01:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3186/3295 [33:12<01:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3186/3295 [33:12<01:08,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3187/3295 [33:12<01:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3187/3295 [33:12<01:07,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3188/3295 [33:14<01:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3188/3295 [33:14<01:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3189/3295 [33:14<01:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3189/3295 [33:14<01:06,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3190/3295 [33:14<01:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3190/3295 [33:14<01:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3191/3295 [33:16<01:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3191/3295 [33:16<01:05,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3192/3295 [33:16<01:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3192/3295 [33:16<01:04,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3193/3295 [33:16<01:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3193/3295 [33:16<01:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3194/3295 [33:19<01:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3194/3295 [33:19<01:03,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3195/3295 [33:19<01:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3195/3295 [33:19<01:02,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3196/3295 [33:19<01:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3196/3295 [33:19<01:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3197/3295 [33:21<01:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3197/3295 [33:21<01:01,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3198/3295 [33:21<01:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3198/3295 [33:21<01:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3199/3295 [33:21<01:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3199/3295 [33:21<01:00,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3200/3295 [33:24<00:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3200/3295 [33:24<00:59,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3201/3295 [33:24<00:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3201/3295 [33:24<00:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3202/3295 [33:24<00:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3202/3295 [33:24<00:58,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3203/3295 [33:26<00:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3203/3295 [33:26<00:57,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3204/3295 [33:26<00:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3204/3295 [33:26<00:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3205/3295 [33:26<00:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3205/3295 [33:26<00:56,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3206/3295 [33:29<00:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3206/3295 [33:29<00:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3207/3295 [33:29<00:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3207/3295 [33:29<00:55,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3208/3295 [33:29<00:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3208/3295 [33:29<00:54,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3209/3295 [33:31<00:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3209/3295 [33:31<00:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3210/3295 [33:31<00:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3210/3295 [33:31<00:53,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3211/3295 [33:31<00:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3211/3295 [33:31<00:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3212/3295 [33:32<00:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  97%|█████████▋| 3212/3295 [33:32<00:52,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3213/3295 [33:32<00:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3213/3295 [33:32<00:51,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3214/3295 [33:33<00:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3214/3295 [33:33<00:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3215/3295 [33:35<00:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3215/3295 [33:35<00:50,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3216/3295 [33:35<00:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3216/3295 [33:35<00:49,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3217/3295 [33:35<00:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3217/3295 [33:35<00:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3218/3295 [33:37<00:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3218/3295 [33:37<00:48,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3219/3295 [33:37<00:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3219/3295 [33:37<00:47,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3220/3295 [33:37<00:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3220/3295 [33:37<00:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3221/3295 [33:39<00:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3221/3295 [33:39<00:46,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3222/3295 [33:39<00:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3222/3295 [33:39<00:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3223/3295 [33:39<00:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3223/3295 [33:39<00:45,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3224/3295 [33:40<00:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3224/3295 [33:40<00:44,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3225/3295 [33:41<00:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3225/3295 [33:41<00:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3226/3295 [33:41<00:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3226/3295 [33:41<00:43,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3227/3295 [33:43<00:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3227/3295 [33:43<00:42,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3228/3295 [33:43<00:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3228/3295 [33:43<00:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3229/3295 [33:43<00:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3229/3295 [33:43<00:41,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3230/3295 [33:44<00:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3230/3295 [33:44<00:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3231/3295 [33:44<00:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3231/3295 [33:44<00:40,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3232/3295 [33:44<00:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3232/3295 [33:44<00:39,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3233/3295 [33:47<00:38,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3233/3295 [33:47<00:38,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3234/3295 [33:47<00:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3234/3295 [33:47<00:38,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3235/3295 [33:47<00:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3235/3295 [33:47<00:37,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3236/3295 [33:48<00:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3236/3295 [33:48<00:36,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3237/3295 [33:48<00:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3237/3295 [33:48<00:36,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3238/3295 [33:48<00:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3238/3295 [33:48<00:35,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3239/3295 [33:50<00:35,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3239/3295 [33:50<00:35,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3240/3295 [33:50<00:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3240/3295 [33:50<00:34,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3241/3295 [33:50<00:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3241/3295 [33:50<00:33,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3242/3295 [33:52<00:33,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3242/3295 [33:52<00:33,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3243/3295 [33:52<00:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3243/3295 [33:52<00:32,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3244/3295 [33:52<00:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3244/3295 [33:52<00:31,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3245/3295 [33:54<00:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  98%|█████████▊| 3245/3295 [33:54<00:31,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3246/3295 [33:54<00:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3246/3295 [33:54<00:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3247/3295 [33:54<00:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3247/3295 [33:54<00:30,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3248/3295 [33:56<00:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3248/3295 [33:56<00:29,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3249/3295 [33:56<00:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3249/3295 [33:56<00:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3250/3295 [33:56<00:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3250/3295 [33:56<00:28,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3251/3295 [33:58<00:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3251/3295 [33:58<00:27,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3252/3295 [33:58<00:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3252/3295 [33:58<00:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3253/3295 [33:58<00:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▊| 3253/3295 [33:58<00:26,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3254/3295 [34:00<00:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3254/3295 [34:00<00:25,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3255/3295 [34:00<00:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3255/3295 [34:00<00:25,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3256/3295 [34:00<00:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3256/3295 [34:00<00:24,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3257/3295 [34:02<00:23,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3257/3295 [34:02<00:23,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3258/3295 [34:02<00:23,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3258/3295 [34:02<00:23,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3259/3295 [34:02<00:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3259/3295 [34:02<00:22,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3260/3295 [34:04<00:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3260/3295 [34:04<00:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3261/3295 [34:04<00:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3261/3295 [34:04<00:21,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3262/3295 [34:04<00:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3262/3295 [34:04<00:20,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3263/3295 [34:06<00:20,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3263/3295 [34:06<00:20,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3264/3295 [34:06<00:19,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3264/3295 [34:06<00:19,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3265/3295 [34:06<00:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3265/3295 [34:06<00:18,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3266/3295 [34:08<00:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3266/3295 [34:08<00:18,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3267/3295 [34:08<00:17,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3267/3295 [34:08<00:17,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3268/3295 [34:08<00:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3268/3295 [34:08<00:16,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3269/3295 [34:10<00:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3269/3295 [34:10<00:16,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3270/3295 [34:10<00:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3270/3295 [34:10<00:15,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3271/3295 [34:10<00:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3271/3295 [34:10<00:15,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3272/3295 [34:12<00:14,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3272/3295 [34:12<00:14,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3273/3295 [34:12<00:13,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3273/3295 [34:12<00:13,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3274/3295 [34:12<00:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3274/3295 [34:12<00:13,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3275/3295 [34:14<00:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3275/3295 [34:14<00:12,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3276/3295 [34:14<00:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3276/3295 [34:14<00:11,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3277/3295 [34:14<00:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3277/3295 [34:14<00:11,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3278/3295 [34:16<00:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3:  99%|█████████▉| 3278/3295 [34:16<00:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3279/3295 [34:16<00:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3279/3295 [34:16<00:10,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3280/3295 [34:16<00:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3280/3295 [34:16<00:09,  1.60it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3281/3295 [34:18<00:08,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3281/3295 [34:18<00:08,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3282/3295 [34:18<00:08,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3282/3295 [34:18<00:08,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3283/3295 [34:18<00:07,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3283/3295 [34:18<00:07,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3284/3295 [34:20<00:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3284/3295 [34:20<00:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3285/3295 [34:20<00:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3285/3295 [34:20<00:06,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3286/3295 [34:20<00:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3286/3295 [34:20<00:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3287/3295 [34:22<00:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3287/3295 [34:22<00:05,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3288/3295 [34:22<00:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3288/3295 [34:22<00:04,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3289/3295 [34:22<00:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3289/3295 [34:22<00:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3290/3295 [34:24<00:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3290/3295 [34:24<00:03,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3291/3295 [34:24<00:02,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3291/3295 [34:24<00:02,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3292/3295 [34:24<00:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3292/3295 [34:24<00:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3293/3295 [34:26<00:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3293/3295 [34:26<00:01,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3294/3295 [34:26<00:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|█████████▉| 3294/3295 [34:26<00:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|██████████| 3295/3295 [34:26<00:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|██████████| 3295/3295 [34:26<00:00,  1.59it/s, R@1 (val)=0.403, R@5 (val)=0.606, R@20 (val)=0.789, loss (train)=0.194, R@1 (train)=0.958]

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation: |          | 0/? [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 0/602 [00:00<?, ?it/s][A

Validation DataLoader 0:   0%|          | 1/602 [00:00<00:14, 40.19it/s][A

Validation DataLoader 0:   0%|          | 2/602 [00:00<00:14, 42.60it/s][A

Validation DataLoader 0:   0%|          | 3/602 [00:00<00:12, 47.22it/s][A

Validation DataLoader 0:   1%|          | 4/602 [00:00<00:11, 49.93it/s][A

Validation DataLoader 0:   1%|          | 5/602 [00:00<00:11, 52.00it/s][A

Validation DataLoader 0:   1%|          | 6/602 [00:00<00:11, 53.34it/s][A

Validation DataLoader 0:   1%|          | 7/602 [00:00<00:35, 16.62it/s][A

Validation DataLoader 0:   1%|▏         | 8/602 [00:00<00:45, 13.14it/s][A

Validation DataLoader 0:   1%|▏         | 9/602 [00:00<00:41, 14.37it/s][A

Validation DataLoader 0:   2%|▏         | 10/602 [00:00<00:50, 11.80it/s][A

Validation DataLoader 0:   2%|▏         | 11/602 [00:01<00:56, 10.37it/s][A

Validation DataLoader 0:   2%|▏         | 12/602 [00:01<00:52, 11.14it/s][A

Validation DataLoader 0:   2%|▏         | 13/602 [00:01<01:00,  9.78it/s][A

Validation DataLoader 0:   2%|▏         | 14/602 [00:01<01:03,  9.32it/s][A

Validation DataLoader 0:   2%|▏         | 15/602 [00:01<00:59,  9.89it/s][A

Validation DataLoader 0:   3%|▎         | 16/602 [00:01<01:07,  8.72it/s][A

Validation DataLoader 0:   3%|▎         | 17/602 [00:01<01:04,  9.01it/s][A

Validation DataLoader 0:   3%|▎         | 18/602 [00:02<01:06,  8.72it/s][A

Validation DataLoader 0:   3%|▎         | 19/602 [00:02<01:13,  7.96it/s][A

Validation DataLoader 0:   3%|▎         | 20/602 [00:02<01:10,  8.24it/s][A

Validation DataLoader 0:   3%|▎         | 21/602 [00:02<01:12,  8.04it/s][A

Validation DataLoader 0:   4%|▎         | 22/602 [00:02<01:13,  7.93it/s][A

Validation DataLoader 0:   4%|▍         | 23/602 [00:02<01:10,  8.24it/s][A

Validation DataLoader 0:   4%|▍         | 24/602 [00:03<01:14,  7.72it/s][A

Validation DataLoader 0:   4%|▍         | 25/602 [00:03<01:15,  7.66it/s][A

Validation DataLoader 0:   4%|▍         | 26/602 [00:03<01:13,  7.80it/s][A

Validation DataLoader 0:   4%|▍         | 27/602 [00:03<01:16,  7.50it/s][A

Validation DataLoader 0:   5%|▍         | 28/602 [00:03<01:18,  7.32it/s][A

Validation DataLoader 0:   5%|▍         | 29/602 [00:03<01:17,  7.42it/s][A

Validation DataLoader 0:   5%|▍         | 30/602 [00:04<01:19,  7.24it/s][A

Validation DataLoader 0:   5%|▌         | 31/602 [00:04<01:19,  7.23it/s][A

Validation DataLoader 0:   5%|▌         | 32/602 [00:04<01:16,  7.43it/s][A

Validation DataLoader 0:   5%|▌         | 33/602 [00:04<01:17,  7.30it/s][A

Validation DataLoader 0:   6%|▌         | 34/602 [00:04<01:20,  7.09it/s][A

Validation DataLoader 0:   6%|▌         | 35/602 [00:04<01:18,  7.25it/s][A

Validation DataLoader 0:   6%|▌         | 36/602 [00:04<01:18,  7.25it/s][A

Validation DataLoader 0:   6%|▌         | 37/602 [00:05<01:19,  7.06it/s][A

Validation DataLoader 0:   6%|▋         | 38/602 [00:05<01:19,  7.11it/s][A

Validation DataLoader 0:   6%|▋         | 39/602 [00:05<01:18,  7.17it/s][A

Validation DataLoader 0:   7%|▋         | 40/602 [00:05<01:20,  7.01it/s][A

Validation DataLoader 0:   7%|▋         | 41/602 [00:05<01:19,  7.10it/s][A

Validation DataLoader 0:   7%|▋         | 42/602 [00:05<01:18,  7.10it/s][A

Validation DataLoader 0:   7%|▋         | 43/602 [00:06<01:19,  7.02it/s][A

Validation DataLoader 0:   7%|▋         | 44/602 [00:06<01:19,  7.00it/s][A

Validation DataLoader 0:   7%|▋         | 45/602 [00:06<01:18,  7.09it/s][A

Validation DataLoader 0:   8%|▊         | 46/602 [00:06<01:19,  6.96it/s][A

Validation DataLoader 0:   8%|▊         | 47/602 [00:06<01:19,  7.01it/s][A

Validation DataLoader 0:   8%|▊         | 48/602 [00:06<01:18,  7.07it/s][A

Validation DataLoader 0:   8%|▊         | 49/602 [00:07<01:20,  6.89it/s][A

Validation DataLoader 0:   8%|▊         | 50/602 [00:07<01:18,  7.02it/s][A

Validation DataLoader 0:   8%|▊         | 51/602 [00:07<01:18,  7.06it/s][A

Validation DataLoader 0:   9%|▊         | 52/602 [00:07<01:20,  6.87it/s][A

Validation DataLoader 0:   9%|▉         | 53/602 [00:07<01:18,  6.99it/s][A

Validation DataLoader 0:   9%|▉         | 54/602 [00:07<01:17,  7.05it/s][A

Validation DataLoader 0:   9%|▉         | 55/602 [00:08<01:20,  6.83it/s][A

Validation DataLoader 0:   9%|▉         | 56/602 [00:08<01:18,  6.94it/s][A

Validation DataLoader 0:   9%|▉         | 57/602 [00:08<01:17,  7.05it/s][A

Validation DataLoader 0:  10%|▉         | 58/602 [00:08<01:20,  6.78it/s][A

Validation DataLoader 0:  10%|▉         | 59/602 [00:08<01:18,  6.89it/s][A

Validation DataLoader 0:  10%|▉         | 60/602 [00:08<01:17,  6.99it/s][A

Validation DataLoader 0:  10%|█         | 61/602 [00:09<01:20,  6.73it/s][A

Validation DataLoader 0:  10%|█         | 62/602 [00:09<01:19,  6.83it/s][A

Validation DataLoader 0:  10%|█         | 63/602 [00:09<01:17,  6.93it/s][A

Validation DataLoader 0:  11%|█         | 64/602 [00:09<01:20,  6.68it/s][A

Validation DataLoader 0:  11%|█         | 65/602 [00:09<01:19,  6.77it/s][A

Validation DataLoader 0:  11%|█         | 66/602 [00:09<01:18,  6.87it/s][A

Validation DataLoader 0:  11%|█         | 67/602 [00:10<01:20,  6.63it/s][A

Validation DataLoader 0:  11%|█▏        | 68/602 [00:10<01:19,  6.72it/s][A

Validation DataLoader 0:  11%|█▏        | 69/602 [00:10<01:18,  6.81it/s][A

Validation DataLoader 0:  12%|█▏        | 70/602 [00:10<01:19,  6.67it/s][A

Validation DataLoader 0:  12%|█▏        | 71/602 [00:10<01:18,  6.75it/s][A

Validation DataLoader 0:  12%|█▏        | 72/602 [00:10<01:17,  6.84it/s][A

Validation DataLoader 0:  12%|█▏        | 73/602 [00:11<01:20,  6.58it/s][A

Validation DataLoader 0:  12%|█▏        | 74/602 [00:11<01:19,  6.66it/s][A

Validation DataLoader 0:  12%|█▏        | 75/602 [00:11<01:18,  6.74it/s][A

Validation DataLoader 0:  13%|█▎        | 76/602 [00:11<01:19,  6.58it/s][A

Validation DataLoader 0:  13%|█▎        | 77/602 [00:11<01:18,  6.66it/s][A

Validation DataLoader 0:  13%|█▎        | 78/602 [00:11<01:17,  6.74it/s][A

Validation DataLoader 0:  13%|█▎        | 79/602 [00:11<01:18,  6.63it/s][A

Validation DataLoader 0:  13%|█▎        | 80/602 [00:11<01:17,  6.71it/s][A

Validation DataLoader 0:  13%|█▎        | 81/602 [00:11<01:16,  6.78it/s][A

Validation DataLoader 0:  14%|█▎        | 82/602 [00:12<01:18,  6.64it/s][A

Validation DataLoader 0:  14%|█▍        | 83/602 [00:12<01:17,  6.71it/s][A

Validation DataLoader 0:  14%|█▍        | 84/602 [00:12<01:16,  6.78it/s][A

Validation DataLoader 0:  14%|█▍        | 85/602 [00:12<01:17,  6.64it/s][A

Validation DataLoader 0:  14%|█▍        | 86/602 [00:12<01:16,  6.71it/s][A

Validation DataLoader 0:  14%|█▍        | 87/602 [00:12<01:16,  6.77it/s][A

Validation DataLoader 0:  15%|█▍        | 88/602 [00:13<01:18,  6.56it/s][A

Validation DataLoader 0:  15%|█▍        | 89/602 [00:13<01:17,  6.62it/s][A

Validation DataLoader 0:  15%|█▍        | 90/602 [00:13<01:16,  6.69it/s][A

Validation DataLoader 0:  15%|█▌        | 91/602 [00:13<01:18,  6.52it/s][A

Validation DataLoader 0:  15%|█▌        | 92/602 [00:13<01:17,  6.59it/s][A

Validation DataLoader 0:  15%|█▌        | 93/602 [00:13<01:16,  6.65it/s][A

Validation DataLoader 0:  16%|█▌        | 94/602 [00:14<01:17,  6.53it/s][A

Validation DataLoader 0:  16%|█▌        | 95/602 [00:14<01:16,  6.59it/s][A

Validation DataLoader 0:  16%|█▌        | 96/602 [00:14<01:16,  6.65it/s][A

Validation DataLoader 0:  16%|█▌        | 97/602 [00:14<01:17,  6.54it/s][A

Validation DataLoader 0:  16%|█▋        | 98/602 [00:14<01:16,  6.61it/s][A

Validation DataLoader 0:  16%|█▋        | 99/602 [00:14<01:15,  6.67it/s][A

Validation DataLoader 0:  17%|█▋        | 100/602 [00:15<01:17,  6.52it/s][A

Validation DataLoader 0:  17%|█▋        | 101/602 [00:15<01:16,  6.57it/s][A

Validation DataLoader 0:  17%|█▋        | 102/602 [00:15<01:15,  6.63it/s][A

Validation DataLoader 0:  17%|█▋        | 103/602 [00:15<01:16,  6.51it/s][A

Validation DataLoader 0:  17%|█▋        | 104/602 [00:15<01:15,  6.56it/s][A

Validation DataLoader 0:  17%|█▋        | 105/602 [00:15<01:15,  6.62it/s][A

Validation DataLoader 0:  18%|█▊        | 106/602 [00:16<01:16,  6.52it/s][A

Validation DataLoader 0:  18%|█▊        | 107/602 [00:16<01:15,  6.57it/s][A

Validation DataLoader 0:  18%|█▊        | 108/602 [00:16<01:14,  6.63it/s][A

Validation DataLoader 0:  18%|█▊        | 109/602 [00:16<01:15,  6.56it/s][A

Validation DataLoader 0:  18%|█▊        | 110/602 [00:16<01:14,  6.61it/s][A

Validation DataLoader 0:  18%|█▊        | 111/602 [00:16<01:13,  6.67it/s][A

Validation DataLoader 0:  19%|█▊        | 112/602 [00:17<01:14,  6.57it/s][A

Validation DataLoader 0:  19%|█▉        | 113/602 [00:17<01:13,  6.62it/s][A

Validation DataLoader 0:  19%|█▉        | 114/602 [00:17<01:13,  6.67it/s][A

Validation DataLoader 0:  19%|█▉        | 115/602 [00:17<01:13,  6.58it/s][A

Validation DataLoader 0:  19%|█▉        | 116/602 [00:17<01:13,  6.63it/s][A

Validation DataLoader 0:  19%|█▉        | 117/602 [00:17<01:12,  6.68it/s][A

Validation DataLoader 0:  20%|█▉        | 118/602 [00:18<01:14,  6.53it/s][A

Validation DataLoader 0:  20%|█▉        | 119/602 [00:18<01:13,  6.58it/s][A

Validation DataLoader 0:  20%|█▉        | 120/602 [00:18<01:12,  6.63it/s][A

Validation DataLoader 0:  20%|██        | 121/602 [00:18<01:13,  6.51it/s][A

Validation DataLoader 0:  20%|██        | 122/602 [00:18<01:13,  6.56it/s][A

Validation DataLoader 0:  20%|██        | 123/602 [00:18<01:12,  6.61it/s][A

Validation DataLoader 0:  21%|██        | 124/602 [00:19<01:13,  6.50it/s][A

Validation DataLoader 0:  21%|██        | 125/602 [00:19<01:12,  6.54it/s][A

Validation DataLoader 0:  21%|██        | 126/602 [00:19<01:12,  6.59it/s][A

Validation DataLoader 0:  21%|██        | 127/602 [00:19<01:13,  6.50it/s][A

Validation DataLoader 0:  21%|██▏       | 128/602 [00:19<01:12,  6.54it/s][A

Validation DataLoader 0:  21%|██▏       | 129/602 [00:19<01:11,  6.59it/s][A

Validation DataLoader 0:  22%|██▏       | 130/602 [00:20<01:13,  6.46it/s][A

Validation DataLoader 0:  22%|██▏       | 131/602 [00:20<01:12,  6.51it/s][A

Validation DataLoader 0:  22%|██▏       | 132/602 [00:20<01:11,  6.55it/s][A

Validation DataLoader 0:  22%|██▏       | 133/602 [00:20<01:12,  6.46it/s][A

Validation DataLoader 0:  22%|██▏       | 134/602 [00:20<01:11,  6.51it/s][A

Validation DataLoader 0:  22%|██▏       | 135/602 [00:20<01:11,  6.55it/s][A

Validation DataLoader 0:  23%|██▎       | 136/602 [00:20<01:11,  6.48it/s][A

Validation DataLoader 0:  23%|██▎       | 137/602 [00:20<01:11,  6.52it/s][A

Validation DataLoader 0:  23%|██▎       | 138/602 [00:21<01:10,  6.57it/s][A

Validation DataLoader 0:  23%|██▎       | 139/602 [00:21<01:11,  6.49it/s][A

Validation DataLoader 0:  23%|██▎       | 140/602 [00:21<01:10,  6.53it/s][A

Validation DataLoader 0:  23%|██▎       | 141/602 [00:21<01:10,  6.58it/s][A

Validation DataLoader 0:  24%|██▎       | 142/602 [00:21<01:10,  6.48it/s][A

Validation DataLoader 0:  24%|██▍       | 143/602 [00:21<01:10,  6.52it/s][A

Validation DataLoader 0:  24%|██▍       | 144/602 [00:21<01:09,  6.57it/s][A

Validation DataLoader 0:  24%|██▍       | 145/602 [00:22<01:10,  6.50it/s][A

Validation DataLoader 0:  24%|██▍       | 146/602 [00:22<01:09,  6.54it/s][A

Validation DataLoader 0:  24%|██▍       | 147/602 [00:22<01:09,  6.58it/s][A

Validation DataLoader 0:  25%|██▍       | 148/602 [00:22<01:09,  6.49it/s][A

Validation DataLoader 0:  25%|██▍       | 149/602 [00:22<01:09,  6.52it/s][A

Validation DataLoader 0:  25%|██▍       | 150/602 [00:22<01:08,  6.56it/s][A

Validation DataLoader 0:  25%|██▌       | 151/602 [00:23<01:09,  6.48it/s][A

Validation DataLoader 0:  25%|██▌       | 152/602 [00:23<01:09,  6.51it/s][A

Validation DataLoader 0:  25%|██▌       | 153/602 [00:23<01:08,  6.55it/s][A

Validation DataLoader 0:  26%|██▌       | 154/602 [00:23<01:09,  6.49it/s][A

Validation DataLoader 0:  26%|██▌       | 155/602 [00:23<01:08,  6.53it/s][A

Validation DataLoader 0:  26%|██▌       | 156/602 [00:23<01:07,  6.57it/s][A

Validation DataLoader 0:  26%|██▌       | 157/602 [00:24<01:08,  6.51it/s][A

Validation DataLoader 0:  26%|██▌       | 158/602 [00:24<01:07,  6.53it/s][A

Validation DataLoader 0:  26%|██▋       | 159/602 [00:24<01:07,  6.57it/s][A

Validation DataLoader 0:  27%|██▋       | 160/602 [00:24<01:07,  6.52it/s][A

Validation DataLoader 0:  27%|██▋       | 161/602 [00:24<01:07,  6.56it/s][A

Validation DataLoader 0:  27%|██▋       | 162/602 [00:24<01:06,  6.60it/s][A

Validation DataLoader 0:  27%|██▋       | 163/602 [00:24<01:07,  6.54it/s][A

Validation DataLoader 0:  27%|██▋       | 164/602 [00:25<01:06,  6.54it/s][A

Validation DataLoader 0:  27%|██▋       | 165/602 [00:25<01:06,  6.57it/s][A

Validation DataLoader 0:  28%|██▊       | 166/602 [00:25<01:06,  6.54it/s][A

Validation DataLoader 0:  28%|██▊       | 167/602 [00:25<01:06,  6.52it/s][A

Validation DataLoader 0:  28%|██▊       | 168/602 [00:25<01:06,  6.55it/s][A

Validation DataLoader 0:  28%|██▊       | 169/602 [00:25<01:06,  6.55it/s][A

Validation DataLoader 0:  28%|██▊       | 170/602 [00:26<01:06,  6.51it/s][A

Validation DataLoader 0:  28%|██▊       | 171/602 [00:26<01:05,  6.55it/s][A

Validation DataLoader 0:  29%|██▊       | 172/602 [00:26<01:05,  6.52it/s][A

Validation DataLoader 0:  29%|██▊       | 173/602 [00:26<01:05,  6.50it/s][A

Validation DataLoader 0:  29%|██▉       | 174/602 [00:26<01:05,  6.54it/s][A

Validation DataLoader 0:  29%|██▉       | 175/602 [00:26<01:05,  6.50it/s][A

Validation DataLoader 0:  29%|██▉       | 176/602 [00:27<01:05,  6.51it/s][A

Validation DataLoader 0:  29%|██▉       | 177/602 [00:27<01:04,  6.55it/s][A

Validation DataLoader 0:  30%|██▉       | 178/602 [00:27<01:05,  6.51it/s][A

Validation DataLoader 0:  30%|██▉       | 179/602 [00:27<01:04,  6.51it/s][A

Validation DataLoader 0:  30%|██▉       | 180/602 [00:27<01:04,  6.54it/s][A

Validation DataLoader 0:  30%|███       | 181/602 [00:27<01:04,  6.52it/s][A

Validation DataLoader 0:  30%|███       | 182/602 [00:27<01:04,  6.52it/s][A

Validation DataLoader 0:  30%|███       | 183/602 [00:27<01:03,  6.55it/s][A

Validation DataLoader 0:  31%|███       | 184/602 [00:28<01:04,  6.52it/s][A

Validation DataLoader 0:  31%|███       | 185/602 [00:28<01:03,  6.52it/s][A

Validation DataLoader 0:  31%|███       | 186/602 [00:28<01:03,  6.56it/s][A

Validation DataLoader 0:  31%|███       | 187/602 [00:28<01:03,  6.54it/s][A

Validation DataLoader 0:  31%|███       | 188/602 [00:28<01:03,  6.54it/s][A

Validation DataLoader 0:  31%|███▏      | 189/602 [00:28<01:02,  6.57it/s][A

Validation DataLoader 0:  32%|███▏      | 190/602 [00:29<01:03,  6.53it/s][A

Validation DataLoader 0:  32%|███▏      | 191/602 [00:29<01:03,  6.52it/s][A

Validation DataLoader 0:  32%|███▏      | 192/602 [00:29<01:02,  6.55it/s][A

Validation DataLoader 0:  32%|███▏      | 193/602 [00:29<01:02,  6.52it/s][A

Validation DataLoader 0:  32%|███▏      | 194/602 [00:29<01:02,  6.50it/s][A

Validation DataLoader 0:  32%|███▏      | 195/602 [00:29<01:02,  6.53it/s][A

Validation DataLoader 0:  33%|███▎      | 196/602 [00:30<01:02,  6.52it/s][A

Validation DataLoader 0:  33%|███▎      | 197/602 [00:30<01:02,  6.50it/s][A

Validation DataLoader 0:  33%|███▎      | 198/602 [00:30<01:01,  6.53it/s][A

Validation DataLoader 0:  33%|███▎      | 199/602 [00:30<01:01,  6.51it/s][A

Validation DataLoader 0:  33%|███▎      | 200/602 [00:30<01:01,  6.49it/s][A

Validation DataLoader 0:  33%|███▎      | 201/602 [00:30<01:01,  6.52it/s][A

Validation DataLoader 0:  34%|███▎      | 202/602 [00:30<01:01,  6.53it/s][A

Validation DataLoader 0:  34%|███▎      | 203/602 [00:31<01:01,  6.48it/s][A

Validation DataLoader 0:  34%|███▍      | 204/602 [00:31<01:01,  6.51it/s][A

Validation DataLoader 0:  34%|███▍      | 205/602 [00:31<01:00,  6.52it/s][A

Validation DataLoader 0:  34%|███▍      | 206/602 [00:31<01:01,  6.48it/s][A

Validation DataLoader 0:  34%|███▍      | 207/602 [00:31<01:00,  6.51it/s][A

Validation DataLoader 0:  35%|███▍      | 208/602 [00:31<01:00,  6.52it/s][A

Validation DataLoader 0:  35%|███▍      | 209/602 [00:32<01:00,  6.48it/s][A

Validation DataLoader 0:  35%|███▍      | 210/602 [00:32<01:00,  6.51it/s][A

Validation DataLoader 0:  35%|███▌      | 211/602 [00:32<01:00,  6.51it/s][A

Validation DataLoader 0:  35%|███▌      | 212/602 [00:32<01:00,  6.48it/s][A

Validation DataLoader 0:  35%|███▌      | 213/602 [00:32<00:59,  6.51it/s][A

Validation DataLoader 0:  36%|███▌      | 214/602 [00:32<00:59,  6.49it/s][A

Validation DataLoader 0:  36%|███▌      | 215/602 [00:33<00:59,  6.49it/s][A

Validation DataLoader 0:  36%|███▌      | 216/602 [00:33<00:59,  6.52it/s][A

Validation DataLoader 0:  36%|███▌      | 217/602 [00:33<00:59,  6.50it/s][A

Validation DataLoader 0:  36%|███▌      | 218/602 [00:33<00:59,  6.50it/s][A

Validation DataLoader 0:  36%|███▋      | 219/602 [00:33<00:58,  6.52it/s][A

Validation DataLoader 0:  37%|███▋      | 220/602 [00:33<00:58,  6.50it/s][A

Validation DataLoader 0:  37%|███▋      | 221/602 [00:34<00:58,  6.50it/s][A

Validation DataLoader 0:  37%|███▋      | 222/602 [00:34<00:58,  6.52it/s][A

Validation DataLoader 0:  37%|███▋      | 223/602 [00:34<00:58,  6.50it/s][A

Validation DataLoader 0:  37%|███▋      | 224/602 [00:34<00:58,  6.51it/s][A

Validation DataLoader 0:  37%|███▋      | 225/602 [00:34<00:57,  6.53it/s][A

Validation DataLoader 0:  38%|███▊      | 226/602 [00:34<00:57,  6.51it/s][A

Validation DataLoader 0:  38%|███▊      | 227/602 [00:35<00:57,  6.48it/s][A

Validation DataLoader 0:  38%|███▊      | 228/602 [00:35<00:57,  6.50it/s][A

Validation DataLoader 0:  38%|███▊      | 229/602 [00:35<00:57,  6.50it/s][A

Validation DataLoader 0:  38%|███▊      | 230/602 [00:35<00:57,  6.46it/s][A

Validation DataLoader 0:  38%|███▊      | 231/602 [00:35<00:57,  6.48it/s][A

Validation DataLoader 0:  39%|███▊      | 232/602 [00:35<00:57,  6.49it/s][A

Validation DataLoader 0:  39%|███▊      | 233/602 [00:36<00:57,  6.44it/s][A

Validation DataLoader 0:  39%|███▉      | 234/602 [00:36<00:56,  6.47it/s][A

Validation DataLoader 0:  39%|███▉      | 235/602 [00:36<00:56,  6.49it/s][A

Validation DataLoader 0:  39%|███▉      | 236/602 [00:36<00:56,  6.45it/s][A

Validation DataLoader 0:  39%|███▉      | 237/602 [00:36<00:56,  6.47it/s][A

Validation DataLoader 0:  40%|███▉      | 238/602 [00:36<00:56,  6.49it/s][A

Validation DataLoader 0:  40%|███▉      | 239/602 [00:37<00:56,  6.45it/s][A

Validation DataLoader 0:  40%|███▉      | 240/602 [00:37<00:55,  6.48it/s][A

Validation DataLoader 0:  40%|████      | 241/602 [00:37<00:55,  6.50it/s][A

Validation DataLoader 0:  40%|████      | 242/602 [00:37<00:55,  6.46it/s][A

Validation DataLoader 0:  40%|████      | 243/602 [00:37<00:55,  6.49it/s][A

Validation DataLoader 0:  41%|████      | 244/602 [00:37<00:54,  6.51it/s][A

Validation DataLoader 0:  41%|████      | 245/602 [00:37<00:55,  6.45it/s][A

Validation DataLoader 0:  41%|████      | 246/602 [00:38<00:54,  6.47it/s][A

Validation DataLoader 0:  41%|████      | 247/602 [00:38<00:54,  6.50it/s][A

Validation DataLoader 0:  41%|████      | 248/602 [00:38<00:54,  6.44it/s][A

Validation DataLoader 0:  41%|████▏     | 249/602 [00:38<00:54,  6.46it/s][A

Validation DataLoader 0:  42%|████▏     | 250/602 [00:38<00:54,  6.49it/s][A

Validation DataLoader 0:  42%|████▏     | 251/602 [00:39<00:54,  6.43it/s][A

Validation DataLoader 0:  42%|████▏     | 252/602 [00:39<00:54,  6.46it/s][A

Validation DataLoader 0:  42%|████▏     | 253/602 [00:39<00:53,  6.48it/s][A

Validation DataLoader 0:  42%|████▏     | 254/602 [00:39<00:53,  6.45it/s][A

Validation DataLoader 0:  42%|████▏     | 255/602 [00:39<00:53,  6.47it/s][A

Validation DataLoader 0:  43%|████▎     | 256/602 [00:39<00:53,  6.49it/s][A

Validation DataLoader 0:  43%|████▎     | 257/602 [00:39<00:53,  6.45it/s][A

Validation DataLoader 0:  43%|████▎     | 258/602 [00:39<00:53,  6.47it/s][A

Validation DataLoader 0:  43%|████▎     | 259/602 [00:39<00:52,  6.49it/s][A

Validation DataLoader 0:  43%|████▎     | 260/602 [00:40<00:53,  6.45it/s][A

Validation DataLoader 0:  43%|████▎     | 261/602 [00:40<00:52,  6.48it/s][A

Validation DataLoader 0:  44%|████▎     | 262/602 [00:40<00:52,  6.50it/s][A

Validation DataLoader 0:  44%|████▎     | 263/602 [00:40<00:52,  6.46it/s][A

Validation DataLoader 0:  44%|████▍     | 264/602 [00:40<00:52,  6.48it/s][A

Validation DataLoader 0:  44%|████▍     | 265/602 [00:40<00:51,  6.50it/s][A

Validation DataLoader 0:  44%|████▍     | 266/602 [00:41<00:52,  6.45it/s][A

Validation DataLoader 0:  44%|████▍     | 267/602 [00:41<00:51,  6.47it/s][A

Validation DataLoader 0:  45%|████▍     | 268/602 [00:41<00:51,  6.49it/s][A

Validation DataLoader 0:  45%|████▍     | 269/602 [00:41<00:51,  6.44it/s][A

Validation DataLoader 0:  45%|████▍     | 270/602 [00:41<00:51,  6.46it/s][A

Validation DataLoader 0:  45%|████▌     | 271/602 [00:41<00:51,  6.49it/s][A

Validation DataLoader 0:  45%|████▌     | 272/602 [00:42<00:51,  6.44it/s][A

Validation DataLoader 0:  45%|████▌     | 273/602 [00:42<00:50,  6.46it/s][A

Validation DataLoader 0:  46%|████▌     | 274/602 [00:42<00:50,  6.48it/s][A

Validation DataLoader 0:  46%|████▌     | 275/602 [00:42<00:50,  6.44it/s][A

Validation DataLoader 0:  46%|████▌     | 276/602 [00:42<00:50,  6.46it/s][A

Validation DataLoader 0:  46%|████▌     | 277/602 [00:42<00:50,  6.48it/s][A

Validation DataLoader 0:  46%|████▌     | 278/602 [00:43<00:50,  6.45it/s][A

Validation DataLoader 0:  46%|████▋     | 279/602 [00:43<00:49,  6.47it/s][A

Validation DataLoader 0:  47%|████▋     | 280/602 [00:43<00:49,  6.49it/s][A

Validation DataLoader 0:  47%|████▋     | 281/602 [00:43<00:49,  6.45it/s][A

Validation DataLoader 0:  47%|████▋     | 282/602 [00:43<00:49,  6.47it/s][A

Validation DataLoader 0:  47%|████▋     | 283/602 [00:43<00:49,  6.49it/s][A

Validation DataLoader 0:  47%|████▋     | 284/602 [00:43<00:49,  6.46it/s][A

Validation DataLoader 0:  47%|████▋     | 285/602 [00:44<00:48,  6.48it/s][A

Validation DataLoader 0:  48%|████▊     | 286/602 [00:44<00:48,  6.50it/s][A

Validation DataLoader 0:  48%|████▊     | 287/602 [00:44<00:48,  6.46it/s][A

Validation DataLoader 0:  48%|████▊     | 288/602 [00:44<00:48,  6.48it/s][A

Validation DataLoader 0:  48%|████▊     | 289/602 [00:44<00:48,  6.50it/s][A

Validation DataLoader 0:  48%|████▊     | 290/602 [00:44<00:48,  6.46it/s][A

Validation DataLoader 0:  48%|████▊     | 291/602 [00:44<00:47,  6.48it/s][A

Validation DataLoader 0:  49%|████▊     | 292/602 [00:44<00:47,  6.50it/s][A

Validation DataLoader 0:  49%|████▊     | 293/602 [00:45<00:47,  6.46it/s][A

Validation DataLoader 0:  49%|████▉     | 294/602 [00:45<00:47,  6.48it/s][A

Validation DataLoader 0:  49%|████▉     | 295/602 [00:45<00:47,  6.50it/s][A

Validation DataLoader 0:  49%|████▉     | 296/602 [00:45<00:47,  6.47it/s][A

Validation DataLoader 0:  49%|████▉     | 297/602 [00:45<00:47,  6.49it/s][A

Validation DataLoader 0:  50%|████▉     | 298/602 [00:45<00:46,  6.51it/s][A

Validation DataLoader 0:  50%|████▉     | 299/602 [00:46<00:46,  6.46it/s][A

Validation DataLoader 0:  50%|████▉     | 300/602 [00:46<00:46,  6.48it/s][A

Validation DataLoader 0:  50%|█████     | 301/602 [00:46<00:46,  6.50it/s][A

Validation DataLoader 0:  50%|█████     | 302/602 [00:46<00:46,  6.47it/s][A

Validation DataLoader 0:  50%|█████     | 303/602 [00:46<00:46,  6.49it/s][A

Validation DataLoader 0:  50%|█████     | 304/602 [00:46<00:45,  6.51it/s][A

Validation DataLoader 0:  51%|█████     | 305/602 [00:47<00:45,  6.47it/s][A

Validation DataLoader 0:  51%|█████     | 306/602 [00:47<00:45,  6.49it/s][A

Validation DataLoader 0:  51%|█████     | 307/602 [00:47<00:45,  6.51it/s][A

Validation DataLoader 0:  51%|█████     | 308/602 [00:47<00:45,  6.48it/s][A

Validation DataLoader 0:  51%|█████▏    | 309/602 [00:47<00:45,  6.50it/s][A

Validation DataLoader 0:  51%|█████▏    | 310/602 [00:47<00:44,  6.52it/s][A

Validation DataLoader 0:  52%|█████▏    | 311/602 [00:48<00:44,  6.47it/s][A

Validation DataLoader 0:  52%|█████▏    | 312/602 [00:48<00:44,  6.49it/s][A

Validation DataLoader 0:  52%|█████▏    | 313/602 [00:48<00:44,  6.51it/s][A

Validation DataLoader 0:  52%|█████▏    | 314/602 [00:48<00:44,  6.46it/s][A

Validation DataLoader 0:  52%|█████▏    | 315/602 [00:48<00:44,  6.48it/s][A

Validation DataLoader 0:  52%|█████▏    | 316/602 [00:48<00:44,  6.50it/s][A

Validation DataLoader 0:  53%|█████▎    | 317/602 [00:49<00:44,  6.46it/s][A

Validation DataLoader 0:  53%|█████▎    | 318/602 [00:49<00:43,  6.48it/s][A

Validation DataLoader 0:  53%|█████▎    | 319/602 [00:49<00:43,  6.50it/s][A

Validation DataLoader 0:  53%|█████▎    | 320/602 [00:49<00:43,  6.46it/s][A

Validation DataLoader 0:  53%|█████▎    | 321/602 [00:49<00:43,  6.48it/s][A

Validation DataLoader 0:  53%|█████▎    | 322/602 [00:49<00:43,  6.50it/s][A

Validation DataLoader 0:  54%|█████▎    | 323/602 [00:49<00:43,  6.47it/s][A

Validation DataLoader 0:  54%|█████▍    | 324/602 [00:49<00:42,  6.49it/s][A

Validation DataLoader 0:  54%|█████▍    | 325/602 [00:49<00:42,  6.51it/s][A

Validation DataLoader 0:  54%|█████▍    | 326/602 [00:50<00:42,  6.48it/s][A

Validation DataLoader 0:  54%|█████▍    | 327/602 [00:50<00:42,  6.50it/s][A

Validation DataLoader 0:  54%|█████▍    | 328/602 [00:50<00:42,  6.52it/s][A

Validation DataLoader 0:  55%|█████▍    | 329/602 [00:50<00:42,  6.48it/s][A

Validation DataLoader 0:  55%|█████▍    | 330/602 [00:50<00:41,  6.50it/s][A

Validation DataLoader 0:  55%|█████▍    | 331/602 [00:50<00:41,  6.52it/s][A

Validation DataLoader 0:  55%|█████▌    | 332/602 [00:51<00:41,  6.50it/s][A

Validation DataLoader 0:  55%|█████▌    | 333/602 [00:51<00:41,  6.51it/s][A

Validation DataLoader 0:  55%|█████▌    | 334/602 [00:51<00:41,  6.53it/s][A

Validation DataLoader 0:  56%|█████▌    | 335/602 [00:51<00:41,  6.50it/s][A

Validation DataLoader 0:  56%|█████▌    | 336/602 [00:51<00:40,  6.51it/s][A

Validation DataLoader 0:  56%|█████▌    | 337/602 [00:51<00:40,  6.53it/s][A

Validation DataLoader 0:  56%|█████▌    | 338/602 [00:51<00:40,  6.50it/s][A

Validation DataLoader 0:  56%|█████▋    | 339/602 [00:52<00:40,  6.51it/s][A

Validation DataLoader 0:  56%|█████▋    | 340/602 [00:52<00:40,  6.53it/s][A

Validation DataLoader 0:  57%|█████▋    | 341/602 [00:52<00:40,  6.51it/s][A

Validation DataLoader 0:  57%|█████▋    | 342/602 [00:52<00:39,  6.51it/s][A

Validation DataLoader 0:  57%|█████▋    | 343/602 [00:52<00:39,  6.53it/s][A

Validation DataLoader 0:  57%|█████▋    | 344/602 [00:52<00:39,  6.52it/s][A

Validation DataLoader 0:  57%|█████▋    | 345/602 [00:52<00:39,  6.52it/s][A

Validation DataLoader 0:  57%|█████▋    | 346/602 [00:52<00:39,  6.54it/s][A

Validation DataLoader 0:  58%|█████▊    | 347/602 [00:53<00:39,  6.52it/s][A

Validation DataLoader 0:  58%|█████▊    | 348/602 [00:53<00:38,  6.52it/s][A

Validation DataLoader 0:  58%|█████▊    | 349/602 [00:53<00:38,  6.54it/s][A

Validation DataLoader 0:  58%|█████▊    | 350/602 [00:53<00:38,  6.52it/s][A

Validation DataLoader 0:  58%|█████▊    | 351/602 [00:53<00:38,  6.52it/s][A

Validation DataLoader 0:  58%|█████▊    | 352/602 [00:53<00:38,  6.54it/s][A

Validation DataLoader 0:  59%|█████▊    | 353/602 [00:54<00:38,  6.52it/s][A

Validation DataLoader 0:  59%|█████▉    | 354/602 [00:54<00:37,  6.53it/s][A

Validation DataLoader 0:  59%|█████▉    | 355/602 [00:54<00:37,  6.55it/s][A

Validation DataLoader 0:  59%|█████▉    | 356/602 [00:54<00:37,  6.53it/s][A

Validation DataLoader 0:  59%|█████▉    | 357/602 [00:54<00:37,  6.54it/s][A

Validation DataLoader 0:  59%|█████▉    | 358/602 [00:54<00:37,  6.56it/s][A

Validation DataLoader 0:  60%|█████▉    | 359/602 [00:54<00:37,  6.53it/s][A

Validation DataLoader 0:  60%|█████▉    | 360/602 [00:55<00:37,  6.54it/s][A

Validation DataLoader 0:  60%|█████▉    | 361/602 [00:55<00:36,  6.55it/s][A

Validation DataLoader 0:  60%|██████    | 362/602 [00:55<00:36,  6.54it/s][A

Validation DataLoader 0:  60%|██████    | 363/602 [00:55<00:36,  6.54it/s][A

Validation DataLoader 0:  60%|██████    | 364/602 [00:55<00:36,  6.56it/s][A

Validation DataLoader 0:  61%|██████    | 365/602 [00:55<00:36,  6.54it/s][A

Validation DataLoader 0:  61%|██████    | 366/602 [00:55<00:36,  6.54it/s][A

Validation DataLoader 0:  61%|██████    | 367/602 [00:55<00:35,  6.56it/s][A

Validation DataLoader 0:  61%|██████    | 368/602 [00:56<00:35,  6.54it/s][A

Validation DataLoader 0:  61%|██████▏   | 369/602 [00:56<00:35,  6.55it/s][A

Validation DataLoader 0:  61%|██████▏   | 370/602 [00:56<00:35,  6.56it/s][A

Validation DataLoader 0:  62%|██████▏   | 371/602 [00:56<00:35,  6.55it/s][A

Validation DataLoader 0:  62%|██████▏   | 372/602 [00:56<00:35,  6.55it/s][A

Validation DataLoader 0:  62%|██████▏   | 373/602 [00:56<00:34,  6.57it/s][A

Validation DataLoader 0:  62%|██████▏   | 374/602 [00:57<00:34,  6.54it/s][A

Validation DataLoader 0:  62%|██████▏   | 375/602 [00:57<00:34,  6.56it/s][A

Validation DataLoader 0:  62%|██████▏   | 376/602 [00:57<00:34,  6.57it/s][A

Validation DataLoader 0:  63%|██████▎   | 377/602 [00:57<00:34,  6.54it/s][A

Validation DataLoader 0:  63%|██████▎   | 378/602 [00:57<00:34,  6.55it/s][A

Validation DataLoader 0:  63%|██████▎   | 379/602 [00:57<00:33,  6.56it/s][A

Validation DataLoader 0:  63%|██████▎   | 380/602 [00:58<00:33,  6.55it/s][A

Validation DataLoader 0:  63%|██████▎   | 381/602 [00:58<00:33,  6.56it/s][A

Validation DataLoader 0:  63%|██████▎   | 382/602 [00:58<00:33,  6.57it/s][A

Validation DataLoader 0:  64%|██████▎   | 383/602 [00:58<00:33,  6.55it/s][A

Validation DataLoader 0:  64%|██████▍   | 384/602 [00:58<00:33,  6.56it/s][A

Validation DataLoader 0:  64%|██████▍   | 385/602 [00:58<00:33,  6.58it/s][A

Validation DataLoader 0:  64%|██████▍   | 386/602 [00:58<00:32,  6.55it/s][A

Validation DataLoader 0:  64%|██████▍   | 387/602 [00:59<00:32,  6.56it/s][A

Validation DataLoader 0:  64%|██████▍   | 388/602 [00:59<00:32,  6.57it/s][A

Validation DataLoader 0:  65%|██████▍   | 389/602 [00:59<00:32,  6.55it/s][A

Validation DataLoader 0:  65%|██████▍   | 390/602 [00:59<00:32,  6.56it/s][A

Validation DataLoader 0:  65%|██████▍   | 391/602 [00:59<00:32,  6.57it/s][A

Validation DataLoader 0:  65%|██████▌   | 392/602 [00:59<00:32,  6.55it/s][A

Validation DataLoader 0:  65%|██████▌   | 393/602 [00:59<00:31,  6.56it/s][A

Validation DataLoader 0:  65%|██████▌   | 394/602 [01:00<00:31,  6.57it/s][A

Validation DataLoader 0:  66%|██████▌   | 395/602 [01:00<00:31,  6.55it/s][A

Validation DataLoader 0:  66%|██████▌   | 396/602 [01:00<00:31,  6.56it/s][A

Validation DataLoader 0:  66%|██████▌   | 397/602 [01:00<00:31,  6.57it/s][A

Validation DataLoader 0:  66%|██████▌   | 398/602 [01:00<00:31,  6.56it/s][A

Validation DataLoader 0:  66%|██████▋   | 399/602 [01:00<00:30,  6.56it/s][A

Validation DataLoader 0:  66%|██████▋   | 400/602 [01:00<00:30,  6.56it/s][A

Validation DataLoader 0:  67%|██████▋   | 401/602 [01:01<00:30,  6.56it/s][A

Validation DataLoader 0:  67%|██████▋   | 402/602 [01:01<00:30,  6.56it/s][A

Validation DataLoader 0:  67%|██████▋   | 403/602 [01:01<00:30,  6.57it/s][A

Validation DataLoader 0:  67%|██████▋   | 404/602 [01:01<00:30,  6.56it/s][A

Validation DataLoader 0:  67%|██████▋   | 405/602 [01:01<00:30,  6.56it/s][A

Validation DataLoader 0:  67%|██████▋   | 406/602 [01:01<00:29,  6.57it/s][A

Validation DataLoader 0:  68%|██████▊   | 407/602 [01:02<00:29,  6.55it/s][A

Validation DataLoader 0:  68%|██████▊   | 408/602 [01:02<00:29,  6.55it/s][A

Validation DataLoader 0:  68%|██████▊   | 409/602 [01:02<00:29,  6.57it/s][A

Validation DataLoader 0:  68%|██████▊   | 410/602 [01:02<00:29,  6.55it/s][A

Validation DataLoader 0:  68%|██████▊   | 411/602 [01:02<00:29,  6.56it/s][A

Validation DataLoader 0:  68%|██████▊   | 412/602 [01:02<00:28,  6.57it/s][A

Validation DataLoader 0:  69%|██████▊   | 413/602 [01:03<00:28,  6.56it/s][A

Validation DataLoader 0:  69%|██████▉   | 414/602 [01:03<00:28,  6.56it/s][A

Validation DataLoader 0:  69%|██████▉   | 415/602 [01:03<00:28,  6.57it/s][A

Validation DataLoader 0:  69%|██████▉   | 416/602 [01:03<00:28,  6.56it/s][A

Validation DataLoader 0:  69%|██████▉   | 417/602 [01:03<00:28,  6.56it/s][A

Validation DataLoader 0:  69%|██████▉   | 418/602 [01:03<00:28,  6.57it/s][A

Validation DataLoader 0:  70%|██████▉   | 419/602 [01:03<00:27,  6.57it/s][A

Validation DataLoader 0:  70%|██████▉   | 420/602 [01:04<00:27,  6.56it/s][A

Validation DataLoader 0:  70%|██████▉   | 421/602 [01:04<00:27,  6.57it/s][A

Validation DataLoader 0:  70%|███████   | 422/602 [01:04<00:27,  6.57it/s][A

Validation DataLoader 0:  70%|███████   | 423/602 [01:04<00:27,  6.56it/s][A

Validation DataLoader 0:  70%|███████   | 424/602 [01:04<00:27,  6.57it/s][A

Validation DataLoader 0:  71%|███████   | 425/602 [01:04<00:26,  6.57it/s][A

Validation DataLoader 0:  71%|███████   | 426/602 [01:04<00:26,  6.56it/s][A

Validation DataLoader 0:  71%|███████   | 427/602 [01:04<00:26,  6.57it/s][A

Validation DataLoader 0:  71%|███████   | 428/602 [01:05<00:26,  6.57it/s][A

Validation DataLoader 0:  71%|███████▏  | 429/602 [01:05<00:26,  6.56it/s][A

Validation DataLoader 0:  71%|███████▏  | 430/602 [01:05<00:26,  6.57it/s][A

Validation DataLoader 0:  72%|███████▏  | 431/602 [01:05<00:26,  6.57it/s][A

Validation DataLoader 0:  72%|███████▏  | 432/602 [01:05<00:25,  6.56it/s][A

Validation DataLoader 0:  72%|███████▏  | 433/602 [01:05<00:25,  6.57it/s][A

Validation DataLoader 0:  72%|███████▏  | 434/602 [01:05<00:25,  6.58it/s][A

Validation DataLoader 0:  72%|███████▏  | 435/602 [01:06<00:25,  6.56it/s][A

Validation DataLoader 0:  72%|███████▏  | 436/602 [01:06<00:25,  6.57it/s][A

Validation DataLoader 0:  73%|███████▎  | 437/602 [01:06<00:25,  6.58it/s][A

Validation DataLoader 0:  73%|███████▎  | 438/602 [01:06<00:24,  6.56it/s][A

Validation DataLoader 0:  73%|███████▎  | 439/602 [01:06<00:24,  6.58it/s][A

Validation DataLoader 0:  73%|███████▎  | 440/602 [01:06<00:24,  6.58it/s][A

Validation DataLoader 0:  73%|███████▎  | 441/602 [01:07<00:24,  6.55it/s][A

Validation DataLoader 0:  73%|███████▎  | 442/602 [01:07<00:24,  6.56it/s][A

Validation DataLoader 0:  74%|███████▎  | 443/602 [01:07<00:24,  6.58it/s][A

Validation DataLoader 0:  74%|███████▍  | 444/602 [01:07<00:24,  6.55it/s][A

Validation DataLoader 0:  74%|███████▍  | 445/602 [01:07<00:23,  6.56it/s][A

Validation DataLoader 0:  74%|███████▍  | 446/602 [01:07<00:23,  6.58it/s][A

Validation DataLoader 0:  74%|███████▍  | 447/602 [01:08<00:23,  6.54it/s][A

Validation DataLoader 0:  74%|███████▍  | 448/602 [01:08<00:23,  6.56it/s][A

Validation DataLoader 0:  75%|███████▍  | 449/602 [01:08<00:23,  6.57it/s][A

Validation DataLoader 0:  75%|███████▍  | 450/602 [01:08<00:23,  6.54it/s][A

Validation DataLoader 0:  75%|███████▍  | 451/602 [01:08<00:23,  6.56it/s][A

Validation DataLoader 0:  75%|███████▌  | 452/602 [01:08<00:22,  6.57it/s][A

Validation DataLoader 0:  75%|███████▌  | 453/602 [01:09<00:22,  6.54it/s][A

Validation DataLoader 0:  75%|███████▌  | 454/602 [01:09<00:22,  6.56it/s][A

Validation DataLoader 0:  76%|███████▌  | 455/602 [01:09<00:22,  6.57it/s][A

Validation DataLoader 0:  76%|███████▌  | 456/602 [01:09<00:22,  6.55it/s][A

Validation DataLoader 0:  76%|███████▌  | 457/602 [01:09<00:22,  6.56it/s][A

Validation DataLoader 0:  76%|███████▌  | 458/602 [01:09<00:21,  6.56it/s][A

Validation DataLoader 0:  76%|███████▌  | 459/602 [01:10<00:21,  6.55it/s][A

Validation DataLoader 0:  76%|███████▋  | 460/602 [01:10<00:21,  6.56it/s][A

Validation DataLoader 0:  77%|███████▋  | 461/602 [01:10<00:21,  6.56it/s][A

Validation DataLoader 0:  77%|███████▋  | 462/602 [01:10<00:21,  6.55it/s][A

Validation DataLoader 0:  77%|███████▋  | 463/602 [01:10<00:21,  6.56it/s][A

Validation DataLoader 0:  77%|███████▋  | 464/602 [01:10<00:21,  6.56it/s][A

Validation DataLoader 0:  77%|███████▋  | 465/602 [01:11<00:20,  6.54it/s][A

Validation DataLoader 0:  77%|███████▋  | 466/602 [01:11<00:20,  6.55it/s][A

Validation DataLoader 0:  78%|███████▊  | 467/602 [01:11<00:20,  6.56it/s][A

Validation DataLoader 0:  78%|███████▊  | 468/602 [01:11<00:20,  6.54it/s][A

Validation DataLoader 0:  78%|███████▊  | 469/602 [01:11<00:20,  6.55it/s][A

Validation DataLoader 0:  78%|███████▊  | 470/602 [01:11<00:20,  6.56it/s][A

Validation DataLoader 0:  78%|███████▊  | 471/602 [01:12<00:20,  6.53it/s][A

Validation DataLoader 0:  78%|███████▊  | 472/602 [01:12<00:19,  6.55it/s][A

Validation DataLoader 0:  79%|███████▊  | 473/602 [01:12<00:19,  6.56it/s][A

Validation DataLoader 0:  79%|███████▊  | 474/602 [01:12<00:19,  6.52it/s][A

Validation DataLoader 0:  79%|███████▉  | 475/602 [01:12<00:19,  6.54it/s][A

Validation DataLoader 0:  79%|███████▉  | 476/602 [01:12<00:19,  6.55it/s][A

Validation DataLoader 0:  79%|███████▉  | 477/602 [01:13<00:19,  6.52it/s][A

Validation DataLoader 0:  79%|███████▉  | 478/602 [01:13<00:18,  6.53it/s][A

Validation DataLoader 0:  80%|███████▉  | 479/602 [01:13<00:18,  6.55it/s][A

Validation DataLoader 0:  80%|███████▉  | 480/602 [01:13<00:18,  6.52it/s][A

Validation DataLoader 0:  80%|███████▉  | 481/602 [01:13<00:18,  6.54it/s][A

Validation DataLoader 0:  80%|████████  | 482/602 [01:13<00:18,  6.55it/s][A

Validation DataLoader 0:  80%|████████  | 483/602 [01:14<00:18,  6.52it/s][A

Validation DataLoader 0:  80%|████████  | 484/602 [01:14<00:18,  6.53it/s][A

Validation DataLoader 0:  81%|████████  | 485/602 [01:14<00:17,  6.54it/s][A

Validation DataLoader 0:  81%|████████  | 486/602 [01:14<00:17,  6.51it/s][A

Validation DataLoader 0:  81%|████████  | 487/602 [01:14<00:17,  6.53it/s][A

Validation DataLoader 0:  81%|████████  | 488/602 [01:14<00:17,  6.54it/s][A

Validation DataLoader 0:  81%|████████  | 489/602 [01:14<00:17,  6.52it/s][A

Validation DataLoader 0:  81%|████████▏ | 490/602 [01:14<00:17,  6.54it/s][A

Validation DataLoader 0:  82%|████████▏ | 491/602 [01:14<00:16,  6.55it/s][A

Validation DataLoader 0:  82%|████████▏ | 492/602 [01:15<00:16,  6.53it/s][A

Validation DataLoader 0:  82%|████████▏ | 493/602 [01:15<00:16,  6.54it/s][A

Validation DataLoader 0:  82%|████████▏ | 494/602 [01:15<00:16,  6.55it/s][A

Validation DataLoader 0:  82%|████████▏ | 495/602 [01:15<00:16,  6.53it/s][A

Validation DataLoader 0:  82%|████████▏ | 496/602 [01:15<00:16,  6.54it/s][A

Validation DataLoader 0:  83%|████████▎ | 497/602 [01:15<00:16,  6.55it/s][A

Validation DataLoader 0:  83%|████████▎ | 498/602 [01:16<00:15,  6.52it/s][A

Validation DataLoader 0:  83%|████████▎ | 499/602 [01:16<00:15,  6.54it/s][A

Validation DataLoader 0:  83%|████████▎ | 500/602 [01:16<00:15,  6.55it/s][A

Validation DataLoader 0:  83%|████████▎ | 501/602 [01:16<00:15,  6.52it/s][A

Validation DataLoader 0:  83%|████████▎ | 502/602 [01:16<00:15,  6.53it/s][A

Validation DataLoader 0:  84%|████████▎ | 503/602 [01:16<00:15,  6.55it/s][A

Validation DataLoader 0:  84%|████████▎ | 504/602 [01:17<00:15,  6.52it/s][A

Validation DataLoader 0:  84%|████████▍ | 505/602 [01:17<00:14,  6.53it/s][A

Validation DataLoader 0:  84%|████████▍ | 506/602 [01:17<00:14,  6.54it/s][A

Validation DataLoader 0:  84%|████████▍ | 507/602 [01:17<00:14,  6.53it/s][A

Validation DataLoader 0:  84%|████████▍ | 508/602 [01:17<00:14,  6.54it/s][A

Validation DataLoader 0:  85%|████████▍ | 509/602 [01:17<00:14,  6.55it/s][A

Validation DataLoader 0:  85%|████████▍ | 510/602 [01:18<00:14,  6.53it/s][A

Validation DataLoader 0:  85%|████████▍ | 511/602 [01:18<00:13,  6.54it/s][A

Validation DataLoader 0:  85%|████████▌ | 512/602 [01:18<00:13,  6.56it/s][A

Validation DataLoader 0:  85%|████████▌ | 513/602 [01:18<00:13,  6.53it/s][A

Validation DataLoader 0:  85%|████████▌ | 514/602 [01:18<00:13,  6.54it/s][A

Validation DataLoader 0:  86%|████████▌ | 515/602 [01:18<00:13,  6.56it/s][A

Validation DataLoader 0:  86%|████████▌ | 516/602 [01:18<00:13,  6.53it/s][A

Validation DataLoader 0:  86%|████████▌ | 517/602 [01:18<00:12,  6.55it/s][A

Validation DataLoader 0:  86%|████████▌ | 518/602 [01:18<00:12,  6.56it/s][A

Validation DataLoader 0:  86%|████████▌ | 519/602 [01:19<00:12,  6.54it/s][A

Validation DataLoader 0:  86%|████████▋ | 520/602 [01:19<00:12,  6.55it/s][A

Validation DataLoader 0:  87%|████████▋ | 521/602 [01:19<00:12,  6.56it/s][A

Validation DataLoader 0:  87%|████████▋ | 522/602 [01:19<00:12,  6.53it/s][A

Validation DataLoader 0:  87%|████████▋ | 523/602 [01:19<00:12,  6.54it/s][A

Validation DataLoader 0:  87%|████████▋ | 524/602 [01:19<00:11,  6.55it/s][A

Validation DataLoader 0:  87%|████████▋ | 525/602 [01:20<00:11,  6.53it/s][A

Validation DataLoader 0:  87%|████████▋ | 526/602 [01:20<00:11,  6.54it/s][A

Validation DataLoader 0:  88%|████████▊ | 527/602 [01:20<00:11,  6.55it/s][A

Validation DataLoader 0:  88%|████████▊ | 528/602 [01:20<00:11,  6.52it/s][A

Validation DataLoader 0:  88%|████████▊ | 529/602 [01:20<00:11,  6.53it/s][A

Validation DataLoader 0:  88%|████████▊ | 530/602 [01:21<00:11,  6.54it/s][A

Validation DataLoader 0:  88%|████████▊ | 531/602 [01:21<00:10,  6.52it/s][A

Validation DataLoader 0:  88%|████████▊ | 532/602 [01:21<00:10,  6.53it/s][A

Validation DataLoader 0:  89%|████████▊ | 533/602 [01:21<00:10,  6.54it/s][A

Validation DataLoader 0:  89%|████████▊ | 534/602 [01:21<00:10,  6.51it/s][A

Validation DataLoader 0:  89%|████████▉ | 535/602 [01:21<00:10,  6.53it/s][A

Validation DataLoader 0:  89%|████████▉ | 536/602 [01:21<00:10,  6.54it/s][A

Validation DataLoader 0:  89%|████████▉ | 537/602 [01:22<00:09,  6.52it/s][A

Validation DataLoader 0:  89%|████████▉ | 538/602 [01:22<00:09,  6.53it/s][A

Validation DataLoader 0:  90%|████████▉ | 539/602 [01:22<00:09,  6.54it/s][A

Validation DataLoader 0:  90%|████████▉ | 540/602 [01:22<00:09,  6.51it/s][A

Validation DataLoader 0:  90%|████████▉ | 541/602 [01:22<00:09,  6.52it/s][A

Validation DataLoader 0:  90%|█████████ | 542/602 [01:22<00:09,  6.54it/s][A

Validation DataLoader 0:  90%|█████████ | 543/602 [01:23<00:09,  6.52it/s][A

Validation DataLoader 0:  90%|█████████ | 544/602 [01:23<00:08,  6.53it/s][A

Validation DataLoader 0:  91%|█████████ | 545/602 [01:23<00:08,  6.54it/s][A

Validation DataLoader 0:  91%|█████████ | 546/602 [01:23<00:08,  6.53it/s][A

Validation DataLoader 0:  91%|█████████ | 547/602 [01:23<00:08,  6.54it/s][A

Validation DataLoader 0:  91%|█████████ | 548/602 [01:23<00:08,  6.55it/s][A

Validation DataLoader 0:  91%|█████████ | 549/602 [01:24<00:08,  6.53it/s][A

Validation DataLoader 0:  91%|█████████▏| 550/602 [01:24<00:07,  6.54it/s][A

Validation DataLoader 0:  92%|█████████▏| 551/602 [01:24<00:07,  6.56it/s][A

Validation DataLoader 0:  92%|█████████▏| 552/602 [01:24<00:07,  6.53it/s][A

Validation DataLoader 0:  92%|█████████▏| 553/602 [01:24<00:07,  6.54it/s][A

Validation DataLoader 0:  92%|█████████▏| 554/602 [01:24<00:07,  6.55it/s][A

Validation DataLoader 0:  92%|█████████▏| 555/602 [01:24<00:07,  6.53it/s][A

Validation DataLoader 0:  92%|█████████▏| 556/602 [01:25<00:07,  6.54it/s][A

Validation DataLoader 0:  93%|█████████▎| 557/602 [01:25<00:06,  6.55it/s][A

Validation DataLoader 0:  93%|█████████▎| 558/602 [01:25<00:06,  6.53it/s][A

Validation DataLoader 0:  93%|█████████▎| 559/602 [01:25<00:06,  6.54it/s][A

Validation DataLoader 0:  93%|█████████▎| 560/602 [01:25<00:06,  6.55it/s][A

Validation DataLoader 0:  93%|█████████▎| 561/602 [01:25<00:06,  6.52it/s][A

Validation DataLoader 0:  93%|█████████▎| 562/602 [01:26<00:06,  6.53it/s][A

Validation DataLoader 0:  94%|█████████▎| 563/602 [01:26<00:05,  6.54it/s][A

Validation DataLoader 0:  94%|█████████▎| 564/602 [01:26<00:05,  6.52it/s][A

Validation DataLoader 0:  94%|█████████▍| 565/602 [01:26<00:05,  6.54it/s][A

Validation DataLoader 0:  94%|█████████▍| 566/602 [01:26<00:05,  6.55it/s][A

Validation DataLoader 0:  94%|█████████▍| 567/602 [01:26<00:05,  6.52it/s][A

Validation DataLoader 0:  94%|█████████▍| 568/602 [01:26<00:05,  6.53it/s][A

Validation DataLoader 0:  95%|█████████▍| 569/602 [01:26<00:05,  6.54it/s][A

Validation DataLoader 0:  95%|█████████▍| 570/602 [01:27<00:04,  6.52it/s][A

Validation DataLoader 0:  95%|█████████▍| 571/602 [01:27<00:04,  6.53it/s][A

Validation DataLoader 0:  95%|█████████▌| 572/602 [01:27<00:04,  6.54it/s][A

Validation DataLoader 0:  95%|█████████▌| 573/602 [01:27<00:04,  6.52it/s][A

Validation DataLoader 0:  95%|█████████▌| 574/602 [01:27<00:04,  6.53it/s][A

Validation DataLoader 0:  96%|█████████▌| 575/602 [01:27<00:04,  6.54it/s][A

Validation DataLoader 0:  96%|█████████▌| 576/602 [01:28<00:03,  6.52it/s][A

Validation DataLoader 0:  96%|█████████▌| 577/602 [01:28<00:03,  6.53it/s][A

Validation DataLoader 0:  96%|█████████▌| 578/602 [01:28<00:03,  6.54it/s][A

Validation DataLoader 0:  96%|█████████▌| 579/602 [01:28<00:03,  6.52it/s][A

Validation DataLoader 0:  96%|█████████▋| 580/602 [01:28<00:03,  6.53it/s][A

Validation DataLoader 0:  97%|█████████▋| 581/602 [01:28<00:03,  6.54it/s][A

Validation DataLoader 0:  97%|█████████▋| 582/602 [01:29<00:03,  6.52it/s][A

Validation DataLoader 0:  97%|█████████▋| 583/602 [01:29<00:02,  6.53it/s][A

Validation DataLoader 0:  97%|█████████▋| 584/602 [01:29<00:02,  6.54it/s][A

Validation DataLoader 0:  97%|█████████▋| 585/602 [01:29<00:02,  6.52it/s][A

Validation DataLoader 0:  97%|█████████▋| 586/602 [01:29<00:02,  6.53it/s][A

Validation DataLoader 0:  98%|█████████▊| 587/602 [01:29<00:02,  6.54it/s][A

Validation DataLoader 0:  98%|█████████▊| 588/602 [01:30<00:02,  6.52it/s][A

Validation DataLoader 0:  98%|█████████▊| 589/602 [01:30<00:01,  6.53it/s][A

Validation DataLoader 0:  98%|█████████▊| 590/602 [01:30<00:01,  6.54it/s][A

Validation DataLoader 0:  98%|█████████▊| 591/602 [01:30<00:01,  6.52it/s][A

Validation DataLoader 0:  98%|█████████▊| 592/602 [01:30<00:01,  6.53it/s][A

Validation DataLoader 0:  99%|█████████▊| 593/602 [01:30<00:01,  6.54it/s][A

Validation DataLoader 0:  99%|█████████▊| 594/602 [01:31<00:01,  6.52it/s][A

Validation DataLoader 0:  99%|█████████▉| 595/602 [01:31<00:01,  6.53it/s][A

Validation DataLoader 0:  99%|█████████▉| 596/602 [01:31<00:00,  6.54it/s][A

Validation DataLoader 0:  99%|█████████▉| 597/602 [01:31<00:00,  6.51it/s][A

Validation DataLoader 0:  99%|█████████▉| 598/602 [01:31<00:00,  6.52it/s][A

Validation DataLoader 0: 100%|█████████▉| 599/602 [01:31<00:00,  6.53it/s][A

Validation DataLoader 0: 100%|█████████▉| 600/602 [01:32<00:00,  6.52it/s][A

Validation DataLoader 0: 100%|█████████▉| 601/602 [01:32<00:00,  6.53it/s][A

Validation DataLoader 0: 100%|██████████| 602/602 [01:32<00:00,  6.54it/s][A

                                                                          [A
Epoch 3: 100%|██████████| 3295/3295 [36:23<00:00,  1.51it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.194, R@1 (train)=0.958]
Epoch 3: 100%|██████████| 3295/3295 [36:23<00:00,  1.51it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]Epoch 3, global step 13180: 'R@1 (val)' reached 0.40743 (best 0.40743), saving model to '/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=03-R@1 (val)=0.407.ckpt' as top 1

Epoch 3:   0%|          | 0/3295 [00:00<?, ?it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]           
Epoch 4:   0%|          | 0/2820 [00:00<?, ?it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 1/2820 [00:02<1:46:10,  0.44it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 1/2820 [00:02<1:46:11,  0.44it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 2/2820 [00:02<53:32,  0.88it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]  
Epoch 4:   0%|          | 2/2820 [00:02<53:32,  0.88it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 3/2820 [00:02<36:07,  1.30it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 3/2820 [00:02<36:08,  1.30it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 4/2820 [00:04<48:42,  0.96it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 4/2820 [00:04<48:42,  0.96it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 5/2820 [00:04<39:14,  1.20it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 5/2820 [00:04<39:14,  1.20it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 6/2820 [00:04<32:53,  1.43it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 6/2820 [00:04<32:54,  1.43it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 7/2820 [00:06<41:09,  1.14it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 7/2820 [00:06<41:10,  1.14it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 8/2820 [00:06<36:11,  1.30it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 8/2820 [00:06<36:11,  1.30it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 9/2820 [00:06<32:18,  1.45it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 9/2820 [00:06<32:18,  1.45it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 10/2820 [00:08<37:31,  1.25it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 10/2820 [00:08<37:31,  1.25it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 11/2820 [00:08<34:13,  1.37it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 11/2820 [00:08<34:13,  1.37it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 12/2820 [00:08<31:27,  1.49it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 12/2820 [00:08<31:28,  1.49it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 13/2820 [00:09<35:16,  1.33it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 13/2820 [00:09<35:16,  1.33it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 14/2820 [00:09<32:49,  1.42it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   0%|          | 14/2820 [00:09<32:50,  1.42it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 15/2820 [00:09<30:42,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 15/2820 [00:09<30:42,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 16/2820 [00:11<34:10,  1.37it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 16/2820 [00:11<34:10,  1.37it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 17/2820 [00:11<32:13,  1.45it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 17/2820 [00:11<32:14,  1.45it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 18/2820 [00:11<30:29,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 18/2820 [00:11<30:30,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 19/2820 [00:13<33:18,  1.40it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 19/2820 [00:13<33:18,  1.40it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 20/2820 [00:13<31:42,  1.47it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 20/2820 [00:13<31:42,  1.47it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 21/2820 [00:13<30:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 21/2820 [00:13<30:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 22/2820 [00:15<32:30,  1.43it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 22/2820 [00:15<32:31,  1.43it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 23/2820 [00:15<31:23,  1.48it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 23/2820 [00:15<31:23,  1.48it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 24/2820 [00:15<30:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 24/2820 [00:15<30:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 25/2820 [00:17<32:02,  1.45it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 25/2820 [00:17<32:02,  1.45it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 26/2820 [00:17<31:30,  1.48it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 26/2820 [00:17<31:30,  1.48it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 27/2820 [00:17<30:22,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 27/2820 [00:17<30:22,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 28/2820 [00:19<31:58,  1.46it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 28/2820 [00:19<31:58,  1.46it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 29/2820 [00:19<31:13,  1.49it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 29/2820 [00:19<31:13,  1.49it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 30/2820 [00:19<30:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 30/2820 [00:19<30:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 31/2820 [00:21<31:31,  1.47it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 31/2820 [00:21<31:31,  1.47it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 32/2820 [00:21<30:54,  1.50it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 32/2820 [00:21<30:54,  1.50it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 33/2820 [00:21<30:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 33/2820 [00:21<30:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 34/2820 [00:22<31:20,  1.48it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 34/2820 [00:22<31:20,  1.48it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 35/2820 [00:23<30:46,  1.51it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|          | 35/2820 [00:23<30:46,  1.51it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 36/2820 [00:23<29:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 36/2820 [00:23<29:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 37/2820 [00:24<31:11,  1.49it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 37/2820 [00:24<31:11,  1.49it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 38/2820 [00:25<30:38,  1.51it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 38/2820 [00:25<30:38,  1.51it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 39/2820 [00:25<29:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 39/2820 [00:25<29:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 40/2820 [00:26<30:52,  1.50it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 40/2820 [00:26<30:52,  1.50it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 41/2820 [00:27<30:31,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 41/2820 [00:27<30:31,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 42/2820 [00:27<29:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   1%|▏         | 42/2820 [00:27<29:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 43/2820 [00:28<30:32,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 43/2820 [00:28<30:32,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 44/2820 [00:28<30:27,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 44/2820 [00:28<30:27,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 45/2820 [00:28<29:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 45/2820 [00:28<29:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 46/2820 [00:30<30:23,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 46/2820 [00:30<30:23,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 47/2820 [00:30<30:17,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 47/2820 [00:30<30:17,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 48/2820 [00:30<29:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 48/2820 [00:30<29:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 49/2820 [00:32<30:21,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 49/2820 [00:32<30:21,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 50/2820 [00:32<30:05,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 50/2820 [00:32<30:05,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 51/2820 [00:32<29:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 51/2820 [00:32<29:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 52/2820 [00:34<30:17,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 52/2820 [00:34<30:17,  1.52it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 53/2820 [00:34<29:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 53/2820 [00:34<29:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 54/2820 [00:34<29:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 54/2820 [00:34<29:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 55/2820 [00:35<30:06,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 55/2820 [00:35<30:06,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 56/2820 [00:36<29:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 56/2820 [00:36<29:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 57/2820 [00:36<29:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 57/2820 [00:36<29:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 58/2820 [00:37<30:06,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 58/2820 [00:37<30:06,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 59/2820 [00:38<29:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 59/2820 [00:38<29:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 60/2820 [00:38<29:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 60/2820 [00:38<29:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 61/2820 [00:39<29:59,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 61/2820 [00:39<29:59,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 62/2820 [00:39<29:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 62/2820 [00:39<29:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 63/2820 [00:40<29:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 63/2820 [00:40<29:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 64/2820 [00:41<29:55,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 64/2820 [00:41<29:55,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 65/2820 [00:41<29:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 65/2820 [00:41<29:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 66/2820 [00:41<29:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 66/2820 [00:41<29:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 67/2820 [00:43<29:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 67/2820 [00:43<29:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 68/2820 [00:44<29:41,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 68/2820 [00:44<29:41,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 69/2820 [00:44<29:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 69/2820 [00:44<29:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 70/2820 [00:45<29:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   2%|▏         | 70/2820 [00:45<29:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 71/2820 [00:45<29:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 71/2820 [00:45<29:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 72/2820 [00:45<29:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 72/2820 [00:45<29:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 73/2820 [00:47<29:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 73/2820 [00:47<29:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 74/2820 [00:47<29:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 74/2820 [00:47<29:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 75/2820 [00:47<29:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 75/2820 [00:47<29:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 76/2820 [00:49<29:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 76/2820 [00:49<29:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 77/2820 [00:49<29:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 77/2820 [00:49<29:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 78/2820 [00:49<28:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 78/2820 [00:49<28:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 79/2820 [00:51<29:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 79/2820 [00:51<29:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 80/2820 [00:51<29:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 80/2820 [00:51<29:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 81/2820 [00:51<29:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 81/2820 [00:51<29:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 82/2820 [00:53<29:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 82/2820 [00:53<29:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 83/2820 [00:53<29:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 83/2820 [00:53<29:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 84/2820 [00:53<29:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 84/2820 [00:53<29:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 85/2820 [00:55<29:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 85/2820 [00:55<29:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 86/2820 [00:55<29:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 86/2820 [00:55<29:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 87/2820 [00:55<29:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 87/2820 [00:55<29:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 88/2820 [00:56<29:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 88/2820 [00:56<29:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 89/2820 [00:57<29:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 89/2820 [00:57<29:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 90/2820 [00:57<29:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 90/2820 [00:57<29:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 91/2820 [00:58<29:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 91/2820 [00:58<29:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 92/2820 [00:59<29:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 92/2820 [00:59<29:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 93/2820 [00:59<29:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 93/2820 [00:59<29:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 94/2820 [01:00<29:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 94/2820 [01:00<29:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 95/2820 [01:01<29:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 95/2820 [01:01<29:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 96/2820 [01:01<29:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 96/2820 [01:01<29:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 97/2820 [01:02<29:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 97/2820 [01:02<29:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 98/2820 [01:03<29:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   3%|▎         | 98/2820 [01:03<29:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 99/2820 [01:03<28:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 99/2820 [01:03<28:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 100/2820 [01:04<29:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 100/2820 [01:04<29:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 101/2820 [01:05<29:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 101/2820 [01:05<29:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 102/2820 [01:05<28:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 102/2820 [01:05<28:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 103/2820 [01:06<29:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 103/2820 [01:06<29:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 104/2820 [01:06<29:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 104/2820 [01:06<29:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 105/2820 [01:06<28:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▎         | 105/2820 [01:06<28:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 106/2820 [01:08<29:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 106/2820 [01:08<29:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 107/2820 [01:08<29:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 107/2820 [01:08<29:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 108/2820 [01:09<28:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 108/2820 [01:09<28:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 109/2820 [01:10<29:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 109/2820 [01:10<29:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 110/2820 [01:10<29:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 110/2820 [01:10<29:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 111/2820 [01:10<28:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 111/2820 [01:10<28:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 112/2820 [01:12<29:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 112/2820 [01:12<29:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 113/2820 [01:12<29:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 113/2820 [01:12<29:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 114/2820 [01:12<28:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 114/2820 [01:12<28:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 115/2820 [01:13<28:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 115/2820 [01:13<28:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 116/2820 [01:14<29:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 116/2820 [01:14<29:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 117/2820 [01:14<28:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 117/2820 [01:14<28:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 118/2820 [01:15<28:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 118/2820 [01:15<28:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 119/2820 [01:16<28:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 119/2820 [01:16<28:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 120/2820 [01:16<28:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 120/2820 [01:16<28:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 121/2820 [01:17<28:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 121/2820 [01:17<28:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 122/2820 [01:18<28:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 122/2820 [01:18<28:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 123/2820 [01:18<28:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 123/2820 [01:18<28:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 124/2820 [01:19<28:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 124/2820 [01:19<28:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 125/2820 [01:20<28:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 125/2820 [01:20<28:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 126/2820 [01:20<28:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   4%|▍         | 126/2820 [01:20<28:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 127/2820 [01:21<28:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 127/2820 [01:21<28:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 128/2820 [01:21<28:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 128/2820 [01:21<28:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 129/2820 [01:22<28:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 129/2820 [01:22<28:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 130/2820 [01:23<28:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 130/2820 [01:23<28:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 131/2820 [01:24<28:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 131/2820 [01:24<28:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 132/2820 [01:24<28:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 132/2820 [01:24<28:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 133/2820 [01:24<28:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 133/2820 [01:24<28:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 134/2820 [01:26<28:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 134/2820 [01:26<28:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 135/2820 [01:26<28:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 135/2820 [01:26<28:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 136/2820 [01:26<28:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 136/2820 [01:26<28:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 137/2820 [01:28<28:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 137/2820 [01:28<28:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 138/2820 [01:28<28:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 138/2820 [01:28<28:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 139/2820 [01:28<28:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 139/2820 [01:28<28:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 140/2820 [01:30<28:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▍         | 140/2820 [01:30<28:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 141/2820 [01:30<28:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 141/2820 [01:30<28:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 142/2820 [01:30<28:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 142/2820 [01:30<28:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 143/2820 [01:31<28:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 143/2820 [01:31<28:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 144/2820 [01:31<28:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 144/2820 [01:31<28:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 145/2820 [01:32<28:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 145/2820 [01:32<28:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 146/2820 [01:33<28:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 146/2820 [01:33<28:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 147/2820 [01:33<28:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 147/2820 [01:33<28:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 148/2820 [01:34<28:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 148/2820 [01:34<28:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 149/2820 [01:35<28:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 149/2820 [01:35<28:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 150/2820 [01:35<28:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 150/2820 [01:35<28:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 151/2820 [01:35<28:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 151/2820 [01:35<28:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 152/2820 [01:37<28:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 152/2820 [01:37<28:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 153/2820 [01:37<28:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 153/2820 [01:37<28:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 154/2820 [01:37<28:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 154/2820 [01:37<28:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 155/2820 [01:39<28:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   5%|▌         | 155/2820 [01:39<28:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 156/2820 [01:39<28:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 156/2820 [01:39<28:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 157/2820 [01:39<28:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 157/2820 [01:39<28:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 158/2820 [01:41<28:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 158/2820 [01:41<28:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 159/2820 [01:41<28:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 159/2820 [01:41<28:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 160/2820 [01:41<28:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 160/2820 [01:41<28:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 161/2820 [01:43<28:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 161/2820 [01:43<28:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 162/2820 [01:43<28:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 162/2820 [01:43<28:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 163/2820 [01:43<28:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 163/2820 [01:43<28:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 164/2820 [01:45<28:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 164/2820 [01:45<28:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 165/2820 [01:45<28:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 165/2820 [01:45<28:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 166/2820 [01:45<28:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 166/2820 [01:45<28:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 167/2820 [01:47<28:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 167/2820 [01:47<28:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 168/2820 [01:47<28:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 168/2820 [01:47<28:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 169/2820 [01:47<28:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 169/2820 [01:47<28:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 170/2820 [01:49<28:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 170/2820 [01:49<28:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 171/2820 [01:49<28:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 171/2820 [01:49<28:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 172/2820 [01:49<27:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 172/2820 [01:49<27:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 173/2820 [01:50<28:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 173/2820 [01:50<28:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 174/2820 [01:50<28:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 174/2820 [01:50<28:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 175/2820 [01:50<27:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 175/2820 [01:50<27:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 176/2820 [01:52<28:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▌         | 176/2820 [01:52<28:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 177/2820 [01:52<28:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 177/2820 [01:52<28:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 178/2820 [01:52<27:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 178/2820 [01:52<27:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 179/2820 [01:54<28:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 179/2820 [01:54<28:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 180/2820 [01:54<28:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 180/2820 [01:54<28:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 181/2820 [01:54<27:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 181/2820 [01:54<27:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 182/2820 [01:56<28:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 182/2820 [01:56<28:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 183/2820 [01:56<28:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   6%|▋         | 183/2820 [01:56<28:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 184/2820 [01:56<27:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 184/2820 [01:56<27:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 185/2820 [01:58<28:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 185/2820 [01:58<28:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 186/2820 [01:58<27:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 186/2820 [01:58<27:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 187/2820 [01:58<27:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 187/2820 [01:58<27:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 188/2820 [02:00<28:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 188/2820 [02:00<28:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 189/2820 [02:00<27:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 189/2820 [02:00<27:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 190/2820 [02:00<27:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 190/2820 [02:00<27:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 191/2820 [02:02<28:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 191/2820 [02:02<28:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 192/2820 [02:02<27:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 192/2820 [02:02<27:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 193/2820 [02:02<27:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 193/2820 [02:02<27:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 194/2820 [02:04<27:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 194/2820 [02:04<27:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 195/2820 [02:04<27:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 195/2820 [02:04<27:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 196/2820 [02:04<27:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 196/2820 [02:04<27:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 197/2820 [02:05<27:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 197/2820 [02:05<27:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 198/2820 [02:06<27:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 198/2820 [02:06<27:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 199/2820 [02:06<27:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 199/2820 [02:06<27:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 200/2820 [02:07<27:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 200/2820 [02:07<27:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 201/2820 [02:07<27:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 201/2820 [02:07<27:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 202/2820 [02:07<27:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 202/2820 [02:07<27:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 203/2820 [02:09<27:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 203/2820 [02:09<27:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 204/2820 [02:09<27:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 204/2820 [02:09<27:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 205/2820 [02:09<27:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 205/2820 [02:09<27:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 206/2820 [02:11<27:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 206/2820 [02:11<27:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 207/2820 [02:11<27:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 207/2820 [02:11<27:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 208/2820 [02:11<27:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 208/2820 [02:11<27:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 209/2820 [02:13<27:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 209/2820 [02:13<27:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 210/2820 [02:13<27:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 210/2820 [02:13<27:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 211/2820 [02:13<27:31,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   7%|▋         | 211/2820 [02:13<27:31,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 212/2820 [02:15<27:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 212/2820 [02:15<27:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 213/2820 [02:15<27:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 213/2820 [02:15<27:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 214/2820 [02:15<27:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 214/2820 [02:15<27:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 215/2820 [02:17<27:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 215/2820 [02:17<27:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 216/2820 [02:17<27:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 216/2820 [02:17<27:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 217/2820 [02:17<27:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 217/2820 [02:17<27:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 218/2820 [02:18<27:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 218/2820 [02:18<27:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 219/2820 [02:18<27:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 219/2820 [02:18<27:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 220/2820 [02:18<27:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 220/2820 [02:18<27:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 221/2820 [02:20<27:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 221/2820 [02:20<27:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 222/2820 [02:20<27:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 222/2820 [02:20<27:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 223/2820 [02:20<27:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 223/2820 [02:20<27:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 224/2820 [02:22<27:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 224/2820 [02:22<27:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 225/2820 [02:22<27:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 225/2820 [02:22<27:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 226/2820 [02:22<27:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 226/2820 [02:22<27:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 227/2820 [02:24<27:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 227/2820 [02:24<27:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 228/2820 [02:24<27:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 228/2820 [02:24<27:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 229/2820 [02:24<27:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 229/2820 [02:24<27:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 230/2820 [02:26<27:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 230/2820 [02:26<27:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 231/2820 [02:26<27:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 231/2820 [02:26<27:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 232/2820 [02:26<27:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 232/2820 [02:26<27:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 233/2820 [02:28<27:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 233/2820 [02:28<27:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 234/2820 [02:28<27:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 234/2820 [02:28<27:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 235/2820 [02:28<27:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 235/2820 [02:28<27:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 236/2820 [02:30<27:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 236/2820 [02:30<27:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 237/2820 [02:30<27:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 237/2820 [02:30<27:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 238/2820 [02:30<27:10,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 238/2820 [02:30<27:10,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 239/2820 [02:32<27:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   8%|▊         | 239/2820 [02:32<27:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 240/2820 [02:32<27:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 240/2820 [02:32<27:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 241/2820 [02:32<27:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 241/2820 [02:32<27:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 242/2820 [02:34<27:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 242/2820 [02:34<27:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 243/2820 [02:34<27:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 243/2820 [02:34<27:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 244/2820 [02:34<27:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 244/2820 [02:34<27:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 245/2820 [02:35<27:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 245/2820 [02:35<27:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 246/2820 [02:35<27:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▊         | 246/2820 [02:35<27:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 247/2820 [02:35<27:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 247/2820 [02:35<27:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 248/2820 [02:37<27:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 248/2820 [02:37<27:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 249/2820 [02:37<27:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 249/2820 [02:37<27:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 250/2820 [02:37<27:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 250/2820 [02:37<27:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 251/2820 [02:39<27:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 251/2820 [02:39<27:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 252/2820 [02:39<27:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 252/2820 [02:39<27:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 253/2820 [02:39<27:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 253/2820 [02:39<27:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 254/2820 [02:41<27:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 254/2820 [02:41<27:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 255/2820 [02:41<27:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 255/2820 [02:41<27:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 256/2820 [02:41<26:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 256/2820 [02:41<26:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 257/2820 [02:43<27:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 257/2820 [02:43<27:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 258/2820 [02:43<27:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 258/2820 [02:43<27:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 259/2820 [02:43<26:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 259/2820 [02:43<26:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 260/2820 [02:45<27:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 260/2820 [02:45<27:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 261/2820 [02:45<27:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 261/2820 [02:45<27:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 262/2820 [02:45<26:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 262/2820 [02:45<26:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 263/2820 [02:47<27:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 263/2820 [02:47<27:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 264/2820 [02:47<26:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 264/2820 [02:47<26:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 265/2820 [02:47<26:51,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 265/2820 [02:47<26:51,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 266/2820 [02:49<27:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 266/2820 [02:49<27:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 267/2820 [02:49<26:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:   9%|▉         | 267/2820 [02:49<26:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 268/2820 [02:49<26:49,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 268/2820 [02:49<26:49,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 269/2820 [02:50<26:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 269/2820 [02:50<26:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 270/2820 [02:50<26:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 270/2820 [02:50<26:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 271/2820 [02:50<26:46,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 271/2820 [02:50<26:46,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 272/2820 [02:52<26:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 272/2820 [02:52<26:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 273/2820 [02:52<26:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 273/2820 [02:52<26:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 274/2820 [02:52<26:43,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 274/2820 [02:52<26:43,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 275/2820 [02:54<26:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 275/2820 [02:54<26:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 276/2820 [02:54<26:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 276/2820 [02:54<26:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 277/2820 [02:54<26:40,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 277/2820 [02:54<26:40,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 278/2820 [02:56<26:51,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 278/2820 [02:56<26:51,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 279/2820 [02:56<26:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 279/2820 [02:56<26:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 280/2820 [02:56<26:39,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 280/2820 [02:56<26:39,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 281/2820 [02:58<26:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|▉         | 281/2820 [02:58<26:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 282/2820 [02:58<26:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 282/2820 [02:58<26:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 283/2820 [02:58<26:38,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 283/2820 [02:58<26:38,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 284/2820 [03:00<26:48,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 284/2820 [03:00<26:48,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 285/2820 [03:00<26:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 285/2820 [03:00<26:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 286/2820 [03:00<26:36,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 286/2820 [03:00<26:36,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 287/2820 [03:02<26:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 287/2820 [03:02<26:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 288/2820 [03:02<26:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 288/2820 [03:02<26:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 289/2820 [03:02<26:36,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 289/2820 [03:02<26:36,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 290/2820 [03:04<26:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 290/2820 [03:04<26:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 291/2820 [03:04<26:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 291/2820 [03:04<26:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 292/2820 [03:04<26:33,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 292/2820 [03:04<26:33,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 293/2820 [03:05<26:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 293/2820 [03:05<26:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 294/2820 [03:05<26:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 294/2820 [03:05<26:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 295/2820 [03:06<26:32,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 295/2820 [03:06<26:32,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 296/2820 [03:07<26:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  10%|█         | 296/2820 [03:07<26:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 297/2820 [03:07<26:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 297/2820 [03:07<26:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 298/2820 [03:08<26:31,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 298/2820 [03:08<26:31,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 299/2820 [03:09<26:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 299/2820 [03:09<26:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 300/2820 [03:09<26:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 300/2820 [03:09<26:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 301/2820 [03:09<26:29,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 301/2820 [03:09<26:29,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 302/2820 [03:11<26:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 302/2820 [03:11<26:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 303/2820 [03:11<26:33,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 303/2820 [03:11<26:33,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 304/2820 [03:11<26:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 304/2820 [03:11<26:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 305/2820 [03:13<26:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 305/2820 [03:13<26:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 306/2820 [03:13<26:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 306/2820 [03:13<26:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 307/2820 [03:13<26:25,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 307/2820 [03:13<26:25,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 308/2820 [03:15<26:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 308/2820 [03:15<26:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 309/2820 [03:15<26:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 309/2820 [03:15<26:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 310/2820 [03:15<26:23,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 310/2820 [03:15<26:23,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 311/2820 [03:17<26:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 311/2820 [03:17<26:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 312/2820 [03:17<26:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 312/2820 [03:17<26:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 313/2820 [03:17<26:21,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 313/2820 [03:17<26:21,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 314/2820 [03:19<26:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 314/2820 [03:19<26:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 315/2820 [03:19<26:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 315/2820 [03:19<26:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 316/2820 [03:19<26:19,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 316/2820 [03:19<26:19,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 317/2820 [03:21<26:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█         | 317/2820 [03:21<26:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 318/2820 [03:21<26:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 318/2820 [03:21<26:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 319/2820 [03:21<26:16,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 319/2820 [03:21<26:16,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 320/2820 [03:22<26:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 320/2820 [03:22<26:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 321/2820 [03:22<26:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 321/2820 [03:22<26:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 322/2820 [03:22<26:14,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 322/2820 [03:22<26:14,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 323/2820 [03:24<26:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 323/2820 [03:24<26:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 324/2820 [03:24<26:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  11%|█▏        | 324/2820 [03:24<26:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 325/2820 [03:24<26:13,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 325/2820 [03:24<26:13,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 326/2820 [03:26<26:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 326/2820 [03:26<26:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 327/2820 [03:26<26:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 327/2820 [03:26<26:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 328/2820 [03:26<26:11,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 328/2820 [03:26<26:11,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 329/2820 [03:28<26:20,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 329/2820 [03:28<26:20,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 330/2820 [03:28<26:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 330/2820 [03:28<26:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 331/2820 [03:28<26:10,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 331/2820 [03:28<26:10,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 332/2820 [03:30<26:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 332/2820 [03:30<26:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 333/2820 [03:30<26:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 333/2820 [03:30<26:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 334/2820 [03:30<26:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 334/2820 [03:30<26:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 335/2820 [03:32<26:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 335/2820 [03:32<26:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 336/2820 [03:32<26:10,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 336/2820 [03:32<26:10,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 337/2820 [03:32<26:05,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 337/2820 [03:32<26:05,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 338/2820 [03:34<26:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 338/2820 [03:34<26:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 339/2820 [03:34<26:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 339/2820 [03:34<26:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 340/2820 [03:34<26:03,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 340/2820 [03:34<26:03,  1.59it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 341/2820 [03:36<26:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 341/2820 [03:36<26:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 342/2820 [03:36<26:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 342/2820 [03:36<26:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 343/2820 [03:36<26:03,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 343/2820 [03:36<26:03,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 344/2820 [03:38<26:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 344/2820 [03:38<26:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 345/2820 [03:38<26:06,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 345/2820 [03:38<26:06,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 346/2820 [03:38<26:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 346/2820 [03:38<26:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 347/2820 [03:40<26:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 347/2820 [03:40<26:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 348/2820 [03:40<26:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 348/2820 [03:40<26:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 349/2820 [03:40<26:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 349/2820 [03:40<26:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 350/2820 [03:42<26:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 350/2820 [03:42<26:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 351/2820 [03:42<26:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 351/2820 [03:42<26:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 352/2820 [03:42<25:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  12%|█▏        | 352/2820 [03:42<25:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 353/2820 [03:44<26:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 353/2820 [03:44<26:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 354/2820 [03:44<26:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 354/2820 [03:44<26:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 355/2820 [03:44<25:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 355/2820 [03:44<25:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 356/2820 [03:46<26:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 356/2820 [03:46<26:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 357/2820 [03:46<26:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 357/2820 [03:46<26:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 358/2820 [03:46<25:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 358/2820 [03:46<25:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 359/2820 [03:48<26:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 359/2820 [03:48<26:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 360/2820 [03:48<25:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 360/2820 [03:48<25:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 361/2820 [03:48<25:54,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 361/2820 [03:48<25:54,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 362/2820 [03:50<26:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 362/2820 [03:50<26:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 363/2820 [03:50<25:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 363/2820 [03:50<25:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 364/2820 [03:50<25:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 364/2820 [03:50<25:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 365/2820 [03:52<26:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 365/2820 [03:52<26:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 366/2820 [03:52<25:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 366/2820 [03:52<25:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 367/2820 [03:52<25:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 367/2820 [03:52<25:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 368/2820 [03:53<25:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 368/2820 [03:53<25:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 369/2820 [03:53<25:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 369/2820 [03:53<25:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 370/2820 [03:53<25:49,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 370/2820 [03:53<25:49,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 371/2820 [03:55<25:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 371/2820 [03:55<25:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 372/2820 [03:55<25:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 372/2820 [03:55<25:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 373/2820 [03:55<25:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 373/2820 [03:55<25:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 374/2820 [03:57<25:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 374/2820 [03:57<25:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 375/2820 [03:57<25:48,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 375/2820 [03:57<25:48,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 376/2820 [03:57<25:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 376/2820 [03:57<25:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 377/2820 [03:59<25:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 377/2820 [03:59<25:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 378/2820 [03:59<25:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 378/2820 [03:59<25:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 379/2820 [03:59<25:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 379/2820 [03:59<25:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 380/2820 [04:01<25:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  13%|█▎        | 380/2820 [04:01<25:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 381/2820 [04:01<25:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 381/2820 [04:01<25:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 382/2820 [04:01<25:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 382/2820 [04:01<25:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 383/2820 [04:03<25:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 383/2820 [04:03<25:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 384/2820 [04:03<25:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 384/2820 [04:03<25:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 385/2820 [04:03<25:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 385/2820 [04:03<25:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 386/2820 [04:05<25:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 386/2820 [04:05<25:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 387/2820 [04:05<25:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▎        | 387/2820 [04:05<25:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 388/2820 [04:05<25:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 388/2820 [04:05<25:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 389/2820 [04:07<25:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 389/2820 [04:07<25:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 390/2820 [04:07<25:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 390/2820 [04:07<25:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 391/2820 [04:07<25:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 391/2820 [04:07<25:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 392/2820 [04:09<25:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 392/2820 [04:09<25:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 393/2820 [04:09<25:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 393/2820 [04:09<25:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 394/2820 [04:09<25:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 394/2820 [04:09<25:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 395/2820 [04:11<25:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 395/2820 [04:11<25:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 396/2820 [04:11<25:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 396/2820 [04:11<25:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 397/2820 [04:11<25:33,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 397/2820 [04:11<25:33,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 398/2820 [04:13<25:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 398/2820 [04:13<25:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 399/2820 [04:13<25:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 399/2820 [04:13<25:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 400/2820 [04:13<25:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 400/2820 [04:13<25:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 401/2820 [04:15<25:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 401/2820 [04:15<25:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 402/2820 [04:15<25:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 402/2820 [04:15<25:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 403/2820 [04:15<25:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 403/2820 [04:15<25:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 404/2820 [04:17<25:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 404/2820 [04:17<25:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 405/2820 [04:17<25:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 405/2820 [04:17<25:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 406/2820 [04:17<25:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 406/2820 [04:17<25:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 407/2820 [04:18<25:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 407/2820 [04:18<25:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 408/2820 [04:18<25:31,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  14%|█▍        | 408/2820 [04:18<25:31,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 409/2820 [04:19<25:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 409/2820 [04:19<25:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 410/2820 [04:20<25:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 410/2820 [04:20<25:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 411/2820 [04:20<25:29,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 411/2820 [04:20<25:29,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 412/2820 [04:20<25:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 412/2820 [04:20<25:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 413/2820 [04:22<25:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 413/2820 [04:22<25:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 414/2820 [04:22<25:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 414/2820 [04:22<25:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 415/2820 [04:22<25:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 415/2820 [04:22<25:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 416/2820 [04:24<25:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 416/2820 [04:24<25:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 417/2820 [04:24<25:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 417/2820 [04:24<25:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 418/2820 [04:24<25:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 418/2820 [04:24<25:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 419/2820 [04:26<25:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 419/2820 [04:26<25:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 420/2820 [04:26<25:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 420/2820 [04:26<25:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 421/2820 [04:26<25:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 421/2820 [04:26<25:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 422/2820 [04:28<25:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▍        | 422/2820 [04:28<25:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 423/2820 [04:28<25:20,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 423/2820 [04:28<25:20,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 424/2820 [04:28<25:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 424/2820 [04:28<25:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 425/2820 [04:30<25:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 425/2820 [04:30<25:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 426/2820 [04:30<25:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 426/2820 [04:30<25:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 427/2820 [04:30<25:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 427/2820 [04:30<25:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 428/2820 [04:32<25:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 428/2820 [04:32<25:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 429/2820 [04:32<25:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 429/2820 [04:32<25:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 430/2820 [04:32<25:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 430/2820 [04:32<25:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 431/2820 [04:34<25:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 431/2820 [04:34<25:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 432/2820 [04:34<25:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 432/2820 [04:34<25:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 433/2820 [04:34<25:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 433/2820 [04:34<25:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 434/2820 [04:36<25:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 434/2820 [04:36<25:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 435/2820 [04:36<25:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 435/2820 [04:36<25:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 436/2820 [04:36<25:09,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 436/2820 [04:36<25:09,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 437/2820 [04:37<25:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  15%|█▌        | 437/2820 [04:37<25:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 438/2820 [04:38<25:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 438/2820 [04:38<25:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 439/2820 [04:38<25:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 439/2820 [04:38<25:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 440/2820 [04:40<25:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 440/2820 [04:40<25:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 441/2820 [04:40<25:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 441/2820 [04:40<25:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 442/2820 [04:40<25:06,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 442/2820 [04:40<25:06,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 443/2820 [04:42<25:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 443/2820 [04:42<25:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 444/2820 [04:42<25:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 444/2820 [04:42<25:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 445/2820 [04:42<25:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 445/2820 [04:42<25:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 446/2820 [04:43<25:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 446/2820 [04:43<25:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 447/2820 [04:43<25:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 447/2820 [04:43<25:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 448/2820 [04:43<25:03,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 448/2820 [04:43<25:03,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 449/2820 [04:45<25:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 449/2820 [04:45<25:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 450/2820 [04:45<25:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 450/2820 [04:45<25:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 451/2820 [04:45<25:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 451/2820 [04:45<25:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 452/2820 [04:47<25:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 452/2820 [04:47<25:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 453/2820 [04:47<25:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 453/2820 [04:47<25:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 454/2820 [04:47<24:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 454/2820 [04:47<24:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 455/2820 [04:49<25:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 455/2820 [04:49<25:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 456/2820 [04:49<25:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 456/2820 [04:49<25:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 457/2820 [04:49<24:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 457/2820 [04:49<24:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 458/2820 [04:51<25:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▌        | 458/2820 [04:51<25:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 459/2820 [04:51<24:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 459/2820 [04:51<24:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 460/2820 [04:51<24:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 460/2820 [04:51<24:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 461/2820 [04:53<25:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 461/2820 [04:53<25:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 462/2820 [04:53<24:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 462/2820 [04:53<24:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 463/2820 [04:53<24:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 463/2820 [04:53<24:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 464/2820 [04:55<24:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 464/2820 [04:55<24:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 465/2820 [04:55<24:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  16%|█▋        | 465/2820 [04:55<24:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 466/2820 [04:55<24:51,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 466/2820 [04:55<24:51,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 467/2820 [04:57<24:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 467/2820 [04:57<24:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 468/2820 [04:57<24:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 468/2820 [04:57<24:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 469/2820 [04:57<24:49,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 469/2820 [04:57<24:49,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 470/2820 [04:58<24:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 470/2820 [04:58<24:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 471/2820 [04:58<24:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 471/2820 [04:58<24:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 472/2820 [04:58<24:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 472/2820 [04:58<24:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 473/2820 [05:00<24:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 473/2820 [05:00<24:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 474/2820 [05:00<24:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 474/2820 [05:00<24:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 475/2820 [05:01<24:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 475/2820 [05:01<24:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 476/2820 [05:02<24:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 476/2820 [05:02<24:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 477/2820 [05:02<24:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 477/2820 [05:02<24:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 478/2820 [05:02<24:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 478/2820 [05:02<24:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 479/2820 [05:04<24:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 479/2820 [05:04<24:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 480/2820 [05:04<24:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 480/2820 [05:04<24:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 481/2820 [05:04<24:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 481/2820 [05:04<24:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 482/2820 [05:06<24:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 482/2820 [05:06<24:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 483/2820 [05:06<24:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 483/2820 [05:06<24:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 484/2820 [05:06<24:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 484/2820 [05:06<24:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 485/2820 [05:08<24:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 485/2820 [05:08<24:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 486/2820 [05:08<24:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 486/2820 [05:08<24:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 487/2820 [05:08<24:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 487/2820 [05:08<24:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 488/2820 [05:10<24:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 488/2820 [05:10<24:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 489/2820 [05:10<24:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 489/2820 [05:10<24:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 490/2820 [05:10<24:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 490/2820 [05:10<24:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 491/2820 [05:12<24:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 491/2820 [05:12<24:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 492/2820 [05:12<24:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 492/2820 [05:12<24:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 493/2820 [05:12<24:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  17%|█▋        | 493/2820 [05:12<24:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 494/2820 [05:14<24:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 494/2820 [05:14<24:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 495/2820 [05:14<24:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 495/2820 [05:14<24:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 496/2820 [05:14<24:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 496/2820 [05:14<24:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 497/2820 [05:16<24:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 497/2820 [05:16<24:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 498/2820 [05:16<24:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 498/2820 [05:16<24:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 499/2820 [05:16<24:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 499/2820 [05:16<24:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 500/2820 [05:18<24:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 500/2820 [05:18<24:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 501/2820 [05:18<24:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 501/2820 [05:18<24:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 502/2820 [05:18<24:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 502/2820 [05:18<24:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 503/2820 [05:19<24:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 503/2820 [05:19<24:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 504/2820 [05:19<24:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 504/2820 [05:19<24:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 505/2820 [05:20<24:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 505/2820 [05:20<24:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 506/2820 [05:21<24:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 506/2820 [05:21<24:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 507/2820 [05:21<24:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 507/2820 [05:21<24:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 508/2820 [05:21<24:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 508/2820 [05:21<24:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 509/2820 [05:23<24:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 509/2820 [05:23<24:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 510/2820 [05:23<24:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 510/2820 [05:23<24:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 511/2820 [05:24<24:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 511/2820 [05:24<24:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 512/2820 [05:25<24:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 512/2820 [05:25<24:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 513/2820 [05:25<24:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 513/2820 [05:25<24:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 514/2820 [05:25<24:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 514/2820 [05:25<24:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 515/2820 [05:27<24:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 515/2820 [05:27<24:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 516/2820 [05:27<24:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 516/2820 [05:27<24:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 517/2820 [05:28<24:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 517/2820 [05:28<24:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 518/2820 [05:29<24:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 518/2820 [05:29<24:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 519/2820 [05:29<24:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 519/2820 [05:29<24:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 520/2820 [05:30<24:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 520/2820 [05:30<24:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 521/2820 [05:31<24:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  18%|█▊        | 521/2820 [05:31<24:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 522/2820 [05:31<24:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 522/2820 [05:31<24:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 523/2820 [05:31<24:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 523/2820 [05:31<24:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 524/2820 [05:33<24:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 524/2820 [05:33<24:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 525/2820 [05:33<24:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 525/2820 [05:33<24:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 526/2820 [05:33<24:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 526/2820 [05:33<24:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 527/2820 [05:35<24:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 527/2820 [05:35<24:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 528/2820 [05:35<24:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▊        | 528/2820 [05:35<24:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 529/2820 [05:35<24:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 529/2820 [05:35<24:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 530/2820 [05:37<24:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 530/2820 [05:37<24:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 531/2820 [05:37<24:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 531/2820 [05:37<24:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 532/2820 [05:37<24:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 532/2820 [05:37<24:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 533/2820 [05:39<24:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 533/2820 [05:39<24:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 534/2820 [05:39<24:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 534/2820 [05:39<24:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 535/2820 [05:39<24:09,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 535/2820 [05:39<24:09,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 536/2820 [05:41<24:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 536/2820 [05:41<24:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 537/2820 [05:41<24:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 537/2820 [05:41<24:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 538/2820 [05:41<24:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 538/2820 [05:41<24:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 539/2820 [05:43<24:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 539/2820 [05:43<24:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 540/2820 [05:43<24:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 540/2820 [05:43<24:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 541/2820 [05:43<24:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 541/2820 [05:43<24:05,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 542/2820 [05:44<24:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 542/2820 [05:44<24:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 543/2820 [05:44<24:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 543/2820 [05:44<24:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 544/2820 [05:44<24:03,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 544/2820 [05:44<24:03,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 545/2820 [05:46<24:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 545/2820 [05:46<24:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 546/2820 [05:46<24:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 546/2820 [05:46<24:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 547/2820 [05:46<24:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 547/2820 [05:46<24:01,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 548/2820 [05:48<24:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 548/2820 [05:48<24:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 549/2820 [05:48<24:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  19%|█▉        | 549/2820 [05:48<24:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 550/2820 [05:48<23:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 550/2820 [05:48<23:59,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 551/2820 [05:50<24:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 551/2820 [05:50<24:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 552/2820 [05:50<24:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 552/2820 [05:50<24:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 553/2820 [05:50<23:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 553/2820 [05:50<23:57,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 554/2820 [05:52<24:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 554/2820 [05:52<24:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 555/2820 [05:52<23:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 555/2820 [05:52<23:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 556/2820 [05:52<23:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 556/2820 [05:52<23:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 557/2820 [05:54<23:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 557/2820 [05:54<23:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 558/2820 [05:54<23:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 558/2820 [05:54<23:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 559/2820 [05:54<23:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 559/2820 [05:54<23:53,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 560/2820 [05:56<23:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 560/2820 [05:56<23:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 561/2820 [05:56<23:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 561/2820 [05:56<23:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 562/2820 [05:56<23:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 562/2820 [05:56<23:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 563/2820 [05:58<23:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|█▉        | 563/2820 [05:58<23:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 564/2820 [05:58<23:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 564/2820 [05:58<23:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 565/2820 [05:58<23:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 565/2820 [05:58<23:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 566/2820 [06:00<23:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 566/2820 [06:00<23:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 567/2820 [06:00<23:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 567/2820 [06:00<23:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 568/2820 [06:00<23:48,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 568/2820 [06:00<23:48,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 569/2820 [06:01<23:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 569/2820 [06:01<23:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 570/2820 [06:02<23:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 570/2820 [06:02<23:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 571/2820 [06:02<23:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 571/2820 [06:02<23:46,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 572/2820 [06:03<23:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 572/2820 [06:03<23:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 573/2820 [06:03<23:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 573/2820 [06:03<23:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 574/2820 [06:03<23:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 574/2820 [06:03<23:44,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 575/2820 [06:05<23:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 575/2820 [06:05<23:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 576/2820 [06:05<23:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 576/2820 [06:05<23:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 577/2820 [06:05<23:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 577/2820 [06:05<23:41,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 578/2820 [06:07<23:46,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  20%|██        | 578/2820 [06:07<23:46,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 579/2820 [06:07<23:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 579/2820 [06:07<23:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 580/2820 [06:07<23:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 580/2820 [06:07<23:40,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 581/2820 [06:09<23:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 581/2820 [06:09<23:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 582/2820 [06:09<23:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 582/2820 [06:09<23:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 583/2820 [06:09<23:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 583/2820 [06:09<23:38,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 584/2820 [06:11<23:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 584/2820 [06:11<23:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 585/2820 [06:11<23:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 585/2820 [06:11<23:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 586/2820 [06:11<23:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 586/2820 [06:11<23:36,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 587/2820 [06:13<23:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 587/2820 [06:13<23:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 588/2820 [06:13<23:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 588/2820 [06:13<23:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 589/2820 [06:13<23:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 589/2820 [06:13<23:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 590/2820 [06:15<23:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 590/2820 [06:15<23:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 591/2820 [06:15<23:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 591/2820 [06:15<23:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 592/2820 [06:15<23:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 592/2820 [06:15<23:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 593/2820 [06:17<23:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 593/2820 [06:17<23:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 594/2820 [06:17<23:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 594/2820 [06:17<23:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 595/2820 [06:17<23:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 595/2820 [06:17<23:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 596/2820 [06:19<23:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 596/2820 [06:19<23:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 597/2820 [06:19<23:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 597/2820 [06:19<23:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 598/2820 [06:19<23:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 598/2820 [06:19<23:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 599/2820 [06:20<23:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██        | 599/2820 [06:20<23:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 600/2820 [06:20<23:29,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 600/2820 [06:20<23:29,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 601/2820 [06:20<23:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 601/2820 [06:20<23:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 602/2820 [06:22<23:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 602/2820 [06:22<23:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 603/2820 [06:22<23:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 603/2820 [06:22<23:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 604/2820 [06:22<23:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 604/2820 [06:22<23:24,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 605/2820 [06:24<23:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 605/2820 [06:24<23:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 606/2820 [06:24<23:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  21%|██▏       | 606/2820 [06:24<23:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 607/2820 [06:24<23:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 607/2820 [06:24<23:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 608/2820 [06:26<23:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 608/2820 [06:26<23:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 609/2820 [06:26<23:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 609/2820 [06:26<23:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 610/2820 [06:26<23:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 610/2820 [06:26<23:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 611/2820 [06:28<23:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 611/2820 [06:28<23:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 612/2820 [06:28<23:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 612/2820 [06:28<23:21,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 613/2820 [06:28<23:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 613/2820 [06:28<23:18,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 614/2820 [06:30<23:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 614/2820 [06:30<23:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 615/2820 [06:30<23:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 615/2820 [06:30<23:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 616/2820 [06:30<23:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 616/2820 [06:30<23:16,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 617/2820 [06:32<23:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 617/2820 [06:32<23:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 618/2820 [06:32<23:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 618/2820 [06:32<23:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 619/2820 [06:32<23:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 619/2820 [06:32<23:14,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 620/2820 [06:34<23:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 620/2820 [06:34<23:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 621/2820 [06:34<23:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 621/2820 [06:34<23:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 622/2820 [06:34<23:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 622/2820 [06:34<23:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 623/2820 [06:36<23:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 623/2820 [06:36<23:17,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 624/2820 [06:36<23:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 624/2820 [06:36<23:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 625/2820 [06:36<23:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 625/2820 [06:36<23:11,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 626/2820 [06:38<23:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 626/2820 [06:38<23:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 627/2820 [06:38<23:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 627/2820 [06:38<23:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 628/2820 [06:38<23:09,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 628/2820 [06:38<23:09,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 629/2820 [06:40<23:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 629/2820 [06:40<23:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 630/2820 [06:40<23:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 630/2820 [06:40<23:10,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 631/2820 [06:40<23:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 631/2820 [06:40<23:07,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 632/2820 [06:41<23:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 632/2820 [06:41<23:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 633/2820 [06:42<23:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 633/2820 [06:42<23:08,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 634/2820 [06:42<23:06,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  22%|██▏       | 634/2820 [06:42<23:06,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 635/2820 [06:43<23:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 635/2820 [06:43<23:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 636/2820 [06:43<23:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 636/2820 [06:43<23:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 637/2820 [06:43<23:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 637/2820 [06:43<23:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 638/2820 [06:45<23:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 638/2820 [06:45<23:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 639/2820 [06:45<23:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 639/2820 [06:45<23:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 640/2820 [06:45<23:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 640/2820 [06:45<23:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 641/2820 [06:47<23:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 641/2820 [06:47<23:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 642/2820 [06:47<23:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 642/2820 [06:47<23:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 643/2820 [06:47<23:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 643/2820 [06:47<23:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 644/2820 [06:49<23:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 644/2820 [06:49<23:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 645/2820 [06:49<23:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 645/2820 [06:49<23:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 646/2820 [06:49<22:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 646/2820 [06:49<22:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 647/2820 [06:51<23:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 647/2820 [06:51<23:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 648/2820 [06:51<22:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 648/2820 [06:51<22:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 649/2820 [06:51<22:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 649/2820 [06:51<22:56,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 650/2820 [06:53<22:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 650/2820 [06:53<22:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 651/2820 [06:53<22:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 651/2820 [06:53<22:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 652/2820 [06:53<22:54,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 652/2820 [06:53<22:54,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 653/2820 [06:55<22:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 653/2820 [06:55<22:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 654/2820 [06:55<22:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 654/2820 [06:55<22:55,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 655/2820 [06:55<22:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 655/2820 [06:55<22:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 656/2820 [06:56<22:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 656/2820 [06:56<22:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 657/2820 [06:57<22:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 657/2820 [06:57<22:52,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 658/2820 [06:57<22:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 658/2820 [06:57<22:50,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 659/2820 [06:59<22:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 659/2820 [06:59<22:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 660/2820 [06:59<22:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 660/2820 [06:59<22:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 661/2820 [06:59<22:49,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 661/2820 [06:59<22:49,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 662/2820 [07:01<22:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  23%|██▎       | 662/2820 [07:01<22:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 663/2820 [07:01<22:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 663/2820 [07:01<22:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 664/2820 [07:01<22:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 664/2820 [07:01<22:47,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 665/2820 [07:02<22:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 665/2820 [07:02<22:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 666/2820 [07:02<22:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 666/2820 [07:02<22:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 667/2820 [07:02<22:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 667/2820 [07:02<22:45,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 668/2820 [07:04<22:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 668/2820 [07:04<22:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 669/2820 [07:04<22:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▎       | 669/2820 [07:04<22:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 670/2820 [07:04<22:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 670/2820 [07:04<22:43,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 671/2820 [07:06<22:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 671/2820 [07:06<22:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 672/2820 [07:06<22:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 672/2820 [07:06<22:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 673/2820 [07:06<22:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 673/2820 [07:06<22:42,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 674/2820 [07:08<22:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 674/2820 [07:08<22:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 675/2820 [07:08<22:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 675/2820 [07:08<22:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 676/2820 [07:08<22:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 676/2820 [07:08<22:39,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 677/2820 [07:10<22:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 677/2820 [07:10<22:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 678/2820 [07:10<22:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 678/2820 [07:10<22:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 679/2820 [07:10<22:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 679/2820 [07:10<22:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 680/2820 [07:12<22:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 680/2820 [07:12<22:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 681/2820 [07:12<22:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 681/2820 [07:12<22:37,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 682/2820 [07:12<22:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 682/2820 [07:12<22:35,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 683/2820 [07:14<22:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 683/2820 [07:14<22:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 684/2820 [07:14<22:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 684/2820 [07:14<22:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 685/2820 [07:14<22:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 685/2820 [07:14<22:34,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 686/2820 [07:16<22:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 686/2820 [07:16<22:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 687/2820 [07:16<22:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 687/2820 [07:16<22:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 688/2820 [07:16<22:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 688/2820 [07:16<22:32,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 689/2820 [07:18<22:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 689/2820 [07:18<22:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 690/2820 [07:18<22:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  24%|██▍       | 690/2820 [07:18<22:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 691/2820 [07:18<22:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 691/2820 [07:18<22:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 692/2820 [07:19<22:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 692/2820 [07:19<22:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 693/2820 [07:19<22:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 693/2820 [07:19<22:30,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 694/2820 [07:20<22:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 694/2820 [07:20<22:27,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 695/2820 [07:21<22:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 695/2820 [07:21<22:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 696/2820 [07:21<22:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 696/2820 [07:21<22:28,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 697/2820 [07:21<22:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 697/2820 [07:21<22:25,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 698/2820 [07:23<22:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 698/2820 [07:23<22:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 699/2820 [07:23<22:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 699/2820 [07:23<22:26,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 700/2820 [07:23<22:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 700/2820 [07:23<22:23,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 701/2820 [07:25<22:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 701/2820 [07:25<22:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 702/2820 [07:25<22:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 702/2820 [07:25<22:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 703/2820 [07:25<22:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 703/2820 [07:25<22:22,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 704/2820 [07:27<22:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▍       | 704/2820 [07:27<22:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 705/2820 [07:27<22:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 705/2820 [07:27<22:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 706/2820 [07:27<22:20,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 706/2820 [07:27<22:20,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 707/2820 [07:29<22:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 707/2820 [07:29<22:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 708/2820 [07:29<22:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 708/2820 [07:29<22:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 709/2820 [07:29<22:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 709/2820 [07:29<22:19,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 710/2820 [07:31<22:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 710/2820 [07:31<22:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 711/2820 [07:31<22:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 711/2820 [07:31<22:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 712/2820 [07:31<22:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 712/2820 [07:31<22:17,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 713/2820 [07:33<22:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 713/2820 [07:33<22:20,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 714/2820 [07:33<22:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 714/2820 [07:33<22:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 715/2820 [07:33<22:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 715/2820 [07:33<22:15,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 716/2820 [07:35<22:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 716/2820 [07:35<22:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 717/2820 [07:35<22:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 717/2820 [07:35<22:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 718/2820 [07:35<22:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 718/2820 [07:35<22:13,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 719/2820 [07:37<22:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  25%|██▌       | 719/2820 [07:37<22:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 720/2820 [07:37<22:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 720/2820 [07:37<22:14,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 721/2820 [07:37<22:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 721/2820 [07:37<22:12,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 722/2820 [07:39<22:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 722/2820 [07:39<22:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 723/2820 [07:39<22:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 723/2820 [07:39<22:12,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 724/2820 [07:39<22:10,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 724/2820 [07:39<22:10,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 725/2820 [07:41<22:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 725/2820 [07:41<22:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 726/2820 [07:41<22:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 726/2820 [07:41<22:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 727/2820 [07:41<22:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 727/2820 [07:41<22:08,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 728/2820 [07:43<22:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 728/2820 [07:43<22:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 729/2820 [07:43<22:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 729/2820 [07:43<22:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 730/2820 [07:43<22:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 730/2820 [07:43<22:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 731/2820 [07:45<22:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 731/2820 [07:45<22:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 732/2820 [07:45<22:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 732/2820 [07:45<22:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 733/2820 [07:45<22:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 733/2820 [07:45<22:04,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 734/2820 [07:47<22:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 734/2820 [07:47<22:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 735/2820 [07:47<22:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 735/2820 [07:47<22:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 736/2820 [07:47<22:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 736/2820 [07:47<22:02,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 737/2820 [07:48<22:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 737/2820 [07:48<22:05,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 738/2820 [07:48<22:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 738/2820 [07:48<22:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 739/2820 [07:49<22:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 739/2820 [07:49<22:00,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 740/2820 [07:50<22:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▌       | 740/2820 [07:50<22:03,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 741/2820 [07:50<22:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 741/2820 [07:50<22:01,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 742/2820 [07:50<21:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 742/2820 [07:50<21:58,  1.58it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 743/2820 [07:53<22:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 743/2820 [07:53<22:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 744/2820 [07:53<22:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 744/2820 [07:53<22:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 745/2820 [07:53<21:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 745/2820 [07:53<21:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 746/2820 [07:55<22:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 746/2820 [07:55<22:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 747/2820 [07:55<21:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  26%|██▋       | 747/2820 [07:55<21:58,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 748/2820 [07:55<21:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 748/2820 [07:55<21:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 749/2820 [07:57<21:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 749/2820 [07:57<21:59,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 750/2820 [07:57<21:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 750/2820 [07:57<21:56,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 751/2820 [07:57<21:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 751/2820 [07:57<21:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 752/2820 [07:58<21:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 752/2820 [07:58<21:57,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 753/2820 [07:58<21:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 753/2820 [07:58<21:54,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 754/2820 [07:59<21:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 754/2820 [07:59<21:52,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 755/2820 [08:00<21:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 755/2820 [08:00<21:55,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 756/2820 [08:00<21:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 756/2820 [08:00<21:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 757/2820 [08:00<21:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 757/2820 [08:00<21:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 758/2820 [08:02<21:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 758/2820 [08:02<21:53,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 759/2820 [08:03<21:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 759/2820 [08:03<21:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 760/2820 [08:03<21:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 760/2820 [08:03<21:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 761/2820 [08:04<21:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 761/2820 [08:04<21:51,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 762/2820 [08:04<21:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 762/2820 [08:04<21:49,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 763/2820 [08:04<21:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 763/2820 [08:04<21:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 764/2820 [08:06<21:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 764/2820 [08:06<21:50,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 765/2820 [08:06<21:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 765/2820 [08:06<21:47,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 766/2820 [08:06<21:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 766/2820 [08:06<21:45,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 767/2820 [08:08<21:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 767/2820 [08:08<21:48,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 768/2820 [08:08<21:46,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 768/2820 [08:08<21:46,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 769/2820 [08:08<21:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 769/2820 [08:08<21:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 770/2820 [08:10<21:46,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 770/2820 [08:10<21:46,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 771/2820 [08:10<21:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 771/2820 [08:10<21:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 772/2820 [08:10<21:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 772/2820 [08:10<21:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 773/2820 [08:12<21:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 773/2820 [08:12<21:44,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 774/2820 [08:12<21:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 774/2820 [08:12<21:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 775/2820 [08:12<21:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  27%|██▋       | 775/2820 [08:12<21:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 776/2820 [08:14<21:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 776/2820 [08:14<21:43,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 777/2820 [08:14<21:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 777/2820 [08:14<21:41,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 778/2820 [08:14<21:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 778/2820 [08:14<21:38,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 779/2820 [08:16<21:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 779/2820 [08:16<21:42,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 780/2820 [08:16<21:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 780/2820 [08:16<21:39,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 781/2820 [08:17<21:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 781/2820 [08:17<21:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 782/2820 [08:18<21:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 782/2820 [08:18<21:40,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 783/2820 [08:18<21:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 783/2820 [08:18<21:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 784/2820 [08:18<21:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 784/2820 [08:18<21:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 785/2820 [08:20<21:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 785/2820 [08:20<21:37,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 786/2820 [08:20<21:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 786/2820 [08:20<21:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 787/2820 [08:20<21:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 787/2820 [08:20<21:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 788/2820 [08:22<21:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 788/2820 [08:22<21:36,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 789/2820 [08:22<21:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 789/2820 [08:22<21:34,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 790/2820 [08:22<21:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 790/2820 [08:22<21:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 791/2820 [08:24<21:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 791/2820 [08:24<21:35,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 792/2820 [08:24<21:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 792/2820 [08:24<21:32,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 793/2820 [08:24<21:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 793/2820 [08:24<21:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 794/2820 [08:26<21:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 794/2820 [08:26<21:33,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 795/2820 [08:27<21:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 795/2820 [08:27<21:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 796/2820 [08:27<21:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 796/2820 [08:27<21:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 797/2820 [08:28<21:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 797/2820 [08:28<21:31,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 798/2820 [08:28<21:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 798/2820 [08:28<21:29,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 799/2820 [08:29<21:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 799/2820 [08:29<21:27,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 800/2820 [08:30<21:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 800/2820 [08:30<21:30,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 801/2820 [08:31<21:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 801/2820 [08:31<21:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 802/2820 [08:31<21:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 802/2820 [08:31<21:25,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 803/2820 [08:32<21:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  28%|██▊       | 803/2820 [08:32<21:28,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 804/2820 [08:32<21:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 804/2820 [08:32<21:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 805/2820 [08:32<21:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 805/2820 [08:32<21:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 806/2820 [08:34<21:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 806/2820 [08:34<21:26,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 807/2820 [08:34<21:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 807/2820 [08:34<21:24,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 808/2820 [08:35<21:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 808/2820 [08:35<21:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 809/2820 [08:37<21:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 809/2820 [08:37<21:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 810/2820 [08:37<21:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▊       | 810/2820 [08:37<21:23,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 811/2820 [08:37<21:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 811/2820 [08:37<21:21,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 812/2820 [08:39<21:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 812/2820 [08:39<21:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 813/2820 [08:39<21:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 813/2820 [08:39<21:22,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 814/2820 [08:39<21:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 814/2820 [08:39<21:19,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 815/2820 [08:41<21:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 815/2820 [08:41<21:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 816/2820 [08:41<21:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 816/2820 [08:41<21:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 817/2820 [08:41<21:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 817/2820 [08:41<21:18,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 818/2820 [08:43<21:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 818/2820 [08:43<21:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 819/2820 [08:43<21:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 819/2820 [08:43<21:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 820/2820 [08:43<21:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 820/2820 [08:43<21:16,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 821/2820 [08:45<21:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 821/2820 [08:45<21:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 822/2820 [08:45<21:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 822/2820 [08:45<21:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 823/2820 [08:45<21:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 823/2820 [08:45<21:15,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 824/2820 [08:47<21:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 824/2820 [08:47<21:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 825/2820 [08:47<21:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 825/2820 [08:47<21:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 826/2820 [08:47<21:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 826/2820 [08:47<21:13,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 827/2820 [08:49<21:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 827/2820 [08:49<21:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 828/2820 [08:49<21:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 828/2820 [08:49<21:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 829/2820 [08:49<21:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 829/2820 [08:49<21:11,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 830/2820 [08:51<21:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 830/2820 [08:51<21:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 831/2820 [08:51<21:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  29%|██▉       | 831/2820 [08:51<21:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 832/2820 [08:51<21:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 832/2820 [08:51<21:09,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 833/2820 [08:53<21:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 833/2820 [08:53<21:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 834/2820 [08:53<21:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 834/2820 [08:53<21:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 835/2820 [08:53<21:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 835/2820 [08:53<21:07,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 836/2820 [08:55<21:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 836/2820 [08:55<21:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 837/2820 [08:55<21:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 837/2820 [08:55<21:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 838/2820 [08:55<21:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 838/2820 [08:55<21:06,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 839/2820 [08:57<21:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 839/2820 [08:57<21:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 840/2820 [08:57<21:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 840/2820 [08:57<21:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 841/2820 [08:57<21:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 841/2820 [08:57<21:04,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 842/2820 [08:59<21:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 842/2820 [08:59<21:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 843/2820 [08:59<21:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 843/2820 [08:59<21:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 844/2820 [08:59<21:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 844/2820 [08:59<21:02,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 845/2820 [09:01<21:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|██▉       | 845/2820 [09:01<21:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 846/2820 [09:01<21:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 846/2820 [09:01<21:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 847/2820 [09:01<21:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 847/2820 [09:01<21:00,  1.57it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 848/2820 [09:03<21:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 848/2820 [09:03<21:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 849/2820 [09:03<21:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 849/2820 [09:03<21:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 850/2820 [09:03<20:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 850/2820 [09:03<20:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 851/2820 [09:05<21:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 851/2820 [09:05<21:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 852/2820 [09:05<20:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 852/2820 [09:05<20:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 853/2820 [09:05<20:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 853/2820 [09:05<20:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 854/2820 [09:07<20:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 854/2820 [09:07<20:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 855/2820 [09:07<20:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 855/2820 [09:07<20:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 856/2820 [09:07<20:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 856/2820 [09:07<20:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 857/2820 [09:09<20:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 857/2820 [09:09<20:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 858/2820 [09:09<20:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 858/2820 [09:09<20:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 859/2820 [09:09<20:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 859/2820 [09:09<20:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 860/2820 [09:11<20:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  30%|███       | 860/2820 [09:11<20:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 861/2820 [09:11<20:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 861/2820 [09:11<20:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 862/2820 [09:11<20:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 862/2820 [09:11<20:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 863/2820 [09:13<20:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 863/2820 [09:13<20:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 864/2820 [09:13<20:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 864/2820 [09:13<20:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 865/2820 [09:13<20:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 865/2820 [09:13<20:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 866/2820 [09:15<20:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 866/2820 [09:15<20:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 867/2820 [09:15<20:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 867/2820 [09:15<20:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 868/2820 [09:15<20:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 868/2820 [09:15<20:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 869/2820 [09:17<20:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 869/2820 [09:17<20:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 870/2820 [09:17<20:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 870/2820 [09:17<20:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 871/2820 [09:17<20:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 871/2820 [09:17<20:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 872/2820 [09:19<20:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 872/2820 [09:19<20:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 873/2820 [09:19<20:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 873/2820 [09:19<20:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 874/2820 [09:19<20:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 874/2820 [09:19<20:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 875/2820 [09:21<20:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 875/2820 [09:21<20:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 876/2820 [09:21<20:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 876/2820 [09:21<20:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 877/2820 [09:21<20:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 877/2820 [09:21<20:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 878/2820 [09:23<20:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 878/2820 [09:23<20:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 879/2820 [09:23<20:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 879/2820 [09:23<20:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 880/2820 [09:23<20:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 880/2820 [09:23<20:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 881/2820 [09:25<20:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███       | 881/2820 [09:25<20:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 882/2820 [09:25<20:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 882/2820 [09:25<20:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 883/2820 [09:25<20:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 883/2820 [09:25<20:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 884/2820 [09:27<20:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 884/2820 [09:27<20:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 885/2820 [09:27<20:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 885/2820 [09:27<20:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 886/2820 [09:27<20:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 886/2820 [09:27<20:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 887/2820 [09:29<20:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 887/2820 [09:29<20:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 888/2820 [09:29<20:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  31%|███▏      | 888/2820 [09:29<20:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 889/2820 [09:29<20:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 889/2820 [09:29<20:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 890/2820 [09:31<20:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 890/2820 [09:31<20:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 891/2820 [09:31<20:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 891/2820 [09:31<20:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 892/2820 [09:31<20:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 892/2820 [09:31<20:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 893/2820 [09:33<20:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 893/2820 [09:33<20:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 894/2820 [09:33<20:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 894/2820 [09:33<20:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 895/2820 [09:33<20:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 895/2820 [09:33<20:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 896/2820 [09:35<20:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 896/2820 [09:35<20:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 897/2820 [09:35<20:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 897/2820 [09:35<20:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 898/2820 [09:35<20:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 898/2820 [09:35<20:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 899/2820 [09:37<20:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 899/2820 [09:37<20:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 900/2820 [09:37<20:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 900/2820 [09:37<20:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 901/2820 [09:37<20:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 901/2820 [09:37<20:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 902/2820 [09:39<20:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 902/2820 [09:39<20:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 903/2820 [09:39<20:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 903/2820 [09:39<20:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 904/2820 [09:39<20:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 904/2820 [09:39<20:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 905/2820 [09:41<20:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 905/2820 [09:41<20:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 906/2820 [09:41<20:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 906/2820 [09:41<20:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 907/2820 [09:41<20:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 907/2820 [09:41<20:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 908/2820 [09:43<20:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 908/2820 [09:43<20:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 909/2820 [09:43<20:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 909/2820 [09:43<20:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 910/2820 [09:43<20:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 910/2820 [09:43<20:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 911/2820 [09:45<20:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 911/2820 [09:45<20:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 912/2820 [09:45<20:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 912/2820 [09:45<20:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 913/2820 [09:45<20:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 913/2820 [09:45<20:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 914/2820 [09:47<20:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 914/2820 [09:47<20:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 915/2820 [09:47<20:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 915/2820 [09:47<20:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 916/2820 [09:47<20:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  32%|███▏      | 916/2820 [09:47<20:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 917/2820 [09:49<20:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 917/2820 [09:49<20:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 918/2820 [09:49<20:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 918/2820 [09:49<20:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 919/2820 [09:49<20:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 919/2820 [09:49<20:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 920/2820 [09:51<20:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 920/2820 [09:51<20:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 921/2820 [09:51<20:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 921/2820 [09:51<20:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 922/2820 [09:51<20:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 922/2820 [09:51<20:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 923/2820 [09:53<20:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 923/2820 [09:53<20:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 924/2820 [09:53<20:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 924/2820 [09:53<20:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 925/2820 [09:53<20:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 925/2820 [09:53<20:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 926/2820 [09:55<20:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 926/2820 [09:55<20:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 927/2820 [09:55<20:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 927/2820 [09:55<20:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 928/2820 [09:55<20:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 928/2820 [09:55<20:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 929/2820 [09:57<20:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 929/2820 [09:57<20:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 930/2820 [09:57<20:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 930/2820 [09:57<20:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 931/2820 [09:57<20:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 931/2820 [09:57<20:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 932/2820 [09:59<20:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 932/2820 [09:59<20:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 933/2820 [09:59<20:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 933/2820 [09:59<20:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 934/2820 [09:59<20:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 934/2820 [09:59<20:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 935/2820 [10:01<20:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 935/2820 [10:01<20:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 936/2820 [10:01<20:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 936/2820 [10:01<20:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 937/2820 [10:01<20:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 937/2820 [10:01<20:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 938/2820 [10:03<20:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 938/2820 [10:03<20:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 939/2820 [10:03<20:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 939/2820 [10:03<20:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 940/2820 [10:03<20:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 940/2820 [10:03<20:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 941/2820 [10:05<20:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 941/2820 [10:05<20:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 942/2820 [10:05<20:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 942/2820 [10:05<20:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 943/2820 [10:05<20:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 943/2820 [10:05<20:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 944/2820 [10:07<20:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  33%|███▎      | 944/2820 [10:07<20:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 945/2820 [10:07<20:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 945/2820 [10:07<20:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 946/2820 [10:07<20:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 946/2820 [10:07<20:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 947/2820 [10:09<20:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 947/2820 [10:09<20:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 948/2820 [10:09<20:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 948/2820 [10:09<20:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 949/2820 [10:09<20:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 949/2820 [10:09<20:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 950/2820 [10:11<20:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 950/2820 [10:11<20:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 951/2820 [10:11<20:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▎      | 951/2820 [10:11<20:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 952/2820 [10:11<19:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 952/2820 [10:11<19:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 953/2820 [10:13<20:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 953/2820 [10:13<20:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 954/2820 [10:13<19:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 954/2820 [10:13<19:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 955/2820 [10:13<19:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 955/2820 [10:13<19:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 956/2820 [10:15<19:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 956/2820 [10:15<19:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 957/2820 [10:15<19:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 957/2820 [10:15<19:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 958/2820 [10:15<19:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 958/2820 [10:15<19:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 959/2820 [10:16<19:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 959/2820 [10:16<19:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 960/2820 [10:16<19:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 960/2820 [10:16<19:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 961/2820 [10:16<19:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 961/2820 [10:16<19:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 962/2820 [10:18<19:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 962/2820 [10:18<19:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 963/2820 [10:18<19:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 963/2820 [10:18<19:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 964/2820 [10:18<19:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 964/2820 [10:18<19:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 965/2820 [10:20<19:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 965/2820 [10:20<19:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 966/2820 [10:20<19:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 966/2820 [10:20<19:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 967/2820 [10:20<19:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 967/2820 [10:20<19:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 968/2820 [10:22<19:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 968/2820 [10:22<19:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 969/2820 [10:22<19:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 969/2820 [10:22<19:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 970/2820 [10:22<19:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 970/2820 [10:22<19:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 971/2820 [10:24<19:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 971/2820 [10:24<19:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 972/2820 [10:24<19:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  34%|███▍      | 972/2820 [10:24<19:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 973/2820 [10:24<19:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 973/2820 [10:24<19:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 974/2820 [10:26<19:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 974/2820 [10:26<19:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 975/2820 [10:26<19:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 975/2820 [10:26<19:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 976/2820 [10:26<19:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 976/2820 [10:26<19:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 977/2820 [10:28<19:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 977/2820 [10:28<19:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 978/2820 [10:28<19:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 978/2820 [10:28<19:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 979/2820 [10:28<19:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 979/2820 [10:28<19:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 980/2820 [10:30<19:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 980/2820 [10:30<19:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 981/2820 [10:30<19:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 981/2820 [10:30<19:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 982/2820 [10:30<19:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 982/2820 [10:30<19:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 983/2820 [10:32<19:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 983/2820 [10:32<19:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 984/2820 [10:32<19:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 984/2820 [10:32<19:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 985/2820 [10:32<19:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 985/2820 [10:32<19:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 986/2820 [10:34<19:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▍      | 986/2820 [10:34<19:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 987/2820 [10:34<19:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 987/2820 [10:34<19:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 988/2820 [10:34<19:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 988/2820 [10:34<19:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 989/2820 [10:36<19:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 989/2820 [10:36<19:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 990/2820 [10:36<19:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 990/2820 [10:36<19:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 991/2820 [10:36<19:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 991/2820 [10:36<19:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 992/2820 [10:38<19:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 992/2820 [10:38<19:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 993/2820 [10:38<19:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 993/2820 [10:38<19:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 994/2820 [10:38<19:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 994/2820 [10:38<19:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 995/2820 [10:40<19:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 995/2820 [10:40<19:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 996/2820 [10:40<19:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 996/2820 [10:40<19:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 997/2820 [10:40<19:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 997/2820 [10:40<19:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 998/2820 [10:42<19:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 998/2820 [10:42<19:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 999/2820 [10:42<19:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 999/2820 [10:42<19:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 1000/2820 [10:42<19:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 1000/2820 [10:42<19:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 1001/2820 [10:44<19:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  35%|███▌      | 1001/2820 [10:44<19:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1002/2820 [10:44<19:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1002/2820 [10:44<19:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1003/2820 [10:44<19:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1003/2820 [10:44<19:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1004/2820 [10:46<19:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1004/2820 [10:46<19:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1005/2820 [10:46<19:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1005/2820 [10:46<19:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1006/2820 [10:46<19:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1006/2820 [10:46<19:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1007/2820 [10:48<19:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1007/2820 [10:48<19:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1008/2820 [10:48<19:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1008/2820 [10:48<19:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1009/2820 [10:48<19:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1009/2820 [10:48<19:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1010/2820 [10:50<19:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1010/2820 [10:50<19:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1011/2820 [10:50<19:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1011/2820 [10:50<19:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1012/2820 [10:50<19:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1012/2820 [10:50<19:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1013/2820 [10:51<19:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1013/2820 [10:51<19:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1014/2820 [10:51<19:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1014/2820 [10:51<19:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1015/2820 [10:51<19:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1015/2820 [10:51<19:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1016/2820 [10:53<19:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1016/2820 [10:53<19:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1017/2820 [10:53<19:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1017/2820 [10:53<19:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1018/2820 [10:53<19:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1018/2820 [10:53<19:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1019/2820 [10:55<19:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1019/2820 [10:55<19:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1020/2820 [10:55<19:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1020/2820 [10:55<19:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1021/2820 [10:55<19:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1021/2820 [10:55<19:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1022/2820 [10:57<19:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▌      | 1022/2820 [10:57<19:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1023/2820 [10:57<19:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1023/2820 [10:57<19:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1024/2820 [10:57<19:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1024/2820 [10:57<19:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1025/2820 [10:59<19:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1025/2820 [10:59<19:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1026/2820 [10:59<19:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1026/2820 [10:59<19:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1027/2820 [10:59<19:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1027/2820 [10:59<19:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1028/2820 [11:01<19:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1028/2820 [11:01<19:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1029/2820 [11:01<19:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  36%|███▋      | 1029/2820 [11:01<19:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1030/2820 [11:01<19:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1030/2820 [11:01<19:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1031/2820 [11:03<19:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1031/2820 [11:03<19:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1032/2820 [11:03<19:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1032/2820 [11:03<19:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1033/2820 [11:03<19:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1033/2820 [11:03<19:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1034/2820 [11:04<19:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1034/2820 [11:04<19:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1035/2820 [11:05<19:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1035/2820 [11:05<19:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1036/2820 [11:05<19:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1036/2820 [11:05<19:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1037/2820 [11:07<19:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1037/2820 [11:07<19:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1038/2820 [11:07<19:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1038/2820 [11:07<19:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1039/2820 [11:07<19:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1039/2820 [11:07<19:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1040/2820 [11:08<19:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1040/2820 [11:08<19:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1041/2820 [11:08<19:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1041/2820 [11:08<19:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1042/2820 [11:09<19:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1042/2820 [11:09<19:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1043/2820 [11:10<19:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1043/2820 [11:10<19:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1044/2820 [11:10<19:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1044/2820 [11:10<19:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1045/2820 [11:10<18:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1045/2820 [11:10<18:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1046/2820 [11:12<19:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1046/2820 [11:12<19:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1047/2820 [11:12<18:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1047/2820 [11:12<18:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1048/2820 [11:12<18:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1048/2820 [11:12<18:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1049/2820 [11:14<18:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1049/2820 [11:14<18:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1050/2820 [11:14<18:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1050/2820 [11:14<18:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1051/2820 [11:14<18:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1051/2820 [11:14<18:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1052/2820 [11:16<18:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1052/2820 [11:16<18:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1053/2820 [11:16<18:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1053/2820 [11:16<18:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1054/2820 [11:16<18:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1054/2820 [11:16<18:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1055/2820 [11:18<18:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1055/2820 [11:18<18:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1056/2820 [11:18<18:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1056/2820 [11:18<18:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1057/2820 [11:18<18:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  37%|███▋      | 1057/2820 [11:18<18:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1058/2820 [11:20<18:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1058/2820 [11:20<18:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1059/2820 [11:20<18:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1059/2820 [11:20<18:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1060/2820 [11:20<18:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1060/2820 [11:20<18:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1061/2820 [11:23<18:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1061/2820 [11:23<18:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1062/2820 [11:23<18:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1062/2820 [11:23<18:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1063/2820 [11:23<18:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1063/2820 [11:23<18:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1064/2820 [11:25<18:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1064/2820 [11:25<18:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1065/2820 [11:25<18:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1065/2820 [11:25<18:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1066/2820 [11:25<18:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1066/2820 [11:25<18:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1067/2820 [11:26<18:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1067/2820 [11:26<18:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1068/2820 [11:27<18:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1068/2820 [11:27<18:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1069/2820 [11:27<18:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1069/2820 [11:27<18:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1070/2820 [11:28<18:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1070/2820 [11:28<18:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1071/2820 [11:28<18:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1071/2820 [11:28<18:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1072/2820 [11:28<18:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1072/2820 [11:28<18:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1073/2820 [11:30<18:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1073/2820 [11:30<18:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1074/2820 [11:30<18:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1074/2820 [11:30<18:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1075/2820 [11:30<18:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1075/2820 [11:30<18:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1076/2820 [11:32<18:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1076/2820 [11:32<18:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1077/2820 [11:32<18:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1077/2820 [11:32<18:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1078/2820 [11:32<18:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1078/2820 [11:32<18:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1079/2820 [11:34<18:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1079/2820 [11:34<18:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1080/2820 [11:34<18:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1080/2820 [11:34<18:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1081/2820 [11:34<18:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1081/2820 [11:34<18:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1082/2820 [11:36<18:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1082/2820 [11:36<18:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1083/2820 [11:36<18:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1083/2820 [11:36<18:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1084/2820 [11:36<18:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1084/2820 [11:36<18:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1085/2820 [11:38<18:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  38%|███▊      | 1085/2820 [11:38<18:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1086/2820 [11:38<18:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1086/2820 [11:38<18:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1087/2820 [11:38<18:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1087/2820 [11:38<18:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1088/2820 [11:40<18:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1088/2820 [11:40<18:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1089/2820 [11:40<18:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1089/2820 [11:40<18:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1090/2820 [11:40<18:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1090/2820 [11:40<18:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1091/2820 [11:42<18:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1091/2820 [11:42<18:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1092/2820 [11:42<18:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▊      | 1092/2820 [11:42<18:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1093/2820 [11:42<18:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1093/2820 [11:42<18:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1094/2820 [11:44<18:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1094/2820 [11:44<18:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1095/2820 [11:44<18:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1095/2820 [11:44<18:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1096/2820 [11:44<18:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1096/2820 [11:44<18:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1097/2820 [11:46<18:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1097/2820 [11:46<18:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1098/2820 [11:46<18:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1098/2820 [11:46<18:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1099/2820 [11:46<18:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1099/2820 [11:46<18:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1100/2820 [11:48<18:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1100/2820 [11:48<18:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1101/2820 [11:48<18:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1101/2820 [11:48<18:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1102/2820 [11:48<18:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1102/2820 [11:48<18:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1103/2820 [11:49<18:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1103/2820 [11:49<18:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1104/2820 [11:49<18:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1104/2820 [11:49<18:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1105/2820 [11:49<18:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1105/2820 [11:49<18:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1106/2820 [11:51<18:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1106/2820 [11:51<18:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1107/2820 [11:51<18:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1107/2820 [11:51<18:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1108/2820 [11:51<18:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1108/2820 [11:51<18:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1109/2820 [11:53<18:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1109/2820 [11:53<18:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1110/2820 [11:53<18:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1110/2820 [11:53<18:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1111/2820 [11:53<18:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1111/2820 [11:53<18:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1112/2820 [11:55<18:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1112/2820 [11:55<18:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1113/2820 [11:55<18:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  39%|███▉      | 1113/2820 [11:55<18:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1114/2820 [11:55<18:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1114/2820 [11:55<18:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1115/2820 [11:57<18:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1115/2820 [11:57<18:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1116/2820 [11:57<18:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1116/2820 [11:57<18:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1117/2820 [11:57<18:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1117/2820 [11:57<18:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1118/2820 [11:59<18:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1118/2820 [11:59<18:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1119/2820 [11:59<18:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1119/2820 [11:59<18:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1120/2820 [11:59<18:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1120/2820 [11:59<18:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1121/2820 [12:00<18:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1121/2820 [12:00<18:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1122/2820 [12:00<18:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1122/2820 [12:00<18:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1123/2820 [12:00<18:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1123/2820 [12:00<18:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1124/2820 [12:02<18:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1124/2820 [12:02<18:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1125/2820 [12:02<18:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1125/2820 [12:02<18:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1126/2820 [12:02<18:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1126/2820 [12:02<18:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1127/2820 [12:04<18:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|███▉      | 1127/2820 [12:04<18:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1128/2820 [12:04<18:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1128/2820 [12:04<18:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1129/2820 [12:04<18:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1129/2820 [12:04<18:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1130/2820 [12:06<18:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1130/2820 [12:06<18:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1131/2820 [12:06<18:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1131/2820 [12:06<18:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1132/2820 [12:06<18:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1132/2820 [12:06<18:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1133/2820 [12:08<18:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1133/2820 [12:08<18:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1134/2820 [12:08<18:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1134/2820 [12:08<18:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1135/2820 [12:08<18:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1135/2820 [12:08<18:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1136/2820 [12:10<18:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1136/2820 [12:10<18:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1137/2820 [12:10<18:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1137/2820 [12:10<18:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1138/2820 [12:10<17:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1138/2820 [12:10<17:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1139/2820 [12:12<18:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1139/2820 [12:12<18:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1140/2820 [12:12<17:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1140/2820 [12:12<17:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1141/2820 [12:12<17:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1141/2820 [12:12<17:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1142/2820 [12:14<17:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  40%|████      | 1142/2820 [12:14<17:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1143/2820 [12:14<17:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1143/2820 [12:14<17:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1144/2820 [12:14<17:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1144/2820 [12:14<17:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1145/2820 [12:16<17:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1145/2820 [12:16<17:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1146/2820 [12:16<17:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1146/2820 [12:16<17:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1147/2820 [12:16<17:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1147/2820 [12:16<17:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1148/2820 [12:18<17:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1148/2820 [12:18<17:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1149/2820 [12:18<17:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1149/2820 [12:18<17:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1150/2820 [12:18<17:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1150/2820 [12:18<17:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1151/2820 [12:20<17:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1151/2820 [12:20<17:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1152/2820 [12:20<17:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1152/2820 [12:20<17:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1153/2820 [12:20<17:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1153/2820 [12:20<17:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1154/2820 [12:22<17:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1154/2820 [12:22<17:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1155/2820 [12:22<17:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1155/2820 [12:22<17:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1156/2820 [12:22<17:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1156/2820 [12:22<17:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1157/2820 [12:24<17:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1157/2820 [12:24<17:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1158/2820 [12:24<17:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1158/2820 [12:24<17:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1159/2820 [12:24<17:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1159/2820 [12:24<17:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1160/2820 [12:26<17:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1160/2820 [12:26<17:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1161/2820 [12:26<17:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1161/2820 [12:26<17:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1162/2820 [12:26<17:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1162/2820 [12:26<17:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1163/2820 [12:28<17:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████      | 1163/2820 [12:28<17:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1164/2820 [12:28<17:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1164/2820 [12:28<17:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1165/2820 [12:28<17:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1165/2820 [12:28<17:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1166/2820 [12:30<17:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1166/2820 [12:30<17:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1167/2820 [12:30<17:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1167/2820 [12:30<17:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1168/2820 [12:30<17:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1168/2820 [12:30<17:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1169/2820 [12:31<17:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1169/2820 [12:31<17:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1170/2820 [12:31<17:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  41%|████▏     | 1170/2820 [12:31<17:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1171/2820 [12:31<17:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1171/2820 [12:31<17:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1172/2820 [12:33<17:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1172/2820 [12:33<17:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1173/2820 [12:33<17:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1173/2820 [12:33<17:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1174/2820 [12:33<17:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1174/2820 [12:33<17:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1175/2820 [12:35<17:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1175/2820 [12:35<17:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1176/2820 [12:35<17:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1176/2820 [12:35<17:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1177/2820 [12:35<17:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1177/2820 [12:35<17:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1178/2820 [12:37<17:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1178/2820 [12:37<17:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1179/2820 [12:37<17:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1179/2820 [12:37<17:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1180/2820 [12:37<17:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1180/2820 [12:37<17:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1181/2820 [12:39<17:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1181/2820 [12:39<17:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1182/2820 [12:39<17:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1182/2820 [12:39<17:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1183/2820 [12:39<17:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1183/2820 [12:39<17:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1184/2820 [12:41<17:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1184/2820 [12:41<17:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1185/2820 [12:41<17:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1185/2820 [12:41<17:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1186/2820 [12:41<17:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1186/2820 [12:41<17:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1187/2820 [12:43<17:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1187/2820 [12:43<17:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1188/2820 [12:43<17:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1188/2820 [12:43<17:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1189/2820 [12:43<17:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1189/2820 [12:43<17:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1190/2820 [12:44<17:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1190/2820 [12:44<17:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1191/2820 [12:45<17:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1191/2820 [12:45<17:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1192/2820 [12:45<17:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1192/2820 [12:45<17:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1193/2820 [12:46<17:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1193/2820 [12:46<17:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1194/2820 [12:46<17:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1194/2820 [12:46<17:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1195/2820 [12:46<17:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1195/2820 [12:46<17:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1196/2820 [12:48<17:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1196/2820 [12:48<17:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1197/2820 [12:48<17:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1197/2820 [12:48<17:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1198/2820 [12:48<17:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  42%|████▏     | 1198/2820 [12:48<17:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1199/2820 [12:50<17:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1199/2820 [12:50<17:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1200/2820 [12:50<17:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1200/2820 [12:50<17:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1201/2820 [12:50<17:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1201/2820 [12:50<17:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1202/2820 [12:52<17:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1202/2820 [12:52<17:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1203/2820 [12:52<17:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1203/2820 [12:52<17:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1204/2820 [12:52<17:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1204/2820 [12:52<17:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1205/2820 [12:54<17:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1205/2820 [12:54<17:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1206/2820 [12:54<17:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1206/2820 [12:54<17:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1207/2820 [12:54<17:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1207/2820 [12:54<17:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1208/2820 [12:56<17:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1208/2820 [12:56<17:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1209/2820 [12:56<17:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1209/2820 [12:56<17:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1210/2820 [12:56<17:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1210/2820 [12:56<17:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1211/2820 [12:58<17:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1211/2820 [12:58<17:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1212/2820 [12:58<17:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1212/2820 [12:58<17:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1213/2820 [12:58<17:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1213/2820 [12:58<17:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1214/2820 [13:00<17:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1214/2820 [13:00<17:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1215/2820 [13:00<17:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1215/2820 [13:00<17:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1216/2820 [13:00<17:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1216/2820 [13:00<17:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1217/2820 [13:01<17:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1217/2820 [13:01<17:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1218/2820 [13:01<17:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1218/2820 [13:01<17:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1219/2820 [13:01<17:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1219/2820 [13:01<17:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1220/2820 [13:03<17:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1220/2820 [13:03<17:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1221/2820 [13:03<17:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1221/2820 [13:03<17:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1222/2820 [13:03<17:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1222/2820 [13:03<17:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1223/2820 [13:05<17:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1223/2820 [13:05<17:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1224/2820 [13:05<17:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1224/2820 [13:05<17:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1225/2820 [13:05<17:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1225/2820 [13:05<17:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1226/2820 [13:07<17:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  43%|████▎     | 1226/2820 [13:07<17:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1227/2820 [13:07<17:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1227/2820 [13:07<17:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1228/2820 [13:07<17:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1228/2820 [13:07<17:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1229/2820 [13:09<17:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1229/2820 [13:09<17:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1230/2820 [13:09<17:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1230/2820 [13:09<17:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1231/2820 [13:09<16:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1231/2820 [13:09<16:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1232/2820 [13:11<16:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1232/2820 [13:11<16:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1233/2820 [13:11<16:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▎     | 1233/2820 [13:11<16:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1234/2820 [13:11<16:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1234/2820 [13:11<16:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1235/2820 [13:13<16:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1235/2820 [13:13<16:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1236/2820 [13:13<16:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1236/2820 [13:13<16:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1237/2820 [13:13<16:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1237/2820 [13:13<16:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1238/2820 [13:15<16:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1238/2820 [13:15<16:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1239/2820 [13:15<16:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1239/2820 [13:15<16:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1240/2820 [13:15<16:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1240/2820 [13:15<16:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1241/2820 [13:17<16:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1241/2820 [13:17<16:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1242/2820 [13:17<16:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1242/2820 [13:17<16:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1243/2820 [13:17<16:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1243/2820 [13:17<16:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1244/2820 [13:19<16:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1244/2820 [13:19<16:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1245/2820 [13:19<16:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1245/2820 [13:19<16:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1246/2820 [13:19<16:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1246/2820 [13:19<16:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1247/2820 [13:21<16:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1247/2820 [13:21<16:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1248/2820 [13:21<16:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1248/2820 [13:21<16:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1249/2820 [13:21<16:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1249/2820 [13:21<16:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1250/2820 [13:22<16:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1250/2820 [13:22<16:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1251/2820 [13:23<16:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1251/2820 [13:23<16:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1252/2820 [13:23<16:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1252/2820 [13:23<16:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1253/2820 [13:24<16:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1253/2820 [13:24<16:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1254/2820 [13:24<16:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  44%|████▍     | 1254/2820 [13:24<16:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1255/2820 [13:25<16:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1255/2820 [13:25<16:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1256/2820 [13:26<16:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1256/2820 [13:26<16:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1257/2820 [13:26<16:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1257/2820 [13:26<16:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1258/2820 [13:26<16:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1258/2820 [13:26<16:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1259/2820 [13:28<16:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1259/2820 [13:28<16:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1260/2820 [13:28<16:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1260/2820 [13:28<16:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1261/2820 [13:28<16:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1261/2820 [13:28<16:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1262/2820 [13:30<16:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1262/2820 [13:30<16:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1263/2820 [13:30<16:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1263/2820 [13:30<16:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1264/2820 [13:30<16:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1264/2820 [13:30<16:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1265/2820 [13:32<16:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1265/2820 [13:32<16:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1266/2820 [13:32<16:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1266/2820 [13:32<16:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1267/2820 [13:32<16:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1267/2820 [13:32<16:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1268/2820 [13:34<16:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▍     | 1268/2820 [13:34<16:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1269/2820 [13:34<16:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1269/2820 [13:34<16:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1270/2820 [13:34<16:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1270/2820 [13:34<16:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1271/2820 [13:36<16:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1271/2820 [13:36<16:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1272/2820 [13:36<16:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1272/2820 [13:36<16:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1273/2820 [13:36<16:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1273/2820 [13:36<16:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1274/2820 [13:38<16:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1274/2820 [13:38<16:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1275/2820 [13:38<16:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1275/2820 [13:38<16:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1276/2820 [13:38<16:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1276/2820 [13:38<16:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1277/2820 [13:40<16:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1277/2820 [13:40<16:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1278/2820 [13:40<16:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1278/2820 [13:40<16:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1279/2820 [13:40<16:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1279/2820 [13:40<16:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1280/2820 [13:41<16:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1280/2820 [13:41<16:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1281/2820 [13:41<16:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1281/2820 [13:41<16:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1282/2820 [13:42<16:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1282/2820 [13:42<16:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1283/2820 [13:43<16:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  45%|████▌     | 1283/2820 [13:43<16:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1284/2820 [13:43<16:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1284/2820 [13:43<16:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1285/2820 [13:43<16:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1285/2820 [13:43<16:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1286/2820 [13:45<16:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1286/2820 [13:45<16:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1287/2820 [13:45<16:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1287/2820 [13:45<16:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1288/2820 [13:45<16:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1288/2820 [13:45<16:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1289/2820 [13:47<16:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1289/2820 [13:47<16:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1290/2820 [13:47<16:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1290/2820 [13:47<16:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1291/2820 [13:47<16:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1291/2820 [13:47<16:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1292/2820 [13:49<16:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1292/2820 [13:49<16:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1293/2820 [13:49<16:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1293/2820 [13:49<16:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1294/2820 [13:49<16:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1294/2820 [13:49<16:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1295/2820 [13:51<16:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1295/2820 [13:51<16:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1296/2820 [13:51<16:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1296/2820 [13:51<16:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1297/2820 [13:51<16:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1297/2820 [13:51<16:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1298/2820 [13:53<16:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1298/2820 [13:53<16:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1299/2820 [13:53<16:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1299/2820 [13:53<16:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1300/2820 [13:53<16:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1300/2820 [13:53<16:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1301/2820 [13:55<16:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1301/2820 [13:55<16:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1302/2820 [13:55<16:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1302/2820 [13:55<16:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1303/2820 [13:55<16:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1303/2820 [13:55<16:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1304/2820 [13:57<16:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▌     | 1304/2820 [13:57<16:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1305/2820 [13:57<16:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1305/2820 [13:57<16:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1306/2820 [13:57<16:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1306/2820 [13:57<16:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1307/2820 [13:59<16:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1307/2820 [13:59<16:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1308/2820 [13:59<16:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1308/2820 [13:59<16:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1309/2820 [13:59<16:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1309/2820 [13:59<16:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1310/2820 [14:01<16:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1310/2820 [14:01<16:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1311/2820 [14:01<16:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  46%|████▋     | 1311/2820 [14:01<16:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1312/2820 [14:01<16:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1312/2820 [14:01<16:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1313/2820 [14:03<16:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1313/2820 [14:03<16:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1314/2820 [14:03<16:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1314/2820 [14:03<16:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1315/2820 [14:03<16:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1315/2820 [14:03<16:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1316/2820 [14:05<16:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1316/2820 [14:05<16:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1317/2820 [14:05<16:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1317/2820 [14:05<16:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1318/2820 [14:05<16:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1318/2820 [14:05<16:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1319/2820 [14:07<16:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1319/2820 [14:07<16:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1320/2820 [14:07<16:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1320/2820 [14:07<16:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1321/2820 [14:07<16:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1321/2820 [14:07<16:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1322/2820 [14:09<16:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1322/2820 [14:09<16:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1323/2820 [14:09<16:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1323/2820 [14:09<16:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1324/2820 [14:09<15:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1324/2820 [14:09<15:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1325/2820 [14:11<16:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1325/2820 [14:11<16:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1326/2820 [14:11<15:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1326/2820 [14:11<15:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1327/2820 [14:11<15:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1327/2820 [14:11<15:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1328/2820 [14:13<15:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1328/2820 [14:13<15:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1329/2820 [14:13<15:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1329/2820 [14:13<15:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1330/2820 [14:13<15:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1330/2820 [14:13<15:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1331/2820 [14:15<15:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1331/2820 [14:15<15:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1332/2820 [14:15<15:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1332/2820 [14:15<15:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1333/2820 [14:15<15:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1333/2820 [14:15<15:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1334/2820 [14:17<15:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1334/2820 [14:17<15:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1335/2820 [14:17<15:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1335/2820 [14:17<15:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1336/2820 [14:17<15:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1336/2820 [14:17<15:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1337/2820 [14:18<15:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1337/2820 [14:18<15:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1338/2820 [14:18<15:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1338/2820 [14:18<15:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1339/2820 [14:19<15:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  47%|████▋     | 1339/2820 [14:19<15:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1340/2820 [14:20<15:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1340/2820 [14:20<15:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1341/2820 [14:20<15:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1341/2820 [14:20<15:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1342/2820 [14:20<15:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1342/2820 [14:20<15:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1343/2820 [14:22<15:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1343/2820 [14:22<15:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1344/2820 [14:22<15:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1344/2820 [14:22<15:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1345/2820 [14:22<15:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1345/2820 [14:22<15:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1346/2820 [14:24<15:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1346/2820 [14:24<15:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1347/2820 [14:24<15:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1347/2820 [14:24<15:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1348/2820 [14:24<15:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1348/2820 [14:24<15:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1349/2820 [14:26<15:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1349/2820 [14:26<15:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1350/2820 [14:26<15:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1350/2820 [14:26<15:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1351/2820 [14:26<15:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1351/2820 [14:26<15:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1352/2820 [14:28<15:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1352/2820 [14:28<15:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1353/2820 [14:28<15:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1353/2820 [14:28<15:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1354/2820 [14:28<15:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1354/2820 [14:28<15:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1355/2820 [14:30<15:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1355/2820 [14:30<15:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1356/2820 [14:30<15:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1356/2820 [14:30<15:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1357/2820 [14:30<15:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1357/2820 [14:30<15:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1358/2820 [14:32<15:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1358/2820 [14:32<15:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1359/2820 [14:32<15:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1359/2820 [14:32<15:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1360/2820 [14:32<15:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1360/2820 [14:32<15:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1361/2820 [14:34<15:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1361/2820 [14:34<15:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1362/2820 [14:34<15:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1362/2820 [14:34<15:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1363/2820 [14:34<15:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1363/2820 [14:34<15:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1364/2820 [14:36<15:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1364/2820 [14:36<15:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1365/2820 [14:36<15:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1365/2820 [14:36<15:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1366/2820 [14:36<15:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1366/2820 [14:36<15:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1367/2820 [14:38<15:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  48%|████▊     | 1367/2820 [14:38<15:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1368/2820 [14:38<15:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1368/2820 [14:38<15:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1369/2820 [14:38<15:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1369/2820 [14:38<15:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1370/2820 [14:40<15:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1370/2820 [14:40<15:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1371/2820 [14:40<15:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1371/2820 [14:40<15:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1372/2820 [14:40<15:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1372/2820 [14:40<15:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1373/2820 [14:41<15:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1373/2820 [14:41<15:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1374/2820 [14:42<15:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▊     | 1374/2820 [14:42<15:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1375/2820 [14:42<15:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1375/2820 [14:42<15:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1376/2820 [14:44<15:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1376/2820 [14:44<15:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1377/2820 [14:44<15:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1377/2820 [14:44<15:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1378/2820 [14:44<15:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1378/2820 [14:44<15:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1379/2820 [14:46<15:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1379/2820 [14:46<15:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1380/2820 [14:46<15:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1380/2820 [14:46<15:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1381/2820 [14:46<15:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1381/2820 [14:46<15:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1382/2820 [14:48<15:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1382/2820 [14:48<15:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1383/2820 [14:48<15:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1383/2820 [14:48<15:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1384/2820 [14:48<15:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1384/2820 [14:48<15:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1385/2820 [14:50<15:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1385/2820 [14:50<15:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1386/2820 [14:50<15:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1386/2820 [14:50<15:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1387/2820 [14:50<15:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1387/2820 [14:50<15:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1388/2820 [14:52<15:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1388/2820 [14:52<15:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1389/2820 [14:52<15:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1389/2820 [14:52<15:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1390/2820 [14:52<15:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1390/2820 [14:52<15:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1391/2820 [14:54<15:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1391/2820 [14:54<15:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1392/2820 [14:54<15:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1392/2820 [14:54<15:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1393/2820 [14:54<15:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1393/2820 [14:54<15:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1394/2820 [14:56<15:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1394/2820 [14:56<15:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1395/2820 [14:56<15:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  49%|████▉     | 1395/2820 [14:56<15:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1396/2820 [14:56<15:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1396/2820 [14:56<15:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1397/2820 [14:58<15:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1397/2820 [14:58<15:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1398/2820 [14:58<15:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1398/2820 [14:58<15:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1399/2820 [14:58<15:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1399/2820 [14:58<15:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1400/2820 [15:00<15:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1400/2820 [15:00<15:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1401/2820 [15:00<15:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1401/2820 [15:00<15:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1402/2820 [15:00<15:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1402/2820 [15:00<15:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1403/2820 [15:02<15:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1403/2820 [15:02<15:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1404/2820 [15:02<15:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1404/2820 [15:02<15:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1405/2820 [15:02<15:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1405/2820 [15:02<15:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1406/2820 [15:03<15:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1406/2820 [15:03<15:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1407/2820 [15:03<15:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1407/2820 [15:03<15:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1408/2820 [15:03<15:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1408/2820 [15:03<15:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1409/2820 [15:05<15:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|████▉     | 1409/2820 [15:05<15:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1410/2820 [15:05<15:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1410/2820 [15:05<15:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1411/2820 [15:05<15:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1411/2820 [15:05<15:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1412/2820 [15:07<15:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1412/2820 [15:07<15:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1413/2820 [15:07<15:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1413/2820 [15:07<15:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1414/2820 [15:07<15:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1414/2820 [15:07<15:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1415/2820 [15:09<15:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1415/2820 [15:09<15:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1416/2820 [15:09<15:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1416/2820 [15:09<15:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1417/2820 [15:09<15:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1417/2820 [15:09<15:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1418/2820 [15:11<15:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1418/2820 [15:11<15:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1419/2820 [15:11<15:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1419/2820 [15:11<15:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1420/2820 [15:11<14:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1420/2820 [15:11<14:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1421/2820 [15:13<14:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1421/2820 [15:13<14:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1422/2820 [15:13<14:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1422/2820 [15:13<14:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1423/2820 [15:13<14:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1423/2820 [15:13<14:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1424/2820 [15:15<14:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  50%|█████     | 1424/2820 [15:15<14:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1425/2820 [15:15<14:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1425/2820 [15:15<14:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1426/2820 [15:15<14:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1426/2820 [15:15<14:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1427/2820 [15:17<14:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1427/2820 [15:17<14:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1428/2820 [15:17<14:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1428/2820 [15:17<14:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1429/2820 [15:17<14:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1429/2820 [15:17<14:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1430/2820 [15:19<14:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1430/2820 [15:19<14:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1431/2820 [15:19<14:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1431/2820 [15:19<14:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1432/2820 [15:19<14:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1432/2820 [15:19<14:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1433/2820 [15:21<14:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1433/2820 [15:21<14:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1434/2820 [15:21<14:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1434/2820 [15:21<14:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1435/2820 [15:21<14:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1435/2820 [15:21<14:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1436/2820 [15:23<14:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1436/2820 [15:23<14:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1437/2820 [15:23<14:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1437/2820 [15:23<14:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1438/2820 [15:23<14:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1438/2820 [15:23<14:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1439/2820 [15:25<14:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1439/2820 [15:25<14:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1440/2820 [15:25<14:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1440/2820 [15:25<14:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1441/2820 [15:25<14:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1441/2820 [15:25<14:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1442/2820 [15:27<14:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1442/2820 [15:27<14:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1443/2820 [15:27<14:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1443/2820 [15:27<14:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1444/2820 [15:27<14:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1444/2820 [15:27<14:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1445/2820 [15:29<14:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████     | 1445/2820 [15:29<14:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1446/2820 [15:29<14:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1446/2820 [15:29<14:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1447/2820 [15:29<14:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1447/2820 [15:29<14:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1448/2820 [15:31<14:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1448/2820 [15:31<14:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1449/2820 [15:31<14:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1449/2820 [15:31<14:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1450/2820 [15:31<14:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1450/2820 [15:31<14:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1451/2820 [15:33<14:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1451/2820 [15:33<14:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1452/2820 [15:33<14:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  51%|█████▏    | 1452/2820 [15:33<14:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1453/2820 [15:33<14:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1453/2820 [15:33<14:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1454/2820 [15:35<14:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1454/2820 [15:35<14:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1455/2820 [15:35<14:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1455/2820 [15:35<14:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1456/2820 [15:35<14:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1456/2820 [15:35<14:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1457/2820 [15:37<14:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1457/2820 [15:37<14:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1458/2820 [15:37<14:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1458/2820 [15:37<14:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1459/2820 [15:37<14:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1459/2820 [15:37<14:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1460/2820 [15:39<14:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1460/2820 [15:39<14:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1461/2820 [15:39<14:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1461/2820 [15:39<14:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1462/2820 [15:39<14:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1462/2820 [15:39<14:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1463/2820 [15:41<14:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1463/2820 [15:41<14:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1464/2820 [15:41<14:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1464/2820 [15:41<14:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1465/2820 [15:41<14:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1465/2820 [15:41<14:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1466/2820 [15:43<14:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1466/2820 [15:43<14:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1467/2820 [15:43<14:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1467/2820 [15:43<14:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1468/2820 [15:43<14:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1468/2820 [15:43<14:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1469/2820 [15:45<14:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1469/2820 [15:45<14:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1470/2820 [15:45<14:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1470/2820 [15:45<14:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1471/2820 [15:45<14:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1471/2820 [15:45<14:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1472/2820 [15:47<14:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1472/2820 [15:47<14:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1473/2820 [15:47<14:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1473/2820 [15:47<14:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1474/2820 [15:47<14:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1474/2820 [15:47<14:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1475/2820 [15:49<14:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1475/2820 [15:49<14:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1476/2820 [15:49<14:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1476/2820 [15:49<14:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1477/2820 [15:49<14:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1477/2820 [15:49<14:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1478/2820 [15:51<14:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1478/2820 [15:51<14:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1479/2820 [15:51<14:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1479/2820 [15:51<14:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1480/2820 [15:51<14:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  52%|█████▏    | 1480/2820 [15:51<14:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1481/2820 [15:53<14:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1481/2820 [15:53<14:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1482/2820 [15:53<14:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1482/2820 [15:53<14:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1483/2820 [15:53<14:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1483/2820 [15:53<14:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1484/2820 [15:55<14:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1484/2820 [15:55<14:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1485/2820 [15:55<14:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1485/2820 [15:55<14:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1486/2820 [15:55<14:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1486/2820 [15:55<14:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1487/2820 [15:56<14:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1487/2820 [15:56<14:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1488/2820 [15:56<14:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1488/2820 [15:56<14:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1489/2820 [15:57<14:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1489/2820 [15:57<14:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1490/2820 [15:58<14:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1490/2820 [15:58<14:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1491/2820 [15:58<14:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1491/2820 [15:58<14:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1492/2820 [15:58<14:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1492/2820 [15:58<14:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1493/2820 [16:00<14:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1493/2820 [16:00<14:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1494/2820 [16:00<14:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1494/2820 [16:00<14:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1495/2820 [16:00<14:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1495/2820 [16:00<14:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1496/2820 [16:02<14:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1496/2820 [16:02<14:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1497/2820 [16:02<14:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1497/2820 [16:02<14:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1498/2820 [16:02<14:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1498/2820 [16:02<14:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1499/2820 [16:04<14:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1499/2820 [16:04<14:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1500/2820 [16:04<14:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1500/2820 [16:04<14:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1501/2820 [16:04<14:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1501/2820 [16:04<14:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1502/2820 [16:06<14:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1502/2820 [16:06<14:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1503/2820 [16:06<14:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1503/2820 [16:06<14:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1504/2820 [16:06<14:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1504/2820 [16:06<14:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1505/2820 [16:08<14:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1505/2820 [16:08<14:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1506/2820 [16:08<14:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1506/2820 [16:08<14:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1507/2820 [16:08<14:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1507/2820 [16:08<14:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1508/2820 [16:10<14:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  53%|█████▎    | 1508/2820 [16:10<14:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1509/2820 [16:10<14:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1509/2820 [16:10<14:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1510/2820 [16:10<14:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1510/2820 [16:10<14:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1511/2820 [16:12<14:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1511/2820 [16:12<14:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1512/2820 [16:12<14:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1512/2820 [16:12<14:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1513/2820 [16:12<13:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1513/2820 [16:12<13:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1514/2820 [16:14<14:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1514/2820 [16:14<14:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1515/2820 [16:14<13:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▎    | 1515/2820 [16:14<13:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1516/2820 [16:14<13:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1516/2820 [16:14<13:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1517/2820 [16:16<13:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1517/2820 [16:16<13:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1518/2820 [16:16<13:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1518/2820 [16:16<13:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1519/2820 [16:16<13:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1519/2820 [16:16<13:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1520/2820 [16:18<13:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1520/2820 [16:18<13:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1521/2820 [16:18<13:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1521/2820 [16:18<13:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1522/2820 [16:18<13:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1522/2820 [16:18<13:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1523/2820 [16:20<13:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1523/2820 [16:20<13:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1524/2820 [16:20<13:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1524/2820 [16:20<13:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1525/2820 [16:20<13:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1525/2820 [16:20<13:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1526/2820 [16:22<13:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1526/2820 [16:22<13:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1527/2820 [16:22<13:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1527/2820 [16:22<13:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1528/2820 [16:22<13:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1528/2820 [16:22<13:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1529/2820 [16:24<13:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1529/2820 [16:24<13:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1530/2820 [16:24<13:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1530/2820 [16:24<13:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1531/2820 [16:24<13:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1531/2820 [16:24<13:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1532/2820 [16:26<13:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1532/2820 [16:26<13:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1533/2820 [16:26<13:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1533/2820 [16:26<13:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1534/2820 [16:26<13:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1534/2820 [16:26<13:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1535/2820 [16:28<13:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1535/2820 [16:28<13:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1536/2820 [16:28<13:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  54%|█████▍    | 1536/2820 [16:28<13:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1537/2820 [16:28<13:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1537/2820 [16:28<13:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1538/2820 [16:30<13:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1538/2820 [16:30<13:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1539/2820 [16:30<13:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1539/2820 [16:30<13:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1540/2820 [16:30<13:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1540/2820 [16:30<13:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1541/2820 [16:32<13:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1541/2820 [16:32<13:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1542/2820 [16:32<13:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1542/2820 [16:32<13:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1543/2820 [16:32<13:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1543/2820 [16:32<13:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1544/2820 [16:34<13:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1544/2820 [16:34<13:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1545/2820 [16:34<13:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1545/2820 [16:34<13:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1546/2820 [16:34<13:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1546/2820 [16:34<13:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1547/2820 [16:35<13:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1547/2820 [16:35<13:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1548/2820 [16:35<13:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1548/2820 [16:35<13:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1549/2820 [16:35<13:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1549/2820 [16:35<13:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1550/2820 [16:37<13:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▍    | 1550/2820 [16:37<13:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1551/2820 [16:37<13:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1551/2820 [16:37<13:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1552/2820 [16:37<13:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1552/2820 [16:37<13:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1553/2820 [16:39<13:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1553/2820 [16:39<13:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1554/2820 [16:39<13:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1554/2820 [16:39<13:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1555/2820 [16:39<13:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1555/2820 [16:39<13:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1556/2820 [16:41<13:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1556/2820 [16:41<13:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1557/2820 [16:42<13:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1557/2820 [16:42<13:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1558/2820 [16:42<13:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1558/2820 [16:42<13:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1559/2820 [16:43<13:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1559/2820 [16:43<13:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1560/2820 [16:43<13:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1560/2820 [16:43<13:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1561/2820 [16:44<13:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1561/2820 [16:44<13:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1562/2820 [16:46<13:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1562/2820 [16:46<13:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1563/2820 [16:46<13:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1563/2820 [16:46<13:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1564/2820 [16:46<13:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1564/2820 [16:46<13:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1565/2820 [16:48<13:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  55%|█████▌    | 1565/2820 [16:48<13:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1566/2820 [16:48<13:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1566/2820 [16:48<13:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1567/2820 [16:48<13:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1567/2820 [16:48<13:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1568/2820 [16:49<13:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1568/2820 [16:49<13:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1569/2820 [16:49<13:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1569/2820 [16:49<13:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1570/2820 [16:50<13:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1570/2820 [16:50<13:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1571/2820 [16:51<13:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1571/2820 [16:51<13:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1572/2820 [16:51<13:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1572/2820 [16:51<13:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1573/2820 [16:51<13:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1573/2820 [16:51<13:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1574/2820 [16:53<13:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1574/2820 [16:53<13:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1575/2820 [16:53<13:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1575/2820 [16:53<13:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1576/2820 [16:53<13:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1576/2820 [16:53<13:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1577/2820 [16:55<13:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1577/2820 [16:55<13:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1578/2820 [16:55<13:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1578/2820 [16:55<13:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1579/2820 [16:55<13:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1579/2820 [16:55<13:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1580/2820 [16:57<13:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1580/2820 [16:57<13:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1581/2820 [16:57<13:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1581/2820 [16:57<13:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1582/2820 [16:57<13:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1582/2820 [16:57<13:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1583/2820 [16:59<13:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1583/2820 [16:59<13:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1584/2820 [16:59<13:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1584/2820 [16:59<13:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1585/2820 [16:59<13:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1585/2820 [16:59<13:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1586/2820 [17:01<13:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▌    | 1586/2820 [17:01<13:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1587/2820 [17:01<13:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1587/2820 [17:01<13:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1588/2820 [17:01<13:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1588/2820 [17:01<13:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1589/2820 [17:03<13:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1589/2820 [17:03<13:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1590/2820 [17:03<13:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1590/2820 [17:03<13:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1591/2820 [17:03<13:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1591/2820 [17:03<13:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1592/2820 [17:04<13:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1592/2820 [17:04<13:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1593/2820 [17:05<13:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  56%|█████▋    | 1593/2820 [17:05<13:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1594/2820 [17:05<13:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1594/2820 [17:05<13:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1595/2820 [17:06<13:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1595/2820 [17:06<13:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1596/2820 [17:06<13:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1596/2820 [17:06<13:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1597/2820 [17:06<13:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1597/2820 [17:06<13:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1598/2820 [17:08<13:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1598/2820 [17:08<13:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1599/2820 [17:08<13:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1599/2820 [17:08<13:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1600/2820 [17:08<13:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1600/2820 [17:08<13:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1601/2820 [17:10<13:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1601/2820 [17:10<13:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1602/2820 [17:10<13:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1602/2820 [17:10<13:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1603/2820 [17:10<13:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1603/2820 [17:10<13:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1604/2820 [17:12<13:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1604/2820 [17:12<13:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1605/2820 [17:12<13:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1605/2820 [17:12<13:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1606/2820 [17:12<13:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1606/2820 [17:12<13:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1607/2820 [17:14<13:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1607/2820 [17:14<13:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1608/2820 [17:14<12:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1608/2820 [17:14<12:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1609/2820 [17:14<12:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1609/2820 [17:14<12:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1610/2820 [17:16<12:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1610/2820 [17:16<12:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1611/2820 [17:16<12:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1611/2820 [17:16<12:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1612/2820 [17:16<12:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1612/2820 [17:16<12:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1613/2820 [17:18<12:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1613/2820 [17:18<12:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1614/2820 [17:18<12:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1614/2820 [17:18<12:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1615/2820 [17:18<12:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1615/2820 [17:18<12:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1616/2820 [17:20<12:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1616/2820 [17:20<12:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1617/2820 [17:20<12:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1617/2820 [17:20<12:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1618/2820 [17:20<12:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1618/2820 [17:20<12:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1619/2820 [17:22<12:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1619/2820 [17:22<12:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1620/2820 [17:22<12:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1620/2820 [17:22<12:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1621/2820 [17:22<12:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  57%|█████▋    | 1621/2820 [17:22<12:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1622/2820 [17:24<12:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1622/2820 [17:24<12:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1623/2820 [17:24<12:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1623/2820 [17:24<12:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1624/2820 [17:24<12:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1624/2820 [17:24<12:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1625/2820 [17:26<12:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1625/2820 [17:26<12:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1626/2820 [17:26<12:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1626/2820 [17:26<12:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1627/2820 [17:26<12:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1627/2820 [17:26<12:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1628/2820 [17:28<12:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1628/2820 [17:28<12:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1629/2820 [17:28<12:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1629/2820 [17:28<12:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1630/2820 [17:28<12:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1630/2820 [17:28<12:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1631/2820 [17:30<12:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1631/2820 [17:30<12:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1632/2820 [17:30<12:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1632/2820 [17:30<12:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1633/2820 [17:30<12:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1633/2820 [17:30<12:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1634/2820 [17:32<12:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1634/2820 [17:32<12:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1635/2820 [17:32<12:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1635/2820 [17:32<12:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1636/2820 [17:32<12:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1636/2820 [17:32<12:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1637/2820 [17:34<12:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1637/2820 [17:34<12:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1638/2820 [17:34<12:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1638/2820 [17:34<12:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1639/2820 [17:34<12:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1639/2820 [17:34<12:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1640/2820 [17:36<12:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1640/2820 [17:36<12:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1641/2820 [17:36<12:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1641/2820 [17:36<12:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1642/2820 [17:36<12:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1642/2820 [17:36<12:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1643/2820 [17:37<12:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1643/2820 [17:37<12:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1644/2820 [17:38<12:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1644/2820 [17:38<12:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1645/2820 [17:38<12:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1645/2820 [17:38<12:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1646/2820 [17:39<12:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1646/2820 [17:39<12:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1647/2820 [17:39<12:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1647/2820 [17:39<12:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1648/2820 [17:39<12:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1648/2820 [17:39<12:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1649/2820 [17:41<12:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  58%|█████▊    | 1649/2820 [17:41<12:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1650/2820 [17:41<12:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1650/2820 [17:41<12:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1651/2820 [17:41<12:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1651/2820 [17:41<12:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1652/2820 [17:43<12:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1652/2820 [17:43<12:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1653/2820 [17:43<12:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1653/2820 [17:43<12:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1654/2820 [17:43<12:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1654/2820 [17:43<12:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1655/2820 [17:45<12:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1655/2820 [17:45<12:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1656/2820 [17:45<12:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▊    | 1656/2820 [17:45<12:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1657/2820 [17:45<12:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1657/2820 [17:45<12:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1658/2820 [17:47<12:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1658/2820 [17:47<12:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1659/2820 [17:47<12:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1659/2820 [17:47<12:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1660/2820 [17:47<12:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1660/2820 [17:47<12:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1661/2820 [17:49<12:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1661/2820 [17:49<12:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1662/2820 [17:49<12:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1662/2820 [17:49<12:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1663/2820 [17:49<12:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1663/2820 [17:49<12:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1664/2820 [17:50<12:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1664/2820 [17:50<12:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1665/2820 [17:50<12:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1665/2820 [17:50<12:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1666/2820 [17:50<12:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1666/2820 [17:50<12:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1667/2820 [17:52<12:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1667/2820 [17:52<12:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1668/2820 [17:52<12:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1668/2820 [17:52<12:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1669/2820 [17:52<12:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1669/2820 [17:52<12:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1670/2820 [17:54<12:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1670/2820 [17:54<12:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1671/2820 [17:54<12:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1671/2820 [17:54<12:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1672/2820 [17:54<12:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1672/2820 [17:54<12:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1673/2820 [17:56<12:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1673/2820 [17:56<12:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1674/2820 [17:56<12:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1674/2820 [17:56<12:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1675/2820 [17:56<12:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1675/2820 [17:56<12:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1676/2820 [17:58<12:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1676/2820 [17:58<12:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1677/2820 [17:58<12:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  59%|█████▉    | 1677/2820 [17:58<12:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1678/2820 [17:58<12:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1678/2820 [17:58<12:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1679/2820 [18:00<12:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1679/2820 [18:00<12:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1680/2820 [18:00<12:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1680/2820 [18:00<12:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1681/2820 [18:00<12:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1681/2820 [18:00<12:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1682/2820 [18:02<12:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1682/2820 [18:02<12:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1683/2820 [18:02<12:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1683/2820 [18:02<12:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1684/2820 [18:02<12:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1684/2820 [18:02<12:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1685/2820 [18:04<12:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1685/2820 [18:04<12:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1686/2820 [18:04<12:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1686/2820 [18:04<12:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1687/2820 [18:04<12:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1687/2820 [18:04<12:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1688/2820 [18:06<12:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1688/2820 [18:06<12:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1689/2820 [18:06<12:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1689/2820 [18:06<12:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1690/2820 [18:06<12:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1690/2820 [18:06<12:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1691/2820 [18:07<12:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|█████▉    | 1691/2820 [18:07<12:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1692/2820 [18:07<12:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1692/2820 [18:07<12:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1693/2820 [18:07<12:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1693/2820 [18:07<12:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1694/2820 [18:09<12:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1694/2820 [18:09<12:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1695/2820 [18:09<12:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1695/2820 [18:09<12:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1696/2820 [18:09<12:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1696/2820 [18:09<12:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1697/2820 [18:11<12:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1697/2820 [18:11<12:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1698/2820 [18:11<12:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1698/2820 [18:11<12:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1699/2820 [18:11<12:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1699/2820 [18:11<12:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1700/2820 [18:13<12:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1700/2820 [18:13<12:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1701/2820 [18:13<11:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1701/2820 [18:13<11:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1702/2820 [18:13<11:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1702/2820 [18:13<11:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1703/2820 [18:15<11:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1703/2820 [18:15<11:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1704/2820 [18:15<11:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1704/2820 [18:15<11:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1705/2820 [18:15<11:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1705/2820 [18:15<11:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1706/2820 [18:17<11:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  60%|██████    | 1706/2820 [18:17<11:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1707/2820 [18:17<11:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1707/2820 [18:17<11:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1708/2820 [18:17<11:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1708/2820 [18:17<11:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1709/2820 [18:19<11:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1709/2820 [18:19<11:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1710/2820 [18:19<11:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1710/2820 [18:19<11:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1711/2820 [18:19<11:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1711/2820 [18:19<11:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1712/2820 [18:20<11:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1712/2820 [18:20<11:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1713/2820 [18:20<11:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1713/2820 [18:20<11:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1714/2820 [18:21<11:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1714/2820 [18:21<11:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1715/2820 [18:22<11:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1715/2820 [18:22<11:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1716/2820 [18:22<11:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1716/2820 [18:22<11:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1717/2820 [18:23<11:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1717/2820 [18:23<11:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1718/2820 [18:24<11:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1718/2820 [18:24<11:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1719/2820 [18:24<11:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1719/2820 [18:24<11:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1720/2820 [18:25<11:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1720/2820 [18:25<11:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1721/2820 [18:27<11:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1721/2820 [18:27<11:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1722/2820 [18:27<11:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1722/2820 [18:27<11:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1723/2820 [18:27<11:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1723/2820 [18:27<11:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1724/2820 [18:29<11:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1724/2820 [18:29<11:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1725/2820 [18:29<11:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1725/2820 [18:29<11:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1726/2820 [18:29<11:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1726/2820 [18:29<11:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1727/2820 [18:31<11:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████    | 1727/2820 [18:31<11:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1728/2820 [18:31<11:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1728/2820 [18:31<11:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1729/2820 [18:31<11:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1729/2820 [18:31<11:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1730/2820 [18:32<11:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1730/2820 [18:32<11:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1731/2820 [18:32<11:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1731/2820 [18:32<11:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1732/2820 [18:32<11:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1732/2820 [18:32<11:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1733/2820 [18:34<11:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1733/2820 [18:34<11:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1734/2820 [18:34<11:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  61%|██████▏   | 1734/2820 [18:34<11:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1735/2820 [18:34<11:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1735/2820 [18:34<11:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1736/2820 [18:36<11:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1736/2820 [18:36<11:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1737/2820 [18:36<11:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1737/2820 [18:36<11:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1738/2820 [18:36<11:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1738/2820 [18:36<11:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1739/2820 [18:38<11:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1739/2820 [18:38<11:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1740/2820 [18:38<11:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1740/2820 [18:38<11:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1741/2820 [18:38<11:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1741/2820 [18:38<11:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1742/2820 [18:40<11:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1742/2820 [18:40<11:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1743/2820 [18:40<11:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1743/2820 [18:40<11:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1744/2820 [18:40<11:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1744/2820 [18:40<11:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1745/2820 [18:42<11:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1745/2820 [18:42<11:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1746/2820 [18:42<11:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1746/2820 [18:42<11:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1747/2820 [18:42<11:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1747/2820 [18:42<11:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1748/2820 [18:44<11:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1748/2820 [18:44<11:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1749/2820 [18:44<11:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1749/2820 [18:44<11:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1750/2820 [18:44<11:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1750/2820 [18:44<11:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1751/2820 [18:46<11:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1751/2820 [18:46<11:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1752/2820 [18:46<11:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1752/2820 [18:46<11:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1753/2820 [18:46<11:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1753/2820 [18:46<11:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1754/2820 [18:48<11:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1754/2820 [18:48<11:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1755/2820 [18:48<11:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1755/2820 [18:48<11:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1756/2820 [18:48<11:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1756/2820 [18:48<11:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1757/2820 [18:50<11:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1757/2820 [18:50<11:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1758/2820 [18:50<11:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1758/2820 [18:50<11:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1759/2820 [18:50<11:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1759/2820 [18:50<11:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1760/2820 [18:52<11:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1760/2820 [18:52<11:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1761/2820 [18:52<11:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1761/2820 [18:52<11:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1762/2820 [18:52<11:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  62%|██████▏   | 1762/2820 [18:52<11:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1763/2820 [18:54<11:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1763/2820 [18:54<11:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1764/2820 [18:54<11:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1764/2820 [18:54<11:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1765/2820 [18:54<11:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1765/2820 [18:54<11:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1766/2820 [18:56<11:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1766/2820 [18:56<11:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1767/2820 [18:56<11:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1767/2820 [18:56<11:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1768/2820 [18:56<11:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1768/2820 [18:56<11:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1769/2820 [18:57<11:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1769/2820 [18:57<11:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1770/2820 [18:58<11:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1770/2820 [18:58<11:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1771/2820 [18:58<11:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1771/2820 [18:58<11:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1772/2820 [18:59<11:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1772/2820 [18:59<11:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1773/2820 [18:59<11:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1773/2820 [18:59<11:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1774/2820 [19:00<11:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1774/2820 [19:00<11:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1775/2820 [19:01<11:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1775/2820 [19:01<11:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1776/2820 [19:01<11:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1776/2820 [19:01<11:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1777/2820 [19:01<11:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1777/2820 [19:01<11:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1778/2820 [19:03<11:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1778/2820 [19:03<11:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1779/2820 [19:03<11:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1779/2820 [19:03<11:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1780/2820 [19:04<11:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1780/2820 [19:04<11:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1781/2820 [19:05<11:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1781/2820 [19:05<11:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1782/2820 [19:05<11:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1782/2820 [19:05<11:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1783/2820 [19:05<11:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1783/2820 [19:05<11:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1784/2820 [19:07<11:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1784/2820 [19:07<11:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1785/2820 [19:07<11:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1785/2820 [19:07<11:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1786/2820 [19:07<11:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1786/2820 [19:07<11:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1787/2820 [19:09<11:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1787/2820 [19:09<11:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1788/2820 [19:09<11:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1788/2820 [19:09<11:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1789/2820 [19:09<11:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1789/2820 [19:09<11:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1790/2820 [19:11<11:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  63%|██████▎   | 1790/2820 [19:11<11:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1791/2820 [19:11<11:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1791/2820 [19:11<11:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1792/2820 [19:11<11:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1792/2820 [19:11<11:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1793/2820 [19:13<11:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1793/2820 [19:13<11:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1794/2820 [19:13<10:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1794/2820 [19:13<10:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1795/2820 [19:13<10:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1795/2820 [19:13<10:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1796/2820 [19:15<10:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1796/2820 [19:15<10:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1797/2820 [19:15<10:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▎   | 1797/2820 [19:15<10:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1798/2820 [19:15<10:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1798/2820 [19:15<10:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1799/2820 [19:17<10:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1799/2820 [19:17<10:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1800/2820 [19:17<10:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1800/2820 [19:17<10:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1801/2820 [19:17<10:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1801/2820 [19:17<10:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1802/2820 [19:18<10:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1802/2820 [19:18<10:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1803/2820 [19:18<10:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1803/2820 [19:18<10:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1804/2820 [19:18<10:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1804/2820 [19:18<10:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1805/2820 [19:20<10:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1805/2820 [19:20<10:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1806/2820 [19:20<10:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1806/2820 [19:20<10:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1807/2820 [19:20<10:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1807/2820 [19:20<10:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1808/2820 [19:22<10:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1808/2820 [19:22<10:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1809/2820 [19:22<10:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1809/2820 [19:22<10:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1810/2820 [19:22<10:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1810/2820 [19:22<10:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1811/2820 [19:24<10:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1811/2820 [19:24<10:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1812/2820 [19:24<10:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1812/2820 [19:24<10:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1813/2820 [19:24<10:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1813/2820 [19:24<10:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1814/2820 [19:26<10:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1814/2820 [19:26<10:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1815/2820 [19:26<10:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1815/2820 [19:26<10:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1816/2820 [19:26<10:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1816/2820 [19:26<10:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1817/2820 [19:28<10:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1817/2820 [19:28<10:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1818/2820 [19:28<10:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  64%|██████▍   | 1818/2820 [19:28<10:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1819/2820 [19:28<10:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1819/2820 [19:28<10:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1820/2820 [19:30<10:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1820/2820 [19:30<10:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1821/2820 [19:30<10:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1821/2820 [19:30<10:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1822/2820 [19:30<10:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1822/2820 [19:30<10:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1823/2820 [19:32<10:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1823/2820 [19:32<10:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1824/2820 [19:32<10:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1824/2820 [19:32<10:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1825/2820 [19:32<10:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1825/2820 [19:32<10:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1826/2820 [19:33<10:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1826/2820 [19:33<10:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1827/2820 [19:33<10:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1827/2820 [19:33<10:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1828/2820 [19:33<10:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1828/2820 [19:33<10:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1829/2820 [19:35<10:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1829/2820 [19:35<10:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1830/2820 [19:35<10:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1830/2820 [19:35<10:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1831/2820 [19:35<10:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1831/2820 [19:35<10:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1832/2820 [19:37<10:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▍   | 1832/2820 [19:37<10:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1833/2820 [19:37<10:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1833/2820 [19:37<10:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1834/2820 [19:37<10:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1834/2820 [19:37<10:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1835/2820 [19:39<10:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1835/2820 [19:39<10:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1836/2820 [19:39<10:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1836/2820 [19:39<10:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1837/2820 [19:39<10:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1837/2820 [19:39<10:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1838/2820 [19:41<10:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1838/2820 [19:41<10:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1839/2820 [19:41<10:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1839/2820 [19:41<10:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1840/2820 [19:41<10:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1840/2820 [19:41<10:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1841/2820 [19:43<10:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1841/2820 [19:43<10:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1842/2820 [19:43<10:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1842/2820 [19:43<10:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1843/2820 [19:43<10:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1843/2820 [19:43<10:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1844/2820 [19:45<10:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1844/2820 [19:45<10:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1845/2820 [19:45<10:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1845/2820 [19:45<10:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1846/2820 [19:45<10:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1846/2820 [19:45<10:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1847/2820 [19:47<10:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  65%|██████▌   | 1847/2820 [19:47<10:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1848/2820 [19:47<10:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1848/2820 [19:47<10:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1849/2820 [19:47<10:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1849/2820 [19:47<10:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1850/2820 [19:49<10:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1850/2820 [19:49<10:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1851/2820 [19:49<10:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1851/2820 [19:49<10:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1852/2820 [19:49<10:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1852/2820 [19:49<10:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1853/2820 [19:51<10:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1853/2820 [19:51<10:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1854/2820 [19:51<10:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1854/2820 [19:51<10:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1855/2820 [19:51<10:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1855/2820 [19:51<10:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1856/2820 [19:53<10:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1856/2820 [19:53<10:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1857/2820 [19:53<10:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1857/2820 [19:53<10:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1858/2820 [19:53<10:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1858/2820 [19:53<10:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1859/2820 [19:55<10:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1859/2820 [19:55<10:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1860/2820 [19:55<10:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1860/2820 [19:55<10:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1861/2820 [19:55<10:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1861/2820 [19:55<10:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1862/2820 [19:57<10:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1862/2820 [19:57<10:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1863/2820 [19:57<10:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1863/2820 [19:57<10:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1864/2820 [19:57<10:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1864/2820 [19:57<10:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1865/2820 [19:58<10:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1865/2820 [19:58<10:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1866/2820 [19:58<10:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1866/2820 [19:58<10:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1867/2820 [19:58<10:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1867/2820 [19:58<10:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1868/2820 [20:00<10:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▌   | 1868/2820 [20:00<10:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1869/2820 [20:00<10:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1869/2820 [20:00<10:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1870/2820 [20:00<10:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1870/2820 [20:00<10:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1871/2820 [20:02<10:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1871/2820 [20:02<10:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1872/2820 [20:02<10:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1872/2820 [20:02<10:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1873/2820 [20:02<10:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1873/2820 [20:02<10:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1874/2820 [20:04<10:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1874/2820 [20:04<10:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1875/2820 [20:04<10:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  66%|██████▋   | 1875/2820 [20:04<10:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1876/2820 [20:04<10:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1876/2820 [20:04<10:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1877/2820 [20:06<10:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1877/2820 [20:06<10:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1878/2820 [20:06<10:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1878/2820 [20:06<10:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1879/2820 [20:06<10:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1879/2820 [20:06<10:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1880/2820 [20:08<10:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1880/2820 [20:08<10:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1881/2820 [20:08<10:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1881/2820 [20:08<10:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1882/2820 [20:08<10:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1882/2820 [20:08<10:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1883/2820 [20:10<10:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1883/2820 [20:10<10:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1884/2820 [20:10<10:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1884/2820 [20:10<10:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1885/2820 [20:10<10:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1885/2820 [20:10<10:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1886/2820 [20:12<10:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1886/2820 [20:12<10:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1887/2820 [20:12<09:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1887/2820 [20:12<09:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1888/2820 [20:12<09:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1888/2820 [20:12<09:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1889/2820 [20:14<09:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1889/2820 [20:14<09:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1890/2820 [20:14<09:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1890/2820 [20:14<09:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1891/2820 [20:14<09:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1891/2820 [20:14<09:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1892/2820 [20:16<09:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1892/2820 [20:16<09:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1893/2820 [20:16<09:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1893/2820 [20:16<09:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1894/2820 [20:16<09:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1894/2820 [20:16<09:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1895/2820 [20:18<09:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1895/2820 [20:18<09:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1896/2820 [20:18<09:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1896/2820 [20:18<09:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1897/2820 [20:18<09:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1897/2820 [20:18<09:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1898/2820 [20:20<09:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1898/2820 [20:20<09:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1899/2820 [20:20<09:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1899/2820 [20:20<09:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1900/2820 [20:20<09:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1900/2820 [20:20<09:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1901/2820 [20:21<09:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1901/2820 [20:21<09:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1902/2820 [20:21<09:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1902/2820 [20:21<09:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1903/2820 [20:22<09:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  67%|██████▋   | 1903/2820 [20:22<09:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1904/2820 [20:23<09:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1904/2820 [20:23<09:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1905/2820 [20:23<09:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1905/2820 [20:23<09:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1906/2820 [20:24<09:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1906/2820 [20:24<09:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1907/2820 [20:25<09:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1907/2820 [20:25<09:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1908/2820 [20:25<09:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1908/2820 [20:25<09:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1909/2820 [20:25<09:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1909/2820 [20:25<09:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1910/2820 [20:27<09:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1910/2820 [20:27<09:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1911/2820 [20:27<09:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1911/2820 [20:27<09:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1912/2820 [20:27<09:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1912/2820 [20:27<09:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1913/2820 [20:29<09:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1913/2820 [20:29<09:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1914/2820 [20:29<09:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1914/2820 [20:29<09:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1915/2820 [20:30<09:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1915/2820 [20:30<09:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1916/2820 [20:31<09:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1916/2820 [20:31<09:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1917/2820 [20:31<09:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1917/2820 [20:31<09:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1918/2820 [20:31<09:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1918/2820 [20:31<09:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1919/2820 [20:33<09:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1919/2820 [20:33<09:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1920/2820 [20:33<09:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1920/2820 [20:33<09:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1921/2820 [20:33<09:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1921/2820 [20:33<09:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1922/2820 [20:35<09:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1922/2820 [20:35<09:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1923/2820 [20:35<09:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1923/2820 [20:35<09:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1924/2820 [20:35<09:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1924/2820 [20:35<09:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1925/2820 [20:37<09:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1925/2820 [20:37<09:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1926/2820 [20:37<09:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1926/2820 [20:37<09:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1927/2820 [20:37<09:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1927/2820 [20:37<09:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1928/2820 [20:39<09:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1928/2820 [20:39<09:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1929/2820 [20:39<09:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1929/2820 [20:39<09:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1930/2820 [20:39<09:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1930/2820 [20:39<09:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1931/2820 [20:41<09:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  68%|██████▊   | 1931/2820 [20:41<09:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1932/2820 [20:41<09:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1932/2820 [20:41<09:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1933/2820 [20:41<09:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1933/2820 [20:41<09:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1934/2820 [20:43<09:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1934/2820 [20:43<09:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1935/2820 [20:43<09:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1935/2820 [20:43<09:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1936/2820 [20:43<09:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1936/2820 [20:43<09:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1937/2820 [20:45<09:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1937/2820 [20:45<09:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1938/2820 [20:45<09:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▊   | 1938/2820 [20:45<09:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1939/2820 [20:45<09:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1939/2820 [20:45<09:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1940/2820 [20:47<09:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1940/2820 [20:47<09:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1941/2820 [20:47<09:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1941/2820 [20:47<09:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1942/2820 [20:47<09:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1942/2820 [20:47<09:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1943/2820 [20:49<09:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1943/2820 [20:49<09:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1944/2820 [20:49<09:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1944/2820 [20:49<09:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1945/2820 [20:49<09:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1945/2820 [20:49<09:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1946/2820 [20:51<09:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1946/2820 [20:51<09:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1947/2820 [20:51<09:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1947/2820 [20:51<09:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1948/2820 [20:51<09:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1948/2820 [20:51<09:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1949/2820 [20:53<09:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1949/2820 [20:53<09:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1950/2820 [20:53<09:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1950/2820 [20:53<09:19,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1951/2820 [20:53<09:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1951/2820 [20:53<09:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1952/2820 [20:55<09:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1952/2820 [20:55<09:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1953/2820 [20:55<09:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1953/2820 [20:55<09:17,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1954/2820 [20:55<09:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1954/2820 [20:55<09:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1955/2820 [20:57<09:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1955/2820 [20:57<09:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1956/2820 [20:57<09:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1956/2820 [20:57<09:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1957/2820 [20:57<09:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1957/2820 [20:57<09:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1958/2820 [20:59<09:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1958/2820 [20:59<09:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1959/2820 [20:59<09:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  69%|██████▉   | 1959/2820 [20:59<09:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1960/2820 [20:59<09:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1960/2820 [20:59<09:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1961/2820 [21:01<09:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1961/2820 [21:01<09:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1962/2820 [21:01<09:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1962/2820 [21:01<09:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1963/2820 [21:01<09:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1963/2820 [21:01<09:10,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1964/2820 [21:03<09:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1964/2820 [21:03<09:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1965/2820 [21:03<09:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1965/2820 [21:03<09:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1966/2820 [21:03<09:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1966/2820 [21:03<09:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1967/2820 [21:05<09:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1967/2820 [21:05<09:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1968/2820 [21:05<09:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1968/2820 [21:05<09:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1969/2820 [21:05<09:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1969/2820 [21:05<09:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1970/2820 [21:07<09:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1970/2820 [21:07<09:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1971/2820 [21:07<09:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1971/2820 [21:07<09:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1972/2820 [21:07<09:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1972/2820 [21:07<09:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1973/2820 [21:09<09:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|██████▉   | 1973/2820 [21:09<09:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1974/2820 [21:09<09:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1974/2820 [21:09<09:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1975/2820 [21:09<09:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1975/2820 [21:09<09:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1976/2820 [21:11<09:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1976/2820 [21:11<09:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1977/2820 [21:11<09:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1977/2820 [21:11<09:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1978/2820 [21:11<09:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1978/2820 [21:11<09:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1979/2820 [21:13<09:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1979/2820 [21:13<09:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1980/2820 [21:13<09:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1980/2820 [21:13<09:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1981/2820 [21:13<08:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1981/2820 [21:13<08:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1982/2820 [21:14<08:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1982/2820 [21:14<08:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1983/2820 [21:14<08:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1983/2820 [21:14<08:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1984/2820 [21:15<08:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1984/2820 [21:15<08:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1985/2820 [21:16<08:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1985/2820 [21:16<08:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1986/2820 [21:16<08:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1986/2820 [21:16<08:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1987/2820 [21:16<08:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1987/2820 [21:16<08:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1988/2820 [21:18<08:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  70%|███████   | 1988/2820 [21:18<08:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1989/2820 [21:18<08:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1989/2820 [21:18<08:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1990/2820 [21:18<08:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1990/2820 [21:18<08:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1991/2820 [21:20<08:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1991/2820 [21:20<08:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1992/2820 [21:20<08:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1992/2820 [21:20<08:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1993/2820 [21:20<08:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1993/2820 [21:20<08:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1994/2820 [21:22<08:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1994/2820 [21:22<08:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1995/2820 [21:22<08:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1995/2820 [21:22<08:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1996/2820 [21:22<08:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1996/2820 [21:22<08:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1997/2820 [21:24<08:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1997/2820 [21:24<08:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1998/2820 [21:24<08:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1998/2820 [21:24<08:48,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1999/2820 [21:24<08:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 1999/2820 [21:24<08:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2000/2820 [21:26<08:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2000/2820 [21:26<08:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2001/2820 [21:26<08:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2001/2820 [21:26<08:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2002/2820 [21:26<08:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2002/2820 [21:26<08:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2003/2820 [21:28<08:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2003/2820 [21:28<08:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2004/2820 [21:28<08:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2004/2820 [21:28<08:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2005/2820 [21:28<08:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2005/2820 [21:28<08:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2006/2820 [21:30<08:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2006/2820 [21:30<08:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2007/2820 [21:30<08:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2007/2820 [21:30<08:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2008/2820 [21:30<08:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2008/2820 [21:30<08:41,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2009/2820 [21:32<08:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████   | 2009/2820 [21:32<08:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2010/2820 [21:32<08:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2010/2820 [21:32<08:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2011/2820 [21:32<08:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2011/2820 [21:32<08:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2012/2820 [21:34<08:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2012/2820 [21:34<08:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2013/2820 [21:34<08:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2013/2820 [21:34<08:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2014/2820 [21:34<08:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2014/2820 [21:34<08:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2015/2820 [21:36<08:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2015/2820 [21:36<08:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2016/2820 [21:36<08:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  71%|███████▏  | 2016/2820 [21:36<08:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2017/2820 [21:36<08:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2017/2820 [21:36<08:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2018/2820 [21:38<08:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2018/2820 [21:38<08:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2019/2820 [21:38<08:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2019/2820 [21:38<08:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2020/2820 [21:38<08:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2020/2820 [21:38<08:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2021/2820 [21:40<08:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2021/2820 [21:40<08:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2022/2820 [21:40<08:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2022/2820 [21:40<08:33,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2023/2820 [21:40<08:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2023/2820 [21:40<08:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2024/2820 [21:41<08:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2024/2820 [21:41<08:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2025/2820 [21:41<08:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2025/2820 [21:41<08:31,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2026/2820 [21:41<08:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2026/2820 [21:41<08:30,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2027/2820 [21:43<08:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2027/2820 [21:43<08:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2028/2820 [21:43<08:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2028/2820 [21:43<08:29,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2029/2820 [21:43<08:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2029/2820 [21:43<08:28,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2030/2820 [21:45<08:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2030/2820 [21:45<08:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2031/2820 [21:45<08:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2031/2820 [21:45<08:27,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2032/2820 [21:45<08:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2032/2820 [21:45<08:26,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2033/2820 [21:47<08:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2033/2820 [21:47<08:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2034/2820 [21:47<08:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2034/2820 [21:47<08:25,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2035/2820 [21:47<08:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2035/2820 [21:47<08:24,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2036/2820 [21:49<08:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2036/2820 [21:49<08:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2037/2820 [21:49<08:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2037/2820 [21:49<08:23,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2038/2820 [21:49<08:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2038/2820 [21:49<08:22,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2039/2820 [21:51<08:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2039/2820 [21:51<08:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2040/2820 [21:51<08:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2040/2820 [21:51<08:21,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2041/2820 [21:51<08:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2041/2820 [21:51<08:20,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2042/2820 [21:53<08:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2042/2820 [21:53<08:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2043/2820 [21:53<08:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2043/2820 [21:53<08:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2044/2820 [21:53<08:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  72%|███████▏  | 2044/2820 [21:53<08:18,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2045/2820 [21:55<08:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2045/2820 [21:55<08:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2046/2820 [21:55<08:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2046/2820 [21:55<08:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2047/2820 [21:55<08:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2047/2820 [21:55<08:16,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2048/2820 [21:57<08:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2048/2820 [21:57<08:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2049/2820 [21:57<08:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2049/2820 [21:57<08:15,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2050/2820 [21:57<08:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2050/2820 [21:57<08:14,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2051/2820 [21:59<08:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2051/2820 [21:59<08:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2052/2820 [21:59<08:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2052/2820 [21:59<08:13,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2053/2820 [21:59<08:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2053/2820 [21:59<08:12,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2054/2820 [22:01<08:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2054/2820 [22:01<08:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2055/2820 [22:01<08:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2055/2820 [22:01<08:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2056/2820 [22:01<08:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2056/2820 [22:01<08:11,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2057/2820 [22:03<08:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2057/2820 [22:03<08:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2058/2820 [22:03<08:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2058/2820 [22:03<08:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2059/2820 [22:03<08:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2059/2820 [22:03<08:09,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2060/2820 [22:05<08:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2060/2820 [22:05<08:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2061/2820 [22:05<08:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2061/2820 [22:05<08:08,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2062/2820 [22:05<08:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2062/2820 [22:05<08:07,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2063/2820 [22:07<08:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2063/2820 [22:07<08:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2064/2820 [22:07<08:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2064/2820 [22:07<08:06,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2065/2820 [22:07<08:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2065/2820 [22:07<08:05,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2066/2820 [22:09<08:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2066/2820 [22:09<08:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2067/2820 [22:09<08:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2067/2820 [22:09<08:04,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2068/2820 [22:09<08:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2068/2820 [22:09<08:03,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2069/2820 [22:11<08:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2069/2820 [22:11<08:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2070/2820 [22:11<08:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2070/2820 [22:11<08:02,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2071/2820 [22:11<08:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2071/2820 [22:11<08:01,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2072/2820 [22:12<08:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  73%|███████▎  | 2072/2820 [22:12<08:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2073/2820 [22:13<08:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2073/2820 [22:13<08:00,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2074/2820 [22:13<07:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2074/2820 [22:13<07:59,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2075/2820 [22:15<07:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2075/2820 [22:15<07:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2076/2820 [22:15<07:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2076/2820 [22:15<07:58,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2077/2820 [22:15<07:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2077/2820 [22:15<07:57,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2078/2820 [22:16<07:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2078/2820 [22:16<07:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2079/2820 [22:16<07:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▎  | 2079/2820 [22:16<07:56,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2080/2820 [22:16<07:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2080/2820 [22:16<07:55,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2081/2820 [22:18<07:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2081/2820 [22:18<07:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2082/2820 [22:18<07:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2082/2820 [22:18<07:54,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2083/2820 [22:18<07:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2083/2820 [22:18<07:53,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2084/2820 [22:20<07:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2084/2820 [22:20<07:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2085/2820 [22:20<07:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2085/2820 [22:20<07:52,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2086/2820 [22:20<07:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2086/2820 [22:20<07:51,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2087/2820 [22:22<07:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2087/2820 [22:22<07:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2088/2820 [22:22<07:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2088/2820 [22:22<07:50,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2089/2820 [22:22<07:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2089/2820 [22:22<07:49,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2090/2820 [22:24<07:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2090/2820 [22:24<07:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2091/2820 [22:24<07:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2091/2820 [22:24<07:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2092/2820 [22:24<07:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2092/2820 [22:24<07:47,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2093/2820 [22:26<07:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2093/2820 [22:26<07:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2094/2820 [22:26<07:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2094/2820 [22:26<07:46,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2095/2820 [22:26<07:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2095/2820 [22:26<07:45,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2096/2820 [22:28<07:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2096/2820 [22:28<07:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2097/2820 [22:28<07:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2097/2820 [22:28<07:44,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2098/2820 [22:28<07:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2098/2820 [22:28<07:43,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2099/2820 [22:30<07:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2099/2820 [22:30<07:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2100/2820 [22:30<07:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  74%|███████▍  | 2100/2820 [22:30<07:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2101/2820 [22:30<07:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2101/2820 [22:30<07:42,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2102/2820 [22:32<07:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2102/2820 [22:32<07:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2103/2820 [22:32<07:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2103/2820 [22:32<07:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2104/2820 [22:32<07:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2104/2820 [22:32<07:40,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2105/2820 [22:34<07:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2105/2820 [22:34<07:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2106/2820 [22:34<07:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2106/2820 [22:34<07:39,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2107/2820 [22:34<07:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2107/2820 [22:34<07:38,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2108/2820 [22:35<07:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2108/2820 [22:35<07:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2109/2820 [22:35<07:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2109/2820 [22:35<07:37,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2110/2820 [22:35<07:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2110/2820 [22:35<07:36,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2111/2820 [22:37<07:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2111/2820 [22:37<07:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2112/2820 [22:37<07:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2112/2820 [22:37<07:35,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2113/2820 [22:37<07:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2113/2820 [22:37<07:34,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2114/2820 [22:40<07:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▍  | 2114/2820 [22:40<07:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2115/2820 [22:40<07:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2115/2820 [22:40<07:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2116/2820 [22:40<07:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2116/2820 [22:40<07:32,  1.56it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2117/2820 [22:42<07:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2117/2820 [22:42<07:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2118/2820 [22:42<07:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2118/2820 [22:42<07:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2119/2820 [22:42<07:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2119/2820 [22:42<07:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2120/2820 [22:45<07:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2120/2820 [22:45<07:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2121/2820 [22:45<07:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2121/2820 [22:45<07:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2122/2820 [22:45<07:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2122/2820 [22:45<07:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2123/2820 [22:47<07:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2123/2820 [22:47<07:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2124/2820 [22:47<07:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2124/2820 [22:47<07:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2125/2820 [22:47<07:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2125/2820 [22:47<07:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2126/2820 [22:49<07:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2126/2820 [22:49<07:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2127/2820 [22:49<07:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2127/2820 [22:49<07:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2128/2820 [22:49<07:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2128/2820 [22:49<07:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2129/2820 [22:51<07:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  75%|███████▌  | 2129/2820 [22:51<07:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2130/2820 [22:51<07:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2130/2820 [22:51<07:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2131/2820 [22:51<07:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2131/2820 [22:51<07:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2132/2820 [22:53<07:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2132/2820 [22:53<07:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2133/2820 [22:53<07:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2133/2820 [22:53<07:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2134/2820 [22:53<07:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2134/2820 [22:53<07:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2135/2820 [22:55<07:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2135/2820 [22:55<07:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2136/2820 [22:55<07:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2136/2820 [22:55<07:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2137/2820 [22:55<07:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2137/2820 [22:55<07:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2138/2820 [22:56<07:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2138/2820 [22:56<07:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2139/2820 [22:56<07:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2139/2820 [22:56<07:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2140/2820 [22:56<07:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2140/2820 [22:56<07:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2141/2820 [22:58<07:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2141/2820 [22:58<07:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2142/2820 [22:58<07:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2142/2820 [22:58<07:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2143/2820 [22:58<07:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2143/2820 [22:58<07:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2144/2820 [23:00<07:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2144/2820 [23:00<07:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2145/2820 [23:00<07:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2145/2820 [23:00<07:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2146/2820 [23:00<07:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2146/2820 [23:00<07:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2147/2820 [23:02<07:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2147/2820 [23:02<07:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2148/2820 [23:02<07:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2148/2820 [23:02<07:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2149/2820 [23:02<07:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2149/2820 [23:02<07:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2150/2820 [23:04<07:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▌  | 2150/2820 [23:04<07:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2151/2820 [23:04<07:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2151/2820 [23:04<07:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2152/2820 [23:04<07:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2152/2820 [23:04<07:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2153/2820 [23:06<07:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2153/2820 [23:06<07:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2154/2820 [23:06<07:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2154/2820 [23:06<07:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2155/2820 [23:06<07:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2155/2820 [23:06<07:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2156/2820 [23:08<07:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2156/2820 [23:08<07:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2157/2820 [23:08<07:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  76%|███████▋  | 2157/2820 [23:08<07:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2158/2820 [23:08<07:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2158/2820 [23:08<07:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2159/2820 [23:10<07:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2159/2820 [23:10<07:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2160/2820 [23:10<07:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2160/2820 [23:10<07:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2161/2820 [23:10<07:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2161/2820 [23:10<07:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2162/2820 [23:12<07:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2162/2820 [23:12<07:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2163/2820 [23:12<07:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2163/2820 [23:12<07:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2164/2820 [23:12<07:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2164/2820 [23:12<07:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2165/2820 [23:14<07:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2165/2820 [23:14<07:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2166/2820 [23:14<07:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2166/2820 [23:14<07:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2167/2820 [23:14<07:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2167/2820 [23:14<07:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2168/2820 [23:16<06:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2168/2820 [23:16<06:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2169/2820 [23:16<06:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2169/2820 [23:16<06:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2170/2820 [23:16<06:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2170/2820 [23:16<06:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2171/2820 [23:18<06:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2171/2820 [23:18<06:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2172/2820 [23:18<06:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2172/2820 [23:18<06:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2173/2820 [23:18<06:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2173/2820 [23:18<06:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2174/2820 [23:20<06:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2174/2820 [23:20<06:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2175/2820 [23:20<06:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2175/2820 [23:20<06:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2176/2820 [23:20<06:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2176/2820 [23:20<06:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2177/2820 [23:22<06:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2177/2820 [23:22<06:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2178/2820 [23:22<06:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2178/2820 [23:22<06:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2179/2820 [23:22<06:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2179/2820 [23:22<06:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2180/2820 [23:23<06:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2180/2820 [23:23<06:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2181/2820 [23:23<06:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2181/2820 [23:23<06:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2182/2820 [23:23<06:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2182/2820 [23:23<06:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2183/2820 [23:26<06:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2183/2820 [23:26<06:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2184/2820 [23:26<06:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2184/2820 [23:26<06:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2185/2820 [23:26<06:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  77%|███████▋  | 2185/2820 [23:26<06:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2186/2820 [23:27<06:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2186/2820 [23:27<06:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2187/2820 [23:28<06:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2187/2820 [23:28<06:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2188/2820 [23:28<06:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2188/2820 [23:28<06:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2189/2820 [23:29<06:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2189/2820 [23:29<06:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2190/2820 [23:29<06:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2190/2820 [23:29<06:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2191/2820 [23:29<06:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2191/2820 [23:29<06:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2192/2820 [23:31<06:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2192/2820 [23:31<06:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2193/2820 [23:31<06:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2193/2820 [23:31<06:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2194/2820 [23:32<06:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2194/2820 [23:32<06:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2195/2820 [23:33<06:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2195/2820 [23:33<06:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2196/2820 [23:33<06:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2196/2820 [23:33<06:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2197/2820 [23:33<06:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2197/2820 [23:33<06:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2198/2820 [23:35<06:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2198/2820 [23:35<06:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2199/2820 [23:35<06:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2199/2820 [23:35<06:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2200/2820 [23:35<06:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2200/2820 [23:35<06:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2201/2820 [23:37<06:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2201/2820 [23:37<06:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2202/2820 [23:37<06:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2202/2820 [23:37<06:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2203/2820 [23:37<06:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2203/2820 [23:37<06:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2204/2820 [23:39<06:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2204/2820 [23:39<06:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2205/2820 [23:39<06:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2205/2820 [23:39<06:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2206/2820 [23:39<06:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2206/2820 [23:39<06:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2207/2820 [23:41<06:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2207/2820 [23:41<06:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2208/2820 [23:41<06:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2208/2820 [23:41<06:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2209/2820 [23:41<06:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2209/2820 [23:41<06:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2210/2820 [23:43<06:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2210/2820 [23:43<06:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2211/2820 [23:43<06:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2211/2820 [23:43<06:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2212/2820 [23:43<06:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2212/2820 [23:43<06:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2213/2820 [23:45<06:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  78%|███████▊  | 2213/2820 [23:45<06:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2214/2820 [23:45<06:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2214/2820 [23:45<06:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2215/2820 [23:45<06:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2215/2820 [23:45<06:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2216/2820 [23:47<06:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2216/2820 [23:47<06:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2217/2820 [23:47<06:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2217/2820 [23:47<06:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2218/2820 [23:47<06:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2218/2820 [23:47<06:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2219/2820 [23:49<06:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2219/2820 [23:49<06:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2220/2820 [23:49<06:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▊  | 2220/2820 [23:49<06:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2221/2820 [23:49<06:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2221/2820 [23:49<06:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2222/2820 [23:51<06:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2222/2820 [23:51<06:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2223/2820 [23:51<06:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2223/2820 [23:51<06:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2224/2820 [23:51<06:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2224/2820 [23:51<06:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2225/2820 [23:52<06:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2225/2820 [23:52<06:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2226/2820 [23:52<06:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2226/2820 [23:52<06:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2227/2820 [23:53<06:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2227/2820 [23:53<06:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2228/2820 [23:54<06:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2228/2820 [23:54<06:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2229/2820 [23:54<06:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2229/2820 [23:54<06:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2230/2820 [23:54<06:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2230/2820 [23:54<06:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2231/2820 [23:56<06:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2231/2820 [23:56<06:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2232/2820 [23:56<06:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2232/2820 [23:56<06:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2233/2820 [23:56<06:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2233/2820 [23:56<06:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2234/2820 [23:58<06:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2234/2820 [23:58<06:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2235/2820 [23:58<06:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2235/2820 [23:58<06:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2236/2820 [23:58<06:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2236/2820 [23:58<06:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2237/2820 [24:00<06:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2237/2820 [24:00<06:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2238/2820 [24:00<06:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2238/2820 [24:00<06:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2239/2820 [24:00<06:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2239/2820 [24:00<06:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2240/2820 [24:02<06:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2240/2820 [24:02<06:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2241/2820 [24:02<06:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  79%|███████▉  | 2241/2820 [24:02<06:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2242/2820 [24:02<06:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2242/2820 [24:02<06:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2243/2820 [24:04<06:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2243/2820 [24:04<06:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2244/2820 [24:04<06:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2244/2820 [24:04<06:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2245/2820 [24:04<06:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2245/2820 [24:04<06:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2246/2820 [24:06<06:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2246/2820 [24:06<06:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2247/2820 [24:06<06:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2247/2820 [24:06<06:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2248/2820 [24:06<06:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2248/2820 [24:06<06:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2249/2820 [24:08<06:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2249/2820 [24:08<06:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2250/2820 [24:08<06:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2250/2820 [24:08<06:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2251/2820 [24:08<06:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2251/2820 [24:08<06:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2252/2820 [24:10<06:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2252/2820 [24:10<06:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2253/2820 [24:10<06:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2253/2820 [24:10<06:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2254/2820 [24:10<06:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2254/2820 [24:10<06:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2255/2820 [24:12<06:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|███████▉  | 2255/2820 [24:12<06:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2256/2820 [24:12<06:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2256/2820 [24:12<06:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2257/2820 [24:12<06:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2257/2820 [24:12<06:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2258/2820 [24:14<06:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2258/2820 [24:14<06:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2259/2820 [24:14<06:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2259/2820 [24:14<06:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2260/2820 [24:14<06:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2260/2820 [24:14<06:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2261/2820 [24:16<06:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2261/2820 [24:16<06:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2262/2820 [24:16<05:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2262/2820 [24:16<05:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2263/2820 [24:16<05:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2263/2820 [24:16<05:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2264/2820 [24:18<05:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2264/2820 [24:18<05:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2265/2820 [24:18<05:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2265/2820 [24:18<05:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2266/2820 [24:18<05:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2266/2820 [24:18<05:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2267/2820 [24:20<05:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2267/2820 [24:20<05:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2268/2820 [24:20<05:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2268/2820 [24:20<05:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2269/2820 [24:20<05:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2269/2820 [24:20<05:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2270/2820 [24:22<05:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  80%|████████  | 2270/2820 [24:22<05:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2271/2820 [24:22<05:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2271/2820 [24:22<05:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2272/2820 [24:22<05:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2272/2820 [24:22<05:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2273/2820 [24:24<05:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2273/2820 [24:24<05:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2274/2820 [24:24<05:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2274/2820 [24:24<05:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2275/2820 [24:24<05:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2275/2820 [24:24<05:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2276/2820 [24:26<05:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2276/2820 [24:26<05:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2277/2820 [24:26<05:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2277/2820 [24:26<05:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2278/2820 [24:26<05:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2278/2820 [24:26<05:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2279/2820 [24:28<05:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2279/2820 [24:28<05:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2280/2820 [24:28<05:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2280/2820 [24:28<05:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2281/2820 [24:28<05:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2281/2820 [24:28<05:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2282/2820 [24:30<05:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2282/2820 [24:30<05:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2283/2820 [24:30<05:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2283/2820 [24:30<05:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2284/2820 [24:30<05:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2284/2820 [24:30<05:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2285/2820 [24:32<05:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2285/2820 [24:32<05:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2286/2820 [24:32<05:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2286/2820 [24:32<05:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2287/2820 [24:32<05:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2287/2820 [24:32<05:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2288/2820 [24:34<05:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2288/2820 [24:34<05:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2289/2820 [24:34<05:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2289/2820 [24:34<05:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2290/2820 [24:34<05:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2290/2820 [24:34<05:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2291/2820 [24:36<05:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████  | 2291/2820 [24:36<05:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2292/2820 [24:36<05:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2292/2820 [24:36<05:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2293/2820 [24:36<05:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2293/2820 [24:36<05:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2294/2820 [24:38<05:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2294/2820 [24:38<05:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2295/2820 [24:38<05:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2295/2820 [24:38<05:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2296/2820 [24:38<05:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2296/2820 [24:38<05:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2297/2820 [24:39<05:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2297/2820 [24:39<05:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2298/2820 [24:39<05:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  81%|████████▏ | 2298/2820 [24:39<05:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2299/2820 [24:40<05:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2299/2820 [24:40<05:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2300/2820 [24:41<05:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2300/2820 [24:41<05:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2301/2820 [24:41<05:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2301/2820 [24:41<05:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2302/2820 [24:41<05:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2302/2820 [24:41<05:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2303/2820 [24:43<05:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2303/2820 [24:43<05:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2304/2820 [24:44<05:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2304/2820 [24:44<05:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2305/2820 [24:44<05:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2305/2820 [24:44<05:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2306/2820 [24:45<05:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2306/2820 [24:45<05:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2307/2820 [24:45<05:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2307/2820 [24:45<05:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2308/2820 [24:46<05:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2308/2820 [24:46<05:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2309/2820 [24:47<05:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2309/2820 [24:47<05:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2310/2820 [24:47<05:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2310/2820 [24:47<05:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2311/2820 [24:47<05:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2311/2820 [24:47<05:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2312/2820 [24:49<05:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2312/2820 [24:49<05:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2313/2820 [24:49<05:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2313/2820 [24:49<05:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2314/2820 [24:49<05:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2314/2820 [24:49<05:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2315/2820 [24:51<05:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2315/2820 [24:51<05:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2316/2820 [24:51<05:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2316/2820 [24:51<05:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2317/2820 [24:51<05:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2317/2820 [24:51<05:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2318/2820 [24:53<05:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2318/2820 [24:53<05:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2319/2820 [24:53<05:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2319/2820 [24:53<05:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2320/2820 [24:53<05:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2320/2820 [24:53<05:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2321/2820 [24:55<05:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2321/2820 [24:55<05:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2322/2820 [24:55<05:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2322/2820 [24:55<05:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2323/2820 [24:55<05:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2323/2820 [24:55<05:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2324/2820 [24:57<05:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2324/2820 [24:57<05:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2325/2820 [24:57<05:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2325/2820 [24:57<05:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2326/2820 [24:57<05:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  82%|████████▏ | 2326/2820 [24:57<05:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2327/2820 [24:59<05:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2327/2820 [24:59<05:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2328/2820 [24:59<05:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2328/2820 [24:59<05:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2329/2820 [24:59<05:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2329/2820 [24:59<05:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2330/2820 [25:01<05:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2330/2820 [25:01<05:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2331/2820 [25:01<05:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2331/2820 [25:01<05:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2332/2820 [25:01<05:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2332/2820 [25:01<05:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2333/2820 [25:03<05:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2333/2820 [25:03<05:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2334/2820 [25:03<05:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2334/2820 [25:03<05:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2335/2820 [25:03<05:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2335/2820 [25:03<05:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2336/2820 [25:05<05:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2336/2820 [25:05<05:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2337/2820 [25:05<05:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2337/2820 [25:05<05:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2338/2820 [25:05<05:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2338/2820 [25:05<05:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2339/2820 [25:07<05:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2339/2820 [25:07<05:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2340/2820 [25:07<05:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2340/2820 [25:07<05:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2341/2820 [25:07<05:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2341/2820 [25:07<05:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2342/2820 [25:09<05:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2342/2820 [25:09<05:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2343/2820 [25:09<05:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2343/2820 [25:09<05:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2344/2820 [25:09<05:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2344/2820 [25:09<05:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2345/2820 [25:11<05:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2345/2820 [25:11<05:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2346/2820 [25:11<05:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2346/2820 [25:11<05:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2347/2820 [25:11<05:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2347/2820 [25:11<05:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2348/2820 [25:13<05:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2348/2820 [25:13<05:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2349/2820 [25:13<05:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2349/2820 [25:13<05:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2350/2820 [25:13<05:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2350/2820 [25:13<05:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2351/2820 [25:15<05:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2351/2820 [25:15<05:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2352/2820 [25:15<05:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2352/2820 [25:15<05:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2353/2820 [25:15<05:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2353/2820 [25:15<05:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2354/2820 [25:17<05:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  83%|████████▎ | 2354/2820 [25:17<05:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2355/2820 [25:17<04:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2355/2820 [25:17<04:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2356/2820 [25:17<04:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2356/2820 [25:17<04:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2357/2820 [25:19<04:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2357/2820 [25:19<04:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2358/2820 [25:19<04:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2358/2820 [25:19<04:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2359/2820 [25:19<04:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2359/2820 [25:19<04:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2360/2820 [25:21<04:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2360/2820 [25:21<04:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2361/2820 [25:21<04:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▎ | 2361/2820 [25:21<04:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2362/2820 [25:21<04:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2362/2820 [25:21<04:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2363/2820 [25:23<04:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2363/2820 [25:23<04:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2364/2820 [25:23<04:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2364/2820 [25:23<04:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2365/2820 [25:23<04:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2365/2820 [25:23<04:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2366/2820 [25:24<04:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2366/2820 [25:24<04:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2367/2820 [25:24<04:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2367/2820 [25:24<04:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2368/2820 [25:24<04:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2368/2820 [25:24<04:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2369/2820 [25:26<04:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2369/2820 [25:26<04:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2370/2820 [25:26<04:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2370/2820 [25:26<04:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2371/2820 [25:26<04:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2371/2820 [25:26<04:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2372/2820 [25:28<04:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2372/2820 [25:28<04:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2373/2820 [25:28<04:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2373/2820 [25:28<04:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2374/2820 [25:28<04:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2374/2820 [25:28<04:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2375/2820 [25:30<04:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2375/2820 [25:30<04:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2376/2820 [25:30<04:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2376/2820 [25:30<04:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2377/2820 [25:30<04:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2377/2820 [25:30<04:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2378/2820 [25:32<04:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2378/2820 [25:32<04:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2379/2820 [25:32<04:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2379/2820 [25:32<04:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2380/2820 [25:32<04:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2380/2820 [25:32<04:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2381/2820 [25:34<04:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2381/2820 [25:34<04:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2382/2820 [25:34<04:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  84%|████████▍ | 2382/2820 [25:34<04:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2383/2820 [25:34<04:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2383/2820 [25:34<04:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2384/2820 [25:36<04:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2384/2820 [25:36<04:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2385/2820 [25:36<04:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2385/2820 [25:36<04:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2386/2820 [25:36<04:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2386/2820 [25:36<04:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2387/2820 [25:37<04:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2387/2820 [25:37<04:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2388/2820 [25:38<04:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2388/2820 [25:38<04:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2389/2820 [25:38<04:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2389/2820 [25:38<04:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2390/2820 [25:39<04:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2390/2820 [25:39<04:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2391/2820 [25:39<04:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2391/2820 [25:39<04:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2392/2820 [25:39<04:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2392/2820 [25:39<04:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2393/2820 [25:41<04:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2393/2820 [25:41<04:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2394/2820 [25:41<04:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2394/2820 [25:41<04:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2395/2820 [25:41<04:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2395/2820 [25:41<04:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2396/2820 [25:43<04:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▍ | 2396/2820 [25:43<04:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2397/2820 [25:43<04:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2397/2820 [25:43<04:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2398/2820 [25:43<04:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2398/2820 [25:43<04:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2399/2820 [25:45<04:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2399/2820 [25:45<04:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2400/2820 [25:45<04:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2400/2820 [25:45<04:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2401/2820 [25:45<04:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2401/2820 [25:45<04:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2402/2820 [25:47<04:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2402/2820 [25:47<04:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2403/2820 [25:47<04:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2403/2820 [25:47<04:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2404/2820 [25:47<04:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2404/2820 [25:47<04:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2405/2820 [25:49<04:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2405/2820 [25:49<04:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2406/2820 [25:49<04:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2406/2820 [25:49<04:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2407/2820 [25:49<04:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2407/2820 [25:49<04:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2408/2820 [25:51<04:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2408/2820 [25:51<04:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2409/2820 [25:51<04:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2409/2820 [25:51<04:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2410/2820 [25:51<04:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2410/2820 [25:51<04:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2411/2820 [25:53<04:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  85%|████████▌ | 2411/2820 [25:53<04:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2412/2820 [25:53<04:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2412/2820 [25:53<04:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2413/2820 [25:53<04:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2413/2820 [25:53<04:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2414/2820 [25:55<04:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2414/2820 [25:55<04:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2415/2820 [25:55<04:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2415/2820 [25:55<04:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2416/2820 [25:55<04:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2416/2820 [25:55<04:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2417/2820 [25:57<04:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2417/2820 [25:57<04:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2418/2820 [25:57<04:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2418/2820 [25:57<04:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2419/2820 [25:57<04:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2419/2820 [25:57<04:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2420/2820 [25:59<04:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2420/2820 [25:59<04:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2421/2820 [25:59<04:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2421/2820 [25:59<04:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2422/2820 [25:59<04:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2422/2820 [25:59<04:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2423/2820 [26:01<04:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2423/2820 [26:01<04:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2424/2820 [26:01<04:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2424/2820 [26:01<04:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2425/2820 [26:01<04:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2425/2820 [26:01<04:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2426/2820 [26:04<04:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2426/2820 [26:04<04:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2427/2820 [26:04<04:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2427/2820 [26:04<04:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2428/2820 [26:04<04:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2428/2820 [26:04<04:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2429/2820 [26:06<04:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2429/2820 [26:06<04:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2430/2820 [26:06<04:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2430/2820 [26:06<04:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2431/2820 [26:06<04:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2431/2820 [26:06<04:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2432/2820 [26:08<04:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▌ | 2432/2820 [26:08<04:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2433/2820 [26:08<04:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2433/2820 [26:08<04:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2434/2820 [26:08<04:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2434/2820 [26:08<04:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2435/2820 [26:10<04:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2435/2820 [26:10<04:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2436/2820 [26:10<04:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2436/2820 [26:10<04:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2437/2820 [26:10<04:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2437/2820 [26:10<04:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2438/2820 [26:12<04:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2438/2820 [26:12<04:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2439/2820 [26:12<04:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  86%|████████▋ | 2439/2820 [26:12<04:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2440/2820 [26:12<04:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2440/2820 [26:12<04:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2441/2820 [26:14<04:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2441/2820 [26:14<04:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2442/2820 [26:14<04:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2442/2820 [26:14<04:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2443/2820 [26:14<04:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2443/2820 [26:14<04:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2444/2820 [26:15<04:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2444/2820 [26:15<04:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2445/2820 [26:15<04:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2445/2820 [26:15<04:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2446/2820 [26:15<04:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2446/2820 [26:15<04:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2447/2820 [26:17<04:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2447/2820 [26:17<04:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2448/2820 [26:17<03:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2448/2820 [26:17<03:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2449/2820 [26:17<03:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2449/2820 [26:17<03:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2450/2820 [26:19<03:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2450/2820 [26:19<03:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2451/2820 [26:19<03:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2451/2820 [26:19<03:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2452/2820 [26:19<03:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2452/2820 [26:19<03:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2453/2820 [26:21<03:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2453/2820 [26:21<03:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2454/2820 [26:21<03:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2454/2820 [26:21<03:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2455/2820 [26:21<03:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2455/2820 [26:21<03:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2456/2820 [26:23<03:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2456/2820 [26:23<03:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2457/2820 [26:23<03:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2457/2820 [26:23<03:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2458/2820 [26:23<03:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2458/2820 [26:23<03:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2459/2820 [26:25<03:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2459/2820 [26:25<03:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2460/2820 [26:25<03:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2460/2820 [26:25<03:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2461/2820 [26:25<03:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2461/2820 [26:25<03:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2462/2820 [26:27<03:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2462/2820 [26:27<03:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2463/2820 [26:27<03:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2463/2820 [26:27<03:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2464/2820 [26:27<03:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2464/2820 [26:27<03:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2465/2820 [26:29<03:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2465/2820 [26:29<03:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2466/2820 [26:29<03:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2466/2820 [26:29<03:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2467/2820 [26:29<03:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  87%|████████▋ | 2467/2820 [26:29<03:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2468/2820 [26:31<03:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2468/2820 [26:31<03:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2469/2820 [26:31<03:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2469/2820 [26:31<03:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2470/2820 [26:31<03:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2470/2820 [26:31<03:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2471/2820 [26:33<03:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2471/2820 [26:33<03:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2472/2820 [26:33<03:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2472/2820 [26:33<03:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2473/2820 [26:33<03:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2473/2820 [26:33<03:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2474/2820 [26:35<03:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2474/2820 [26:35<03:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2475/2820 [26:35<03:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2475/2820 [26:35<03:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2476/2820 [26:35<03:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2476/2820 [26:35<03:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2477/2820 [26:37<03:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2477/2820 [26:37<03:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2478/2820 [26:37<03:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2478/2820 [26:37<03:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2479/2820 [26:37<03:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2479/2820 [26:37<03:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2480/2820 [26:39<03:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2480/2820 [26:39<03:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2481/2820 [26:39<03:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2481/2820 [26:39<03:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2482/2820 [26:39<03:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2482/2820 [26:39<03:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2483/2820 [26:42<03:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2483/2820 [26:42<03:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2484/2820 [26:42<03:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2484/2820 [26:42<03:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2485/2820 [26:42<03:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2485/2820 [26:42<03:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2486/2820 [26:44<03:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2486/2820 [26:44<03:35,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2487/2820 [26:44<03:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2487/2820 [26:44<03:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2488/2820 [26:44<03:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2488/2820 [26:44<03:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2489/2820 [26:46<03:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2489/2820 [26:46<03:33,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2490/2820 [26:46<03:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2490/2820 [26:46<03:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2491/2820 [26:46<03:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2491/2820 [26:46<03:32,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2492/2820 [26:48<03:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2492/2820 [26:48<03:31,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2493/2820 [26:48<03:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2493/2820 [26:48<03:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2494/2820 [26:48<03:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2494/2820 [26:48<03:30,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2495/2820 [26:50<03:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  88%|████████▊ | 2495/2820 [26:50<03:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2496/2820 [26:50<03:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2496/2820 [26:50<03:29,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2497/2820 [26:50<03:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2497/2820 [26:50<03:28,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2498/2820 [26:52<03:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2498/2820 [26:52<03:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2499/2820 [26:52<03:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2499/2820 [26:52<03:27,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2500/2820 [26:52<03:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2500/2820 [26:52<03:26,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2501/2820 [26:54<03:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2501/2820 [26:54<03:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2502/2820 [26:54<03:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▊ | 2502/2820 [26:54<03:25,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2503/2820 [26:54<03:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2503/2820 [26:54<03:24,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2504/2820 [26:56<03:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2504/2820 [26:56<03:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2505/2820 [26:56<03:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2505/2820 [26:56<03:23,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2506/2820 [26:56<03:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2506/2820 [26:56<03:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2507/2820 [26:58<03:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2507/2820 [26:58<03:22,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2508/2820 [26:58<03:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2508/2820 [26:58<03:21,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2509/2820 [26:58<03:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2509/2820 [26:58<03:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2510/2820 [27:00<03:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2510/2820 [27:00<03:20,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2511/2820 [27:00<03:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2511/2820 [27:00<03:19,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2512/2820 [27:00<03:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2512/2820 [27:00<03:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2513/2820 [27:02<03:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2513/2820 [27:02<03:18,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2514/2820 [27:02<03:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2514/2820 [27:02<03:17,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2515/2820 [27:02<03:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2515/2820 [27:02<03:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2516/2820 [27:04<03:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2516/2820 [27:04<03:16,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2517/2820 [27:04<03:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2517/2820 [27:04<03:15,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2518/2820 [27:04<03:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2518/2820 [27:04<03:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2519/2820 [27:06<03:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2519/2820 [27:06<03:14,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2520/2820 [27:06<03:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2520/2820 [27:06<03:13,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2521/2820 [27:06<03:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2521/2820 [27:06<03:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2522/2820 [27:08<03:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2522/2820 [27:08<03:12,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2523/2820 [27:08<03:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  89%|████████▉ | 2523/2820 [27:08<03:11,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2524/2820 [27:08<03:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2524/2820 [27:08<03:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2525/2820 [27:09<03:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2525/2820 [27:09<03:10,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2526/2820 [27:09<03:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2526/2820 [27:09<03:09,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2527/2820 [27:09<03:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2527/2820 [27:09<03:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2528/2820 [27:11<03:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2528/2820 [27:11<03:08,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2529/2820 [27:11<03:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2529/2820 [27:11<03:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2530/2820 [27:11<03:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2530/2820 [27:11<03:07,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2531/2820 [27:13<03:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2531/2820 [27:13<03:06,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2532/2820 [27:13<03:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2532/2820 [27:13<03:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2533/2820 [27:13<03:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2533/2820 [27:13<03:05,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2534/2820 [27:15<03:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2534/2820 [27:15<03:04,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2535/2820 [27:15<03:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2535/2820 [27:15<03:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2536/2820 [27:15<03:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2536/2820 [27:15<03:03,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2537/2820 [27:17<03:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|████████▉ | 2537/2820 [27:17<03:02,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2538/2820 [27:17<03:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2538/2820 [27:17<03:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2539/2820 [27:17<03:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2539/2820 [27:17<03:01,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2540/2820 [27:19<03:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2540/2820 [27:19<03:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2541/2820 [27:19<03:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2541/2820 [27:19<03:00,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2542/2820 [27:19<02:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2542/2820 [27:19<02:59,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2543/2820 [27:21<02:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2543/2820 [27:21<02:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2544/2820 [27:21<02:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2544/2820 [27:21<02:58,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2545/2820 [27:21<02:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2545/2820 [27:21<02:57,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2546/2820 [27:23<02:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2546/2820 [27:23<02:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2547/2820 [27:23<02:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2547/2820 [27:23<02:56,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2548/2820 [27:23<02:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2548/2820 [27:23<02:55,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2549/2820 [27:25<02:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2549/2820 [27:25<02:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2550/2820 [27:25<02:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2550/2820 [27:25<02:54,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2551/2820 [27:25<02:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2551/2820 [27:25<02:53,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2552/2820 [27:27<02:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  90%|█████████ | 2552/2820 [27:27<02:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2553/2820 [27:27<02:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2553/2820 [27:27<02:52,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2554/2820 [27:27<02:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2554/2820 [27:27<02:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2555/2820 [27:29<02:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2555/2820 [27:29<02:51,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2556/2820 [27:29<02:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2556/2820 [27:29<02:50,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2557/2820 [27:29<02:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2557/2820 [27:29<02:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2558/2820 [27:32<02:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2558/2820 [27:32<02:49,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2559/2820 [27:32<02:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2559/2820 [27:32<02:48,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2560/2820 [27:32<02:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2560/2820 [27:32<02:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2561/2820 [27:34<02:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2561/2820 [27:34<02:47,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2562/2820 [27:34<02:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2562/2820 [27:34<02:46,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2563/2820 [27:34<02:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2563/2820 [27:34<02:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2564/2820 [27:36<02:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2564/2820 [27:36<02:45,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2565/2820 [27:36<02:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2565/2820 [27:36<02:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2566/2820 [27:36<02:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2566/2820 [27:36<02:44,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2567/2820 [27:39<02:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2567/2820 [27:39<02:43,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2568/2820 [27:39<02:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2568/2820 [27:39<02:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2569/2820 [27:39<02:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2569/2820 [27:39<02:42,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2570/2820 [27:42<02:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2570/2820 [27:42<02:41,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2571/2820 [27:42<02:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2571/2820 [27:42<02:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2572/2820 [27:42<02:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2572/2820 [27:42<02:40,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2573/2820 [27:45<02:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████ | 2573/2820 [27:45<02:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2574/2820 [27:45<02:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2574/2820 [27:45<02:39,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2575/2820 [27:45<02:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2575/2820 [27:45<02:38,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2576/2820 [27:47<02:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2576/2820 [27:47<02:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2577/2820 [27:47<02:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2577/2820 [27:47<02:37,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2578/2820 [27:47<02:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2578/2820 [27:47<02:36,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2579/2820 [27:50<02:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2579/2820 [27:50<02:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2580/2820 [27:50<02:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  91%|█████████▏| 2580/2820 [27:50<02:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2581/2820 [27:50<02:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2581/2820 [27:50<02:34,  1.55it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2582/2820 [27:52<02:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2582/2820 [27:52<02:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2583/2820 [27:52<02:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2583/2820 [27:52<02:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2584/2820 [27:52<02:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2584/2820 [27:52<02:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2585/2820 [27:54<02:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2585/2820 [27:54<02:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2586/2820 [27:54<02:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2586/2820 [27:54<02:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2587/2820 [27:54<02:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2587/2820 [27:54<02:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2588/2820 [27:57<02:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2588/2820 [27:57<02:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2589/2820 [27:57<02:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2589/2820 [27:57<02:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2590/2820 [27:57<02:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2590/2820 [27:57<02:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2591/2820 [28:00<02:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2591/2820 [28:00<02:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2592/2820 [28:00<02:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2592/2820 [28:00<02:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2593/2820 [28:00<02:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2593/2820 [28:00<02:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2594/2820 [28:04<02:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2594/2820 [28:04<02:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2595/2820 [28:04<02:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2595/2820 [28:04<02:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2596/2820 [28:04<02:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2596/2820 [28:04<02:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2597/2820 [28:06<02:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2597/2820 [28:06<02:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2598/2820 [28:06<02:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2598/2820 [28:06<02:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2599/2820 [28:06<02:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2599/2820 [28:06<02:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2600/2820 [28:08<02:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2600/2820 [28:08<02:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2601/2820 [28:08<02:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2601/2820 [28:08<02:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2602/2820 [28:08<02:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2602/2820 [28:08<02:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2603/2820 [28:10<02:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2603/2820 [28:10<02:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2604/2820 [28:10<02:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2604/2820 [28:10<02:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2605/2820 [28:10<02:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2605/2820 [28:10<02:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2606/2820 [28:12<02:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2606/2820 [28:12<02:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2607/2820 [28:12<02:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2607/2820 [28:12<02:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2608/2820 [28:12<02:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  92%|█████████▏| 2608/2820 [28:12<02:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2609/2820 [28:14<02:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2609/2820 [28:14<02:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2610/2820 [28:14<02:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2610/2820 [28:14<02:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2611/2820 [28:14<02:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2611/2820 [28:14<02:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2612/2820 [28:16<02:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2612/2820 [28:16<02:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2613/2820 [28:16<02:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2613/2820 [28:16<02:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2614/2820 [28:16<02:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2614/2820 [28:16<02:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2615/2820 [28:18<02:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2615/2820 [28:18<02:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2616/2820 [28:18<02:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2616/2820 [28:18<02:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2617/2820 [28:18<02:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2617/2820 [28:18<02:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2618/2820 [28:20<02:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2618/2820 [28:20<02:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2619/2820 [28:20<02:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2619/2820 [28:20<02:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2620/2820 [28:20<02:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2620/2820 [28:20<02:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2621/2820 [28:22<02:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2621/2820 [28:22<02:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2622/2820 [28:22<02:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2622/2820 [28:22<02:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2623/2820 [28:22<02:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2623/2820 [28:22<02:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2624/2820 [28:24<02:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2624/2820 [28:24<02:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2625/2820 [28:24<02:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2625/2820 [28:24<02:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2626/2820 [28:24<02:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2626/2820 [28:24<02:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2627/2820 [28:26<02:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2627/2820 [28:26<02:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2628/2820 [28:26<02:04,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2628/2820 [28:26<02:04,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2629/2820 [28:26<02:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2629/2820 [28:26<02:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2630/2820 [28:28<02:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2630/2820 [28:28<02:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2631/2820 [28:28<02:02,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2631/2820 [28:28<02:02,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2632/2820 [28:28<02:02,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2632/2820 [28:28<02:02,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2633/2820 [28:30<02:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2633/2820 [28:30<02:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2634/2820 [28:30<02:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2634/2820 [28:30<02:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2635/2820 [28:30<02:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2635/2820 [28:30<02:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2636/2820 [28:32<01:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  93%|█████████▎| 2636/2820 [28:32<01:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2637/2820 [28:32<01:58,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2637/2820 [28:32<01:58,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2638/2820 [28:32<01:58,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2638/2820 [28:32<01:58,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2639/2820 [28:34<01:57,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2639/2820 [28:34<01:57,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2640/2820 [28:34<01:56,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2640/2820 [28:34<01:56,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2641/2820 [28:34<01:56,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2641/2820 [28:34<01:56,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2642/2820 [28:36<01:55,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2642/2820 [28:36<01:55,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2643/2820 [28:36<01:54,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▎| 2643/2820 [28:36<01:54,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2644/2820 [28:36<01:54,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2644/2820 [28:36<01:54,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2645/2820 [28:38<01:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2645/2820 [28:38<01:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2646/2820 [28:38<01:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2646/2820 [28:38<01:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2647/2820 [28:38<01:52,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2647/2820 [28:38<01:52,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2648/2820 [28:41<01:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2648/2820 [28:41<01:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2649/2820 [28:41<01:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2649/2820 [28:41<01:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2650/2820 [28:41<01:50,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2650/2820 [28:41<01:50,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2651/2820 [28:43<01:49,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2651/2820 [28:43<01:49,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2652/2820 [28:43<01:49,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2652/2820 [28:43<01:49,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2653/2820 [28:43<01:48,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2653/2820 [28:43<01:48,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2654/2820 [28:45<01:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2654/2820 [28:45<01:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2655/2820 [28:45<01:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2655/2820 [28:45<01:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2656/2820 [28:45<01:46,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2656/2820 [28:45<01:46,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2657/2820 [28:47<01:45,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2657/2820 [28:47<01:45,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2658/2820 [28:47<01:45,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2658/2820 [28:47<01:45,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2659/2820 [28:47<01:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2659/2820 [28:47<01:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2660/2820 [28:49<01:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2660/2820 [28:49<01:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2661/2820 [28:49<01:43,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2661/2820 [28:49<01:43,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2662/2820 [28:49<01:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2662/2820 [28:49<01:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2663/2820 [28:51<01:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2663/2820 [28:51<01:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2664/2820 [28:51<01:41,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  94%|█████████▍| 2664/2820 [28:51<01:41,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2665/2820 [28:51<01:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2665/2820 [28:51<01:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2666/2820 [28:53<01:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2666/2820 [28:53<01:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2667/2820 [28:53<01:39,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2667/2820 [28:53<01:39,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2668/2820 [28:53<01:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2668/2820 [28:53<01:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2669/2820 [28:55<01:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2669/2820 [28:55<01:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2670/2820 [28:55<01:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2670/2820 [28:55<01:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2671/2820 [28:55<01:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2671/2820 [28:55<01:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2672/2820 [28:57<01:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2672/2820 [28:57<01:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2673/2820 [28:57<01:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2673/2820 [28:57<01:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2674/2820 [28:57<01:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2674/2820 [28:57<01:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2675/2820 [28:59<01:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2675/2820 [28:59<01:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2676/2820 [28:59<01:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2676/2820 [28:59<01:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2677/2820 [28:59<01:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2677/2820 [28:59<01:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2678/2820 [29:01<01:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▍| 2678/2820 [29:01<01:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2679/2820 [29:01<01:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2679/2820 [29:01<01:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2680/2820 [29:01<01:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2680/2820 [29:01<01:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2681/2820 [29:03<01:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2681/2820 [29:03<01:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2682/2820 [29:03<01:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2682/2820 [29:03<01:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2683/2820 [29:03<01:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2683/2820 [29:03<01:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2684/2820 [29:05<01:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2684/2820 [29:05<01:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2685/2820 [29:05<01:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2685/2820 [29:05<01:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2686/2820 [29:05<01:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2686/2820 [29:05<01:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2687/2820 [29:07<01:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2687/2820 [29:07<01:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2688/2820 [29:07<01:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2688/2820 [29:07<01:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2689/2820 [29:07<01:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2689/2820 [29:07<01:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2690/2820 [29:08<01:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2690/2820 [29:08<01:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2691/2820 [29:08<01:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2691/2820 [29:08<01:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2692/2820 [29:08<01:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2692/2820 [29:08<01:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2693/2820 [29:10<01:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  95%|█████████▌| 2693/2820 [29:10<01:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2694/2820 [29:10<01:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2694/2820 [29:10<01:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2695/2820 [29:10<01:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2695/2820 [29:10<01:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2696/2820 [29:12<01:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2696/2820 [29:12<01:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2697/2820 [29:12<01:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2697/2820 [29:12<01:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2698/2820 [29:12<01:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2698/2820 [29:12<01:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2699/2820 [29:14<01:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2699/2820 [29:14<01:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2700/2820 [29:14<01:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2700/2820 [29:14<01:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2701/2820 [29:14<01:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2701/2820 [29:14<01:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2702/2820 [29:16<01:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2702/2820 [29:16<01:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2703/2820 [29:16<01:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2703/2820 [29:16<01:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2704/2820 [29:16<01:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2704/2820 [29:16<01:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2705/2820 [29:18<01:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2705/2820 [29:18<01:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2706/2820 [29:18<01:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2706/2820 [29:18<01:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2707/2820 [29:18<01:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2707/2820 [29:18<01:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2708/2820 [29:20<01:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2708/2820 [29:20<01:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2709/2820 [29:20<01:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2709/2820 [29:20<01:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2710/2820 [29:20<01:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2710/2820 [29:20<01:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2711/2820 [29:22<01:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2711/2820 [29:22<01:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2712/2820 [29:22<01:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2712/2820 [29:22<01:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2713/2820 [29:22<01:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2713/2820 [29:22<01:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2714/2820 [29:24<01:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▌| 2714/2820 [29:24<01:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2715/2820 [29:24<01:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2715/2820 [29:24<01:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2716/2820 [29:24<01:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2716/2820 [29:24<01:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2717/2820 [29:26<01:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2717/2820 [29:26<01:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2718/2820 [29:26<01:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2718/2820 [29:26<01:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2719/2820 [29:26<01:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2719/2820 [29:26<01:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2720/2820 [29:28<01:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2720/2820 [29:28<01:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2721/2820 [29:28<01:04,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  96%|█████████▋| 2721/2820 [29:28<01:04,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2722/2820 [29:28<01:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2722/2820 [29:28<01:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2723/2820 [29:29<01:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2723/2820 [29:29<01:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2724/2820 [29:29<01:02,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2724/2820 [29:29<01:02,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2725/2820 [29:30<01:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2725/2820 [29:30<01:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2726/2820 [29:31<01:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2726/2820 [29:31<01:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2727/2820 [29:31<01:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2727/2820 [29:31<01:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2728/2820 [29:31<00:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2728/2820 [29:31<00:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2729/2820 [29:33<00:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2729/2820 [29:33<00:59,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2730/2820 [29:33<00:58,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2730/2820 [29:33<00:58,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2731/2820 [29:33<00:57,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2731/2820 [29:33<00:57,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2732/2820 [29:35<00:57,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2732/2820 [29:35<00:57,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2733/2820 [29:35<00:56,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2733/2820 [29:35<00:56,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2734/2820 [29:35<00:55,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2734/2820 [29:35<00:55,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2735/2820 [29:37<00:55,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2735/2820 [29:37<00:55,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2736/2820 [29:37<00:54,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2736/2820 [29:37<00:54,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2737/2820 [29:37<00:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2737/2820 [29:37<00:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2738/2820 [29:39<00:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2738/2820 [29:39<00:53,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2739/2820 [29:39<00:52,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2739/2820 [29:39<00:52,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2740/2820 [29:39<00:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2740/2820 [29:39<00:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2741/2820 [29:41<00:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2741/2820 [29:41<00:51,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2742/2820 [29:41<00:50,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2742/2820 [29:41<00:50,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2743/2820 [29:41<00:50,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2743/2820 [29:41<00:50,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2744/2820 [29:43<00:49,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2744/2820 [29:43<00:49,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2745/2820 [29:43<00:48,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2745/2820 [29:43<00:48,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2746/2820 [29:43<00:48,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2746/2820 [29:43<00:48,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2747/2820 [29:45<00:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2747/2820 [29:45<00:47,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2748/2820 [29:45<00:46,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2748/2820 [29:45<00:46,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2749/2820 [29:45<00:46,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  97%|█████████▋| 2749/2820 [29:45<00:46,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2750/2820 [29:47<00:45,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2750/2820 [29:47<00:45,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2751/2820 [29:47<00:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2751/2820 [29:47<00:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2752/2820 [29:47<00:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2752/2820 [29:47<00:44,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2753/2820 [29:48<00:43,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2753/2820 [29:48<00:43,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2754/2820 [29:48<00:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2754/2820 [29:48<00:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2755/2820 [29:49<00:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2755/2820 [29:49<00:42,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2756/2820 [29:50<00:41,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2756/2820 [29:50<00:41,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2757/2820 [29:50<00:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2757/2820 [29:50<00:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2758/2820 [29:50<00:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2758/2820 [29:50<00:40,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2759/2820 [29:52<00:39,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2759/2820 [29:52<00:39,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2760/2820 [29:52<00:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2760/2820 [29:52<00:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2761/2820 [29:52<00:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2761/2820 [29:52<00:38,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2762/2820 [29:54<00:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2762/2820 [29:54<00:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2763/2820 [29:54<00:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2763/2820 [29:54<00:37,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2764/2820 [29:54<00:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2764/2820 [29:54<00:36,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2765/2820 [29:57<00:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2765/2820 [29:57<00:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2766/2820 [29:57<00:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2766/2820 [29:57<00:35,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2767/2820 [29:57<00:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2767/2820 [29:57<00:34,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2768/2820 [29:59<00:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2768/2820 [29:59<00:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2769/2820 [29:59<00:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2769/2820 [29:59<00:33,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2770/2820 [29:59<00:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2770/2820 [29:59<00:32,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2771/2820 [30:01<00:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2771/2820 [30:01<00:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2772/2820 [30:01<00:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2772/2820 [30:01<00:31,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2773/2820 [30:01<00:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2773/2820 [30:01<00:30,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2774/2820 [30:03<00:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2774/2820 [30:03<00:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2775/2820 [30:03<00:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2775/2820 [30:03<00:29,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2776/2820 [30:03<00:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2776/2820 [30:03<00:28,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2777/2820 [30:05<00:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  98%|█████████▊| 2777/2820 [30:05<00:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2778/2820 [30:05<00:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2778/2820 [30:05<00:27,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2779/2820 [30:05<00:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2779/2820 [30:05<00:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2780/2820 [30:08<00:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2780/2820 [30:08<00:26,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2781/2820 [30:08<00:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2781/2820 [30:08<00:25,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2782/2820 [30:08<00:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2782/2820 [30:08<00:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2783/2820 [30:10<00:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2783/2820 [30:10<00:24,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2784/2820 [30:10<00:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▊| 2784/2820 [30:10<00:23,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2785/2820 [30:10<00:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2785/2820 [30:10<00:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2786/2820 [30:12<00:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2786/2820 [30:12<00:22,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2787/2820 [30:12<00:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2787/2820 [30:12<00:21,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2788/2820 [30:12<00:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2788/2820 [30:12<00:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2789/2820 [30:14<00:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2789/2820 [30:14<00:20,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2790/2820 [30:14<00:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2790/2820 [30:14<00:19,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2791/2820 [30:14<00:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2791/2820 [30:14<00:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2792/2820 [30:16<00:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2792/2820 [30:16<00:18,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2793/2820 [30:16<00:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2793/2820 [30:16<00:17,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2794/2820 [30:16<00:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2794/2820 [30:16<00:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2795/2820 [30:18<00:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2795/2820 [30:18<00:16,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2796/2820 [30:18<00:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2796/2820 [30:18<00:15,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2797/2820 [30:18<00:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2797/2820 [30:18<00:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2798/2820 [30:21<00:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2798/2820 [30:21<00:14,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2799/2820 [30:21<00:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2799/2820 [30:21<00:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2800/2820 [30:21<00:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2800/2820 [30:21<00:13,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2801/2820 [30:23<00:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2801/2820 [30:23<00:12,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2802/2820 [30:23<00:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2802/2820 [30:23<00:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2803/2820 [30:23<00:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2803/2820 [30:23<00:11,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2804/2820 [30:26<00:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2804/2820 [30:26<00:10,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2805/2820 [30:26<00:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4:  99%|█████████▉| 2805/2820 [30:26<00:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2806/2820 [30:26<00:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2806/2820 [30:26<00:09,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2807/2820 [30:28<00:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2807/2820 [30:28<00:08,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2808/2820 [30:28<00:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2808/2820 [30:28<00:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2809/2820 [30:28<00:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2809/2820 [30:28<00:07,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2810/2820 [30:30<00:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2810/2820 [30:30<00:06,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2811/2820 [30:30<00:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2811/2820 [30:30<00:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2812/2820 [30:30<00:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2812/2820 [30:30<00:05,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2813/2820 [30:32<00:04,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2813/2820 [30:32<00:04,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2814/2820 [30:32<00:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2814/2820 [30:32<00:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2815/2820 [30:32<00:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2815/2820 [30:32<00:03,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2816/2820 [30:34<00:02,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2816/2820 [30:34<00:02,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2817/2820 [30:34<00:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2817/2820 [30:34<00:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2818/2820 [30:34<00:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2818/2820 [30:34<00:01,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2819/2820 [30:36<00:00,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|█████████▉| 2819/2820 [30:36<00:00,  1.53it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|██████████| 2820/2820 [30:36<00:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|██████████| 2820/2820 [30:36<00:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.112, R@1 (train)=0.983]
Epoch 4: 100%|██████████| 2820/2820 [30:36<00:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.0888, R@1 (train)=0.990]Epoch 4, global step 16000: 'R@1 (val)' was not in top 1
`Trainer.fit` stopped: `max_steps=16000` reached.

Epoch 4: 100%|██████████| 2820/2820 [30:36<00:00,  1.54it/s, R@1 (val)=0.407, R@5 (val)=0.610, R@20 (val)=0.793, loss (train)=0.0888, R@1 (train)=0.990]Restoring states from the checkpoint path at /Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=03-R@1 (val)=0.407.ckpt
Loaded model weights from the checkpoint at /Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=03-R@1 (val)=0.407.ckpt


Validation: |          | 0/? [00:00<?, ?it/s]
Validation: |          | 0/? [00:00<?, ?it/s]
Validation DataLoader 0:   0%|          | 0/602 [00:00<?, ?it/s]
Validation DataLoader 0:   0%|          | 1/602 [00:00<00:21, 28.30it/s]
Validation DataLoader 0:   0%|          | 2/602 [00:00<00:17, 35.05it/s]
Validation DataLoader 0:   0%|          | 3/602 [00:00<00:14, 41.38it/s]
Validation DataLoader 0:   1%|          | 4/602 [00:00<00:44, 13.39it/s]
Validation DataLoader 0:   1%|          | 5/602 [00:00<00:37, 15.98it/s]
Validation DataLoader 0:   1%|          | 6/602 [00:00<00:50, 11.79it/s]
Validation DataLoader 0:   1%|          | 7/602 [00:00<01:07,  8.83it/s]
Validation DataLoader 0:   1%|▏         | 8/602 [00:00<00:59,  9.92it/s]
Validation DataLoader 0:   1%|▏         | 9/602 [00:00<01:04,  9.16it/s]
Validation DataLoader 0:   2%|▏         | 10/602 [00:01<01:11,  8.28it/s]
Validation DataLoader 0:   2%|▏         | 11/602 [00:01<01:05,  9.01it/s]
Validation DataLoader 0:   2%|▏         | 12/602 [00:01<01:08,  8.61it/s]
Validation DataLoader 0:   2%|▏         | 13/602 [00:01<01:20,  7.36it/s]
Validation DataLoader 0:   2%|▏         | 14/602 [00:01<01:14,  7.85it/s]
Validation DataLoader 0:   2%|▏         | 15/602 [00:01<01:10,  8.28it/s]
Validation DataLoader 0:   3%|▎         | 16/602 [00:02<01:24,  6.91it/s]
Validation DataLoader 0:   3%|▎         | 17/602 [00:02<01:20,  7.30it/s]
Validation DataLoader 0:   3%|▎         | 18/602 [00:02<01:16,  7.68it/s]
Validation DataLoader 0:   3%|▎         | 19/602 [00:02<01:27,  6.67it/s]
Validation DataLoader 0:   3%|▎         | 20/602 [00:02<01:23,  6.99it/s]
Validation DataLoader 0:   3%|▎         | 21/602 [00:02<01:19,  7.30it/s]
Validation DataLoader 0:   4%|▎         | 22/602 [00:03<01:26,  6.68it/s]
Validation DataLoader 0:   4%|▍         | 23/602 [00:03<01:23,  6.95it/s]
Validation DataLoader 0:   4%|▍         | 24/602 [00:03<01:19,  7.23it/s]
Validation DataLoader 0:   4%|▍         | 25/602 [00:03<01:26,  6.67it/s]
Validation DataLoader 0:   4%|▍         | 26/602 [00:03<01:23,  6.91it/s]
Validation DataLoader 0:   4%|▍         | 27/602 [00:03<01:20,  7.14it/s]
Validation DataLoader 0:   5%|▍         | 28/602 [00:04<01:25,  6.68it/s]
Validation DataLoader 0:   5%|▍         | 29/602 [00:04<01:23,  6.90it/s]
Validation DataLoader 0:   5%|▍         | 30/602 [00:04<01:20,  7.11it/s]
Validation DataLoader 0:   5%|▌         | 31/602 [00:04<01:26,  6.64it/s]
Validation DataLoader 0:   5%|▌         | 32/602 [00:04<01:23,  6.83it/s]
Validation DataLoader 0:   5%|▌         | 33/602 [00:04<01:21,  7.02it/s]
Validation DataLoader 0:   6%|▌         | 34/602 [00:05<01:26,  6.58it/s]
Validation DataLoader 0:   6%|▌         | 35/602 [00:05<01:23,  6.76it/s]
Validation DataLoader 0:   6%|▌         | 36/602 [00:05<01:21,  6.93it/s]
Validation DataLoader 0:   6%|▌         | 37/602 [00:05<01:24,  6.69it/s]
Validation DataLoader 0:   6%|▋         | 38/602 [00:05<01:22,  6.86it/s]
Validation DataLoader 0:   6%|▋         | 39/602 [00:05<01:20,  7.02it/s]
Validation DataLoader 0:   7%|▋         | 40/602 [00:12<02:53,  3.25it/s]
Validation DataLoader 0:   7%|▋         | 41/602 [00:12<02:49,  3.31it/s]
Validation DataLoader 0:   7%|▋         | 42/602 [00:12<02:45,  3.39it/s]
Validation DataLoader 0:   7%|▋         | 43/602 [00:12<02:41,  3.47it/s]
Validation DataLoader 0:   7%|▋         | 44/602 [00:12<02:37,  3.54it/s]
Validation DataLoader 0:   7%|▋         | 45/602 [00:12<02:34,  3.62it/s]
Validation DataLoader 0:   8%|▊         | 46/602 [00:12<02:30,  3.69it/s]
Validation DataLoader 0:   8%|▊         | 47/602 [00:12<02:31,  3.67it/s]
Validation DataLoader 0:   8%|▊         | 48/602 [00:13<02:30,  3.69it/s]
Validation DataLoader 0:   8%|▊         | 49/602 [00:13<02:27,  3.76it/s]
Validation DataLoader 0:   8%|▊         | 50/602 [00:13<02:27,  3.75it/s]
Validation DataLoader 0:   8%|▊         | 51/602 [00:13<02:24,  3.80it/s]
Validation DataLoader 0:   9%|▊         | 52/602 [00:13<02:22,  3.87it/s]
Validation DataLoader 0:   9%|▉         | 53/602 [00:13<02:23,  3.83it/s]
Validation DataLoader 0:   9%|▉         | 54/602 [00:13<02:20,  3.90it/s]
Validation DataLoader 0:   9%|▉         | 55/602 [00:13<02:17,  3.96it/s]
Validation DataLoader 0:   9%|▉         | 56/602 [00:14<02:19,  3.90it/s]
Validation DataLoader 0:   9%|▉         | 57/602 [00:14<02:17,  3.97it/s]
Validation DataLoader 0:  10%|▉         | 58/602 [00:14<02:14,  4.03it/s]
Validation DataLoader 0:  10%|▉         | 59/602 [00:14<02:15,  4.00it/s]
Validation DataLoader 0:  10%|▉         | 60/602 [00:14<02:13,  4.07it/s]
Validation DataLoader 0:  10%|█         | 61/602 [00:14<02:10,  4.13it/s]
Validation DataLoader 0:  10%|█         | 62/602 [00:15<02:12,  4.07it/s]
Validation DataLoader 0:  10%|█         | 63/602 [00:15<02:10,  4.13it/s]
Validation DataLoader 0:  11%|█         | 64/602 [00:15<02:08,  4.19it/s]
Validation DataLoader 0:  11%|█         | 65/602 [00:15<02:09,  4.14it/s]
Validation DataLoader 0:  11%|█         | 66/602 [00:15<02:07,  4.20it/s]
Validation DataLoader 0:  11%|█         | 67/602 [00:15<02:05,  4.26it/s]
Validation DataLoader 0:  11%|█▏        | 68/602 [00:16<02:06,  4.21it/s]
Validation DataLoader 0:  11%|█▏        | 69/602 [00:16<02:04,  4.27it/s]
Validation DataLoader 0:  12%|█▏        | 70/602 [00:16<02:02,  4.33it/s]
Validation DataLoader 0:  12%|█▏        | 71/602 [00:16<02:04,  4.25it/s]
Validation DataLoader 0:  12%|█▏        | 72/602 [00:16<02:03,  4.31it/s]
Validation DataLoader 0:  12%|█▏        | 73/602 [00:16<02:01,  4.36it/s]
Validation DataLoader 0:  12%|█▏        | 74/602 [00:17<02:02,  4.31it/s]
Validation DataLoader 0:  12%|█▏        | 75/602 [00:17<02:00,  4.37it/s]
Validation DataLoader 0:  13%|█▎        | 76/602 [00:17<01:59,  4.42it/s]
Validation DataLoader 0:  13%|█▎        | 77/602 [00:17<01:59,  4.38it/s]
Validation DataLoader 0:  13%|█▎        | 78/602 [00:17<01:58,  4.43it/s]
Validation DataLoader 0:  13%|█▎        | 79/602 [00:17<01:56,  4.48it/s]
Validation DataLoader 0:  13%|█▎        | 80/602 [00:18<01:57,  4.43it/s]
Validation DataLoader 0:  13%|█▎        | 81/602 [00:18<01:56,  4.48it/s]
Validation DataLoader 0:  14%|█▎        | 82/602 [00:18<01:54,  4.53it/s]
Validation DataLoader 0:  14%|█▍        | 83/602 [00:18<01:55,  4.48it/s]
Validation DataLoader 0:  14%|█▍        | 84/602 [00:18<01:54,  4.53it/s]
Validation DataLoader 0:  14%|█▍        | 85/602 [00:18<01:52,  4.58it/s]
Validation DataLoader 0:  14%|█▍        | 86/602 [00:18<01:53,  4.53it/s]
Validation DataLoader 0:  14%|█▍        | 87/602 [00:19<01:52,  4.58it/s]
Validation DataLoader 0:  15%|█▍        | 88/602 [00:19<01:51,  4.63it/s]
Validation DataLoader 0:  15%|█▍        | 89/602 [00:19<01:51,  4.59it/s]
Validation DataLoader 0:  15%|█▍        | 90/602 [00:19<01:50,  4.64it/s]
Validation DataLoader 0:  15%|█▌        | 91/602 [00:19<01:49,  4.68it/s]
Validation DataLoader 0:  15%|█▌        | 92/602 [00:19<01:49,  4.64it/s]
Validation DataLoader 0:  15%|█▌        | 93/602 [00:19<01:48,  4.69it/s]
Validation DataLoader 0:  16%|█▌        | 94/602 [00:19<01:47,  4.73it/s]
Validation DataLoader 0:  16%|█▌        | 95/602 [00:20<01:48,  4.69it/s]
Validation DataLoader 0:  16%|█▌        | 96/602 [00:20<01:46,  4.74it/s]
Validation DataLoader 0:  16%|█▌        | 97/602 [00:20<01:45,  4.79it/s]
Validation DataLoader 0:  16%|█▋        | 98/602 [00:20<01:46,  4.75it/s]
Validation DataLoader 0:  16%|█▋        | 99/602 [00:20<01:44,  4.80it/s]
Validation DataLoader 0:  17%|█▋        | 100/602 [00:20<01:43,  4.84it/s]
Validation DataLoader 0:  17%|█▋        | 101/602 [00:21<01:44,  4.80it/s]
Validation DataLoader 0:  17%|█▋        | 102/602 [00:21<01:43,  4.85it/s]
Validation DataLoader 0:  17%|█▋        | 103/602 [00:21<01:42,  4.89it/s]
Validation DataLoader 0:  17%|█▋        | 104/602 [00:21<01:42,  4.85it/s]
Validation DataLoader 0:  17%|█▋        | 105/602 [00:21<01:41,  4.90it/s]
Validation DataLoader 0:  18%|█▊        | 106/602 [00:21<01:40,  4.94it/s]
Validation DataLoader 0:  18%|█▊        | 107/602 [00:21<01:41,  4.88it/s]
Validation DataLoader 0:  18%|█▊        | 108/602 [00:21<01:40,  4.93it/s]
Validation DataLoader 0:  18%|█▊        | 109/602 [00:21<01:39,  4.97it/s]
Validation DataLoader 0:  18%|█▊        | 110/602 [00:22<01:39,  4.93it/s]
Validation DataLoader 0:  18%|█▊        | 111/602 [00:22<01:38,  4.98it/s]
Validation DataLoader 0:  19%|█▊        | 112/602 [00:22<01:37,  5.02it/s]
Validation DataLoader 0:  19%|█▉        | 113/602 [00:22<01:38,  4.96it/s]
Validation DataLoader 0:  19%|█▉        | 114/602 [00:22<01:37,  5.00it/s]
Validation DataLoader 0:  19%|█▉        | 115/602 [00:22<01:36,  5.04it/s]
Validation DataLoader 0:  19%|█▉        | 116/602 [00:23<01:37,  5.01it/s]
Validation DataLoader 0:  19%|█▉        | 117/602 [00:23<01:36,  5.05it/s]
Validation DataLoader 0:  20%|█▉        | 118/602 [00:23<01:35,  5.09it/s]
Validation DataLoader 0:  20%|█▉        | 119/602 [00:23<01:35,  5.04it/s]
Validation DataLoader 0:  20%|█▉        | 120/602 [00:23<01:34,  5.08it/s]
Validation DataLoader 0:  20%|██        | 121/602 [00:23<01:34,  5.11it/s]
Validation DataLoader 0:  20%|██        | 122/602 [00:24<01:34,  5.08it/s]
Validation DataLoader 0:  20%|██        | 123/602 [00:24<01:33,  5.11it/s]
Validation DataLoader 0:  21%|██        | 124/602 [00:24<01:32,  5.15it/s]
Validation DataLoader 0:  21%|██        | 125/602 [00:24<01:33,  5.10it/s]
Validation DataLoader 0:  21%|██        | 126/602 [00:24<01:32,  5.14it/s]
Validation DataLoader 0:  21%|██        | 127/602 [00:24<01:31,  5.18it/s]
Validation DataLoader 0:  21%|██▏       | 128/602 [00:24<01:32,  5.12it/s]
Validation DataLoader 0:  21%|██▏       | 129/602 [00:25<01:31,  5.16it/s]
Validation DataLoader 0:  22%|██▏       | 130/602 [00:25<01:30,  5.19it/s]
Validation DataLoader 0:  22%|██▏       | 131/602 [00:25<01:31,  5.16it/s]
Validation DataLoader 0:  22%|██▏       | 132/602 [00:25<01:30,  5.20it/s]
Validation DataLoader 0:  22%|██▏       | 133/602 [00:25<01:29,  5.22it/s]
Validation DataLoader 0:  22%|██▏       | 134/602 [00:25<01:30,  5.18it/s]
Validation DataLoader 0:  22%|██▏       | 135/602 [00:25<01:29,  5.22it/s]
Validation DataLoader 0:  23%|██▎       | 136/602 [00:25<01:28,  5.25it/s]
Validation DataLoader 0:  23%|██▎       | 137/602 [00:26<01:29,  5.19it/s]
Validation DataLoader 0:  23%|██▎       | 138/602 [00:26<01:28,  5.22it/s]
Validation DataLoader 0:  23%|██▎       | 139/602 [00:26<01:28,  5.26it/s]
Validation DataLoader 0:  23%|██▎       | 140/602 [00:26<01:28,  5.23it/s]
Validation DataLoader 0:  23%|██▎       | 141/602 [00:26<01:27,  5.27it/s]
Validation DataLoader 0:  24%|██▎       | 142/602 [00:26<01:26,  5.29it/s]
Validation DataLoader 0:  24%|██▍       | 143/602 [00:27<01:27,  5.24it/s]
Validation DataLoader 0:  24%|██▍       | 144/602 [00:27<01:26,  5.28it/s]
Validation DataLoader 0:  24%|██▍       | 145/602 [00:27<01:26,  5.31it/s]
Validation DataLoader 0:  24%|██▍       | 146/602 [00:27<01:26,  5.27it/s]
Validation DataLoader 0:  24%|██▍       | 147/602 [00:27<01:25,  5.30it/s]
Validation DataLoader 0:  25%|██▍       | 148/602 [00:27<01:25,  5.34it/s]
Validation DataLoader 0:  25%|██▍       | 149/602 [00:28<01:25,  5.28it/s]
Validation DataLoader 0:  25%|██▍       | 150/602 [00:28<01:25,  5.32it/s]
Validation DataLoader 0:  25%|██▌       | 151/602 [00:28<01:24,  5.35it/s]
Validation DataLoader 0:  25%|██▌       | 152/602 [00:28<01:24,  5.30it/s]
Validation DataLoader 0:  25%|██▌       | 153/602 [00:28<01:24,  5.33it/s]
Validation DataLoader 0:  26%|██▌       | 154/602 [00:28<01:23,  5.36it/s]
Validation DataLoader 0:  26%|██▌       | 155/602 [00:29<01:23,  5.33it/s]
Validation DataLoader 0:  26%|██▌       | 156/602 [00:29<01:23,  5.36it/s]
Validation DataLoader 0:  26%|██▌       | 157/602 [00:29<01:22,  5.37it/s]
Validation DataLoader 0:  26%|██▌       | 158/602 [00:29<01:22,  5.35it/s]
Validation DataLoader 0:  26%|██▋       | 159/602 [00:29<01:22,  5.38it/s]
Validation DataLoader 0:  27%|██▋       | 160/602 [00:29<01:22,  5.38it/s]
Validation DataLoader 0:  27%|██▋       | 161/602 [00:30<01:22,  5.37it/s]
Validation DataLoader 0:  27%|██▋       | 162/602 [00:30<01:21,  5.40it/s]
Validation DataLoader 0:  27%|██▋       | 163/602 [00:30<01:21,  5.39it/s]
Validation DataLoader 0:  27%|██▋       | 164/602 [00:30<01:21,  5.38it/s]
Validation DataLoader 0:  27%|██▋       | 165/602 [00:30<01:20,  5.41it/s]
Validation DataLoader 0:  28%|██▊       | 166/602 [00:30<01:20,  5.42it/s]
Validation DataLoader 0:  28%|██▊       | 167/602 [00:30<01:20,  5.40it/s]
Validation DataLoader 0:  28%|██▊       | 168/602 [00:30<01:19,  5.43it/s]
Validation DataLoader 0:  28%|██▊       | 169/602 [00:30<01:19,  5.46it/s]
Validation DataLoader 0:  28%|██▊       | 170/602 [00:31<01:19,  5.41it/s]
Validation DataLoader 0:  28%|██▊       | 171/602 [00:31<01:19,  5.44it/s]
Validation DataLoader 0:  29%|██▊       | 172/602 [00:31<01:18,  5.47it/s]
Validation DataLoader 0:  29%|██▊       | 173/602 [00:31<01:19,  5.43it/s]
Validation DataLoader 0:  29%|██▉       | 174/602 [00:31<01:18,  5.46it/s]
Validation DataLoader 0:  29%|██▉       | 175/602 [00:31<01:17,  5.49it/s]
Validation DataLoader 0:  29%|██▉       | 176/602 [00:32<01:18,  5.44it/s]
Validation DataLoader 0:  29%|██▉       | 177/602 [00:32<01:17,  5.47it/s]
Validation DataLoader 0:  30%|██▉       | 178/602 [00:32<01:17,  5.49it/s]
Validation DataLoader 0:  30%|██▉       | 179/602 [00:32<01:17,  5.45it/s]
Validation DataLoader 0:  30%|██▉       | 180/602 [00:32<01:16,  5.48it/s]
Validation DataLoader 0:  30%|███       | 181/602 [00:32<01:16,  5.50it/s]
Validation DataLoader 0:  30%|███       | 182/602 [00:33<01:16,  5.46it/s]
Validation DataLoader 0:  30%|███       | 183/602 [00:33<01:16,  5.49it/s]
Validation DataLoader 0:  31%|███       | 184/602 [00:33<01:15,  5.51it/s]
Validation DataLoader 0:  31%|███       | 185/602 [00:33<01:16,  5.48it/s]
Validation DataLoader 0:  31%|███       | 186/602 [00:33<01:15,  5.51it/s]
Validation DataLoader 0:  31%|███       | 187/602 [00:33<01:14,  5.54it/s]
Validation DataLoader 0:  31%|███       | 188/602 [00:34<01:15,  5.50it/s]
Validation DataLoader 0:  31%|███▏      | 189/602 [00:34<01:14,  5.53it/s]
Validation DataLoader 0:  32%|███▏      | 190/602 [00:34<01:14,  5.54it/s]
Validation DataLoader 0:  32%|███▏      | 191/602 [00:34<01:14,  5.52it/s]
Validation DataLoader 0:  32%|███▏      | 192/602 [00:34<01:13,  5.55it/s]
Validation DataLoader 0:  32%|███▏      | 193/602 [00:34<01:13,  5.57it/s]
Validation DataLoader 0:  32%|███▏      | 194/602 [00:35<01:13,  5.53it/s]
Validation DataLoader 0:  32%|███▏      | 195/602 [00:35<01:13,  5.55it/s]
Validation DataLoader 0:  33%|███▎      | 196/602 [00:35<01:12,  5.58it/s]
Validation DataLoader 0:  33%|███▎      | 197/602 [00:35<01:13,  5.54it/s]
Validation DataLoader 0:  33%|███▎      | 198/602 [00:35<01:12,  5.57it/s]
Validation DataLoader 0:  33%|███▎      | 199/602 [00:35<01:12,  5.58it/s]
Validation DataLoader 0:  33%|███▎      | 200/602 [00:36<01:12,  5.55it/s]
Validation DataLoader 0:  33%|███▎      | 201/602 [00:36<01:11,  5.58it/s]
Validation DataLoader 0:  34%|███▎      | 202/602 [00:36<01:11,  5.59it/s]
Validation DataLoader 0:  34%|███▎      | 203/602 [00:36<01:11,  5.58it/s]
Validation DataLoader 0:  34%|███▍      | 204/602 [00:36<01:11,  5.60it/s]
Validation DataLoader 0:  34%|███▍      | 205/602 [00:36<01:10,  5.60it/s]
Validation DataLoader 0:  34%|███▍      | 206/602 [00:36<01:10,  5.60it/s]
Validation DataLoader 0:  34%|███▍      | 207/602 [00:36<01:10,  5.63it/s]
Validation DataLoader 0:  35%|███▍      | 208/602 [00:37<01:10,  5.62it/s]
Validation DataLoader 0:  35%|███▍      | 209/602 [00:37<01:10,  5.61it/s]
Validation DataLoader 0:  35%|███▍      | 210/602 [00:37<01:09,  5.64it/s]
Validation DataLoader 0:  35%|███▌      | 211/602 [00:37<01:09,  5.63it/s]
Validation DataLoader 0:  35%|███▌      | 212/602 [00:37<01:09,  5.63it/s]
Validation DataLoader 0:  35%|███▌      | 213/602 [00:37<01:08,  5.65it/s]
Validation DataLoader 0:  36%|███▌      | 214/602 [00:37<01:08,  5.65it/s]
Validation DataLoader 0:  36%|███▌      | 215/602 [00:38<01:08,  5.64it/s]
Validation DataLoader 0:  36%|███▌      | 216/602 [00:38<01:08,  5.66it/s]
Validation DataLoader 0:  36%|███▌      | 217/602 [00:38<01:07,  5.66it/s]
Validation DataLoader 0:  36%|███▌      | 218/602 [00:38<01:08,  5.65it/s]
Validation DataLoader 0:  36%|███▋      | 219/602 [00:38<01:07,  5.67it/s]
Validation DataLoader 0:  37%|███▋      | 220/602 [00:38<01:07,  5.65it/s]
Validation DataLoader 0:  37%|███▋      | 221/602 [00:39<01:07,  5.64it/s]
Validation DataLoader 0:  37%|███▋      | 222/602 [00:39<01:07,  5.67it/s]
Validation DataLoader 0:  37%|███▋      | 223/602 [00:39<01:06,  5.67it/s]
Validation DataLoader 0:  37%|███▋      | 224/602 [00:39<01:06,  5.66it/s]
Validation DataLoader 0:  37%|███▋      | 225/602 [00:39<01:06,  5.68it/s]
Validation DataLoader 0:  38%|███▊      | 226/602 [00:39<01:06,  5.67it/s]
Validation DataLoader 0:  38%|███▊      | 227/602 [00:40<01:06,  5.67it/s]
Validation DataLoader 0:  38%|███▊      | 228/602 [00:40<01:05,  5.69it/s]
Validation DataLoader 0:  38%|███▊      | 229/602 [00:40<01:05,  5.68it/s]
Validation DataLoader 0:  38%|███▊      | 230/602 [00:40<01:05,  5.67it/s]
Validation DataLoader 0:  38%|███▊      | 231/602 [00:40<01:05,  5.69it/s]
Validation DataLoader 0:  39%|███▊      | 232/602 [00:40<01:05,  5.68it/s]
Validation DataLoader 0:  39%|███▊      | 233/602 [00:40<01:04,  5.69it/s]
Validation DataLoader 0:  39%|███▉      | 234/602 [00:40<01:04,  5.71it/s]
Validation DataLoader 0:  39%|███▉      | 235/602 [00:41<01:04,  5.69it/s]
Validation DataLoader 0:  39%|███▉      | 236/602 [00:41<01:04,  5.70it/s]
Validation DataLoader 0:  39%|███▉      | 237/602 [00:41<01:03,  5.72it/s]
Validation DataLoader 0:  40%|███▉      | 238/602 [00:41<01:03,  5.71it/s]
Validation DataLoader 0:  40%|███▉      | 239/602 [00:41<01:03,  5.72it/s]
Validation DataLoader 0:  40%|███▉      | 240/602 [00:41<01:03,  5.74it/s]
Validation DataLoader 0:  40%|████      | 241/602 [00:42<01:03,  5.72it/s]
Validation DataLoader 0:  40%|████      | 242/602 [00:42<01:02,  5.73it/s]
Validation DataLoader 0:  40%|████      | 243/602 [00:42<01:02,  5.75it/s]
Validation DataLoader 0:  41%|████      | 244/602 [00:42<01:02,  5.74it/s]
Validation DataLoader 0:  41%|████      | 245/602 [00:42<01:02,  5.73it/s]
Validation DataLoader 0:  41%|████      | 246/602 [00:42<01:01,  5.75it/s]
Validation DataLoader 0:  41%|████      | 247/602 [00:42<01:01,  5.75it/s]
Validation DataLoader 0:  41%|████      | 248/602 [00:43<01:01,  5.73it/s]
Validation DataLoader 0:  41%|████▏     | 249/602 [00:43<01:01,  5.75it/s]
Validation DataLoader 0:  42%|████▏     | 250/602 [00:43<01:01,  5.76it/s]
Validation DataLoader 0:  42%|████▏     | 251/602 [00:43<01:01,  5.75it/s]
Validation DataLoader 0:  42%|████▏     | 252/602 [00:43<01:00,  5.77it/s]
Validation DataLoader 0:  42%|████▏     | 253/602 [00:43<01:00,  5.77it/s]
Validation DataLoader 0:  42%|████▏     | 254/602 [00:44<01:00,  5.75it/s]
Validation DataLoader 0:  42%|████▏     | 255/602 [00:44<01:00,  5.77it/s]
Validation DataLoader 0:  43%|████▎     | 256/602 [00:44<00:59,  5.79it/s]
Validation DataLoader 0:  43%|████▎     | 257/602 [00:44<01:00,  5.75it/s]
Validation DataLoader 0:  43%|████▎     | 258/602 [00:44<00:59,  5.77it/s]
Validation DataLoader 0:  43%|████▎     | 259/602 [00:44<00:59,  5.79it/s]
Validation DataLoader 0:  43%|████▎     | 260/602 [00:45<00:59,  5.75it/s]
Validation DataLoader 0:  43%|████▎     | 261/602 [00:45<00:59,  5.77it/s]
Validation DataLoader 0:  44%|████▎     | 262/602 [00:45<00:58,  5.79it/s]
Validation DataLoader 0:  44%|████▎     | 263/602 [00:45<00:58,  5.77it/s]
Validation DataLoader 0:  44%|████▍     | 264/602 [00:45<00:58,  5.79it/s]
Validation DataLoader 0:  44%|████▍     | 265/602 [00:45<00:57,  5.81it/s]
Validation DataLoader 0:  44%|████▍     | 266/602 [00:46<00:58,  5.78it/s]
Validation DataLoader 0:  44%|████▍     | 267/602 [00:46<00:57,  5.80it/s]
Validation DataLoader 0:  45%|████▍     | 268/602 [00:46<00:57,  5.82it/s]
Validation DataLoader 0:  45%|████▍     | 269/602 [00:46<00:57,  5.78it/s]
Validation DataLoader 0:  45%|████▍     | 270/602 [00:46<00:57,  5.80it/s]
Validation DataLoader 0:  45%|████▌     | 271/602 [00:46<00:56,  5.82it/s]
Validation DataLoader 0:  45%|████▌     | 272/602 [00:46<00:57,  5.79it/s]
Validation DataLoader 0:  45%|████▌     | 273/602 [00:47<00:56,  5.81it/s]
Validation DataLoader 0:  46%|████▌     | 274/602 [00:47<00:56,  5.83it/s]
Validation DataLoader 0:  46%|████▌     | 275/602 [00:47<00:56,  5.79it/s]
Validation DataLoader 0:  46%|████▌     | 276/602 [00:47<00:56,  5.81it/s]
Validation DataLoader 0:  46%|████▌     | 277/602 [00:47<00:55,  5.83it/s]
Validation DataLoader 0:  46%|████▌     | 278/602 [00:47<00:55,  5.80it/s]
Validation DataLoader 0:  46%|████▋     | 279/602 [00:47<00:55,  5.82it/s]
Validation DataLoader 0:  47%|████▋     | 280/602 [00:47<00:55,  5.84it/s]
Validation DataLoader 0:  47%|████▋     | 281/602 [00:48<00:55,  5.79it/s]
Validation DataLoader 0:  47%|████▋     | 282/602 [00:48<00:55,  5.81it/s]
Validation DataLoader 0:  47%|████▋     | 283/602 [00:48<00:54,  5.83it/s]
Validation DataLoader 0:  47%|████▋     | 284/602 [00:48<00:54,  5.80it/s]
Validation DataLoader 0:  47%|████▋     | 285/602 [00:48<00:54,  5.82it/s]
Validation DataLoader 0:  48%|████▊     | 286/602 [00:49<00:54,  5.84it/s]
Validation DataLoader 0:  48%|████▊     | 287/602 [00:49<00:54,  5.80it/s]
Validation DataLoader 0:  48%|████▊     | 288/602 [00:49<00:53,  5.82it/s]
Validation DataLoader 0:  48%|████▊     | 289/602 [00:49<00:53,  5.84it/s]
Validation DataLoader 0:  48%|████▊     | 290/602 [00:49<00:53,  5.80it/s]
Validation DataLoader 0:  48%|████▊     | 291/602 [00:49<00:53,  5.82it/s]
Validation DataLoader 0:  49%|████▊     | 292/602 [00:49<00:53,  5.84it/s]
Validation DataLoader 0:  49%|████▊     | 293/602 [00:50<00:53,  5.81it/s]
Validation DataLoader 0:  49%|████▉     | 294/602 [00:50<00:52,  5.83it/s]
Validation DataLoader 0:  49%|████▉     | 295/602 [00:50<00:52,  5.85it/s]
Validation DataLoader 0:  49%|████▉     | 296/602 [00:50<00:52,  5.82it/s]
Validation DataLoader 0:  49%|████▉     | 297/602 [00:50<00:52,  5.83it/s]
Validation DataLoader 0:  50%|████▉     | 298/602 [00:50<00:51,  5.85it/s]
Validation DataLoader 0:  50%|████▉     | 299/602 [00:51<00:52,  5.82it/s]
Validation DataLoader 0:  50%|████▉     | 300/602 [00:51<00:51,  5.84it/s]
Validation DataLoader 0:  50%|█████     | 301/602 [00:51<00:51,  5.86it/s]
Validation DataLoader 0:  50%|█████     | 302/602 [00:51<00:51,  5.82it/s]
Validation DataLoader 0:  50%|█████     | 303/602 [00:51<00:51,  5.83it/s]
Validation DataLoader 0:  50%|█████     | 304/602 [00:51<00:50,  5.85it/s]
Validation DataLoader 0:  51%|█████     | 305/602 [00:52<00:51,  5.82it/s]
Validation DataLoader 0:  51%|█████     | 306/602 [00:52<00:50,  5.84it/s]
Validation DataLoader 0:  51%|█████     | 307/602 [00:52<00:50,  5.85it/s]
Validation DataLoader 0:  51%|█████     | 308/602 [00:52<00:50,  5.83it/s]
Validation DataLoader 0:  51%|█████▏    | 309/602 [00:52<00:50,  5.85it/s]
Validation DataLoader 0:  51%|█████▏    | 310/602 [00:52<00:49,  5.86it/s]
Validation DataLoader 0:  52%|█████▏    | 311/602 [00:53<00:49,  5.84it/s]
Validation DataLoader 0:  52%|█████▏    | 312/602 [00:53<00:49,  5.86it/s]
Validation DataLoader 0:  52%|█████▏    | 313/602 [00:53<00:49,  5.88it/s]
Validation DataLoader 0:  52%|█████▏    | 314/602 [00:53<00:49,  5.84it/s]
Validation DataLoader 0:  52%|█████▏    | 315/602 [00:53<00:48,  5.86it/s]
Validation DataLoader 0:  52%|█████▏    | 316/602 [00:53<00:48,  5.88it/s]
Validation DataLoader 0:  53%|█████▎    | 317/602 [00:54<00:48,  5.85it/s]
Validation DataLoader 0:  53%|█████▎    | 318/602 [00:54<00:48,  5.87it/s]
Validation DataLoader 0:  53%|█████▎    | 319/602 [00:54<00:48,  5.89it/s]
Validation DataLoader 0:  53%|█████▎    | 320/602 [00:54<00:48,  5.85it/s]
Validation DataLoader 0:  53%|█████▎    | 321/602 [00:54<00:47,  5.87it/s]
Validation DataLoader 0:  53%|█████▎    | 322/602 [00:54<00:47,  5.89it/s]
Validation DataLoader 0:  54%|█████▎    | 323/602 [00:55<00:47,  5.85it/s]
Validation DataLoader 0:  54%|█████▍    | 324/602 [00:55<00:47,  5.87it/s]
Validation DataLoader 0:  54%|█████▍    | 325/602 [00:55<00:47,  5.89it/s]
Validation DataLoader 0:  54%|█████▍    | 326/602 [00:55<00:47,  5.86it/s]
Validation DataLoader 0:  54%|█████▍    | 327/602 [00:55<00:46,  5.88it/s]
Validation DataLoader 0:  54%|█████▍    | 328/602 [00:55<00:46,  5.90it/s]
Validation DataLoader 0:  55%|█████▍    | 329/602 [00:56<00:46,  5.87it/s]
Validation DataLoader 0:  55%|█████▍    | 330/602 [00:56<00:46,  5.88it/s]
Validation DataLoader 0:  55%|█████▍    | 331/602 [00:56<00:45,  5.90it/s]
Validation DataLoader 0:  55%|█████▌    | 332/602 [00:56<00:45,  5.88it/s]
Validation DataLoader 0:  55%|█████▌    | 333/602 [00:56<00:45,  5.89it/s]
Validation DataLoader 0:  55%|█████▌    | 334/602 [00:56<00:45,  5.91it/s]
Validation DataLoader 0:  56%|█████▌    | 335/602 [00:56<00:45,  5.88it/s]
Validation DataLoader 0:  56%|█████▌    | 336/602 [00:56<00:45,  5.90it/s]
Validation DataLoader 0:  56%|█████▌    | 337/602 [00:56<00:44,  5.92it/s]
Validation DataLoader 0:  56%|█████▌    | 338/602 [00:57<00:44,  5.89it/s]
Validation DataLoader 0:  56%|█████▋    | 339/602 [00:57<00:44,  5.91it/s]
Validation DataLoader 0:  56%|█████▋    | 340/602 [00:57<00:44,  5.92it/s]
Validation DataLoader 0:  57%|█████▋    | 341/602 [00:57<00:44,  5.90it/s]
Validation DataLoader 0:  57%|█████▋    | 342/602 [00:57<00:43,  5.91it/s]
Validation DataLoader 0:  57%|█████▋    | 343/602 [00:57<00:43,  5.93it/s]
Validation DataLoader 0:  57%|█████▋    | 344/602 [00:58<00:43,  5.91it/s]
Validation DataLoader 0:  57%|█████▋    | 345/602 [00:58<00:43,  5.93it/s]
Validation DataLoader 0:  57%|█████▋    | 346/602 [00:58<00:43,  5.95it/s]
Validation DataLoader 0:  58%|█████▊    | 347/602 [00:58<00:43,  5.92it/s]
Validation DataLoader 0:  58%|█████▊    | 348/602 [00:58<00:42,  5.94it/s]
Validation DataLoader 0:  58%|█████▊    | 349/602 [00:58<00:42,  5.95it/s]
Validation DataLoader 0:  58%|█████▊    | 350/602 [00:59<00:42,  5.92it/s]
Validation DataLoader 0:  58%|█████▊    | 351/602 [00:59<00:42,  5.94it/s]
Validation DataLoader 0:  58%|█████▊    | 352/602 [00:59<00:41,  5.95it/s]
Validation DataLoader 0:  59%|█████▊    | 353/602 [00:59<00:42,  5.92it/s]
Validation DataLoader 0:  59%|█████▉    | 354/602 [00:59<00:41,  5.94it/s]
Validation DataLoader 0:  59%|█████▉    | 355/602 [00:59<00:41,  5.95it/s]
Validation DataLoader 0:  59%|█████▉    | 356/602 [01:00<00:41,  5.93it/s]
Validation DataLoader 0:  59%|█████▉    | 357/602 [01:00<00:41,  5.94it/s]
Validation DataLoader 0:  59%|█████▉    | 358/602 [01:00<00:40,  5.96it/s]
Validation DataLoader 0:  60%|█████▉    | 359/602 [01:00<00:40,  5.94it/s]
Validation DataLoader 0:  60%|█████▉    | 360/602 [01:00<00:40,  5.95it/s]
Validation DataLoader 0:  60%|█████▉    | 361/602 [01:00<00:40,  5.97it/s]
Validation DataLoader 0:  60%|██████    | 362/602 [01:00<00:40,  5.94it/s]
Validation DataLoader 0:  60%|██████    | 363/602 [01:00<00:40,  5.95it/s]
Validation DataLoader 0:  60%|██████    | 364/602 [01:01<00:39,  5.97it/s]
Validation DataLoader 0:  61%|██████    | 365/602 [01:01<00:39,  5.94it/s]
Validation DataLoader 0:  61%|██████    | 366/602 [01:01<00:39,  5.96it/s]
Validation DataLoader 0:  61%|██████    | 367/602 [01:01<00:39,  5.97it/s]
Validation DataLoader 0:  61%|██████    | 368/602 [01:01<00:39,  5.95it/s]
Validation DataLoader 0:  61%|██████▏   | 369/602 [01:01<00:39,  5.96it/s]
Validation DataLoader 0:  61%|██████▏   | 370/602 [01:01<00:38,  5.98it/s]
Validation DataLoader 0:  62%|██████▏   | 371/602 [01:02<00:38,  5.95it/s]
Validation DataLoader 0:  62%|██████▏   | 372/602 [01:02<00:38,  5.96it/s]
Validation DataLoader 0:  62%|██████▏   | 373/602 [01:02<00:38,  5.98it/s]
Validation DataLoader 0:  62%|██████▏   | 374/602 [01:02<00:38,  5.95it/s]
Validation DataLoader 0:  62%|██████▏   | 375/602 [01:02<00:38,  5.97it/s]
Validation DataLoader 0:  62%|██████▏   | 376/602 [01:02<00:37,  5.98it/s]
Validation DataLoader 0:  63%|██████▎   | 377/602 [01:03<00:37,  5.96it/s]
Validation DataLoader 0:  63%|██████▎   | 378/602 [01:03<00:37,  5.97it/s]
Validation DataLoader 0:  63%|██████▎   | 379/602 [01:03<00:37,  5.98it/s]
Validation DataLoader 0:  63%|██████▎   | 380/602 [01:03<00:37,  5.96it/s]
Validation DataLoader 0:  63%|██████▎   | 381/602 [01:03<00:37,  5.97it/s]
Validation DataLoader 0:  63%|██████▎   | 382/602 [01:03<00:36,  5.99it/s]
Validation DataLoader 0:  64%|██████▎   | 383/602 [01:04<00:36,  5.96it/s]
Validation DataLoader 0:  64%|██████▍   | 384/602 [01:04<00:36,  5.98it/s]
Validation DataLoader 0:  64%|██████▍   | 385/602 [01:04<00:36,  5.99it/s]
Validation DataLoader 0:  64%|██████▍   | 386/602 [01:04<00:36,  5.97it/s]
Validation DataLoader 0:  64%|██████▍   | 387/602 [01:04<00:35,  5.98it/s]
Validation DataLoader 0:  64%|██████▍   | 388/602 [01:04<00:35,  6.00it/s]
Validation DataLoader 0:  65%|██████▍   | 389/602 [01:05<00:35,  5.98it/s]
Validation DataLoader 0:  65%|██████▍   | 390/602 [01:05<00:35,  5.99it/s]
Validation DataLoader 0:  65%|██████▍   | 391/602 [01:05<00:35,  6.00it/s]
Validation DataLoader 0:  65%|██████▌   | 392/602 [01:05<00:35,  5.99it/s]
Validation DataLoader 0:  65%|██████▌   | 393/602 [01:05<00:34,  6.00it/s]
Validation DataLoader 0:  65%|██████▌   | 394/602 [01:05<00:34,  6.02it/s]
Validation DataLoader 0:  66%|██████▌   | 395/602 [01:05<00:34,  6.00it/s]
Validation DataLoader 0:  66%|██████▌   | 396/602 [01:05<00:34,  6.01it/s]
Validation DataLoader 0:  66%|██████▌   | 397/602 [01:05<00:34,  6.02it/s]
Validation DataLoader 0:  66%|██████▌   | 398/602 [01:06<00:34,  6.00it/s]
Validation DataLoader 0:  66%|██████▋   | 399/602 [01:06<00:33,  6.00it/s]
Validation DataLoader 0:  66%|██████▋   | 400/602 [01:06<00:33,  6.02it/s]
Validation DataLoader 0:  67%|██████▋   | 401/602 [01:06<00:33,  6.00it/s]
Validation DataLoader 0:  67%|██████▋   | 402/602 [01:06<00:33,  6.01it/s]
Validation DataLoader 0:  67%|██████▋   | 403/602 [01:06<00:33,  6.02it/s]
Validation DataLoader 0:  67%|██████▋   | 404/602 [01:07<00:32,  6.00it/s]
Validation DataLoader 0:  67%|██████▋   | 405/602 [01:07<00:32,  6.01it/s]
Validation DataLoader 0:  67%|██████▋   | 406/602 [01:07<00:32,  6.03it/s]
Validation DataLoader 0:  68%|██████▊   | 407/602 [01:07<00:32,  6.00it/s]
Validation DataLoader 0:  68%|██████▊   | 408/602 [01:07<00:32,  6.01it/s]
Validation DataLoader 0:  68%|██████▊   | 409/602 [01:07<00:32,  6.03it/s]
Validation DataLoader 0:  68%|██████▊   | 410/602 [01:08<00:31,  6.00it/s]
Validation DataLoader 0:  68%|██████▊   | 411/602 [01:08<00:31,  6.02it/s]
Validation DataLoader 0:  68%|██████▊   | 412/602 [01:08<00:31,  6.03it/s]
Validation DataLoader 0:  69%|██████▊   | 413/602 [01:08<00:31,  6.01it/s]
Validation DataLoader 0:  69%|██████▉   | 414/602 [01:08<00:31,  6.02it/s]
Validation DataLoader 0:  69%|██████▉   | 415/602 [01:08<00:30,  6.03it/s]
Validation DataLoader 0:  69%|██████▉   | 416/602 [01:09<00:30,  6.01it/s]
Validation DataLoader 0:  69%|██████▉   | 417/602 [01:09<00:30,  6.02it/s]
Validation DataLoader 0:  69%|██████▉   | 418/602 [01:09<00:30,  6.04it/s]
Validation DataLoader 0:  70%|██████▉   | 419/602 [01:09<00:30,  6.01it/s]
Validation DataLoader 0:  70%|██████▉   | 420/602 [01:09<00:30,  6.02it/s]
Validation DataLoader 0:  70%|██████▉   | 421/602 [01:09<00:29,  6.04it/s]
Validation DataLoader 0:  70%|███████   | 422/602 [01:10<00:29,  6.02it/s]
Validation DataLoader 0:  70%|███████   | 423/602 [01:10<00:29,  6.03it/s]
Validation DataLoader 0:  70%|███████   | 424/602 [01:10<00:29,  6.04it/s]
Validation DataLoader 0:  71%|███████   | 425/602 [01:10<00:29,  6.02it/s]
Validation DataLoader 0:  71%|███████   | 426/602 [01:10<00:29,  6.03it/s]
Validation DataLoader 0:  71%|███████   | 427/602 [01:10<00:28,  6.04it/s]
Validation DataLoader 0:  71%|███████   | 428/602 [01:10<00:28,  6.03it/s]
Validation DataLoader 0:  71%|███████▏  | 429/602 [01:10<00:28,  6.04it/s]
Validation DataLoader 0:  71%|███████▏  | 430/602 [01:11<00:28,  6.06it/s]
Validation DataLoader 0:  72%|███████▏  | 431/602 [01:11<00:28,  6.03it/s]
Validation DataLoader 0:  72%|███████▏  | 432/602 [01:11<00:28,  6.04it/s]
Validation DataLoader 0:  72%|███████▏  | 433/602 [01:11<00:27,  6.05it/s]
Validation DataLoader 0:  72%|███████▏  | 434/602 [01:11<00:27,  6.03it/s]
Validation DataLoader 0:  72%|███████▏  | 435/602 [01:11<00:27,  6.05it/s]
Validation DataLoader 0:  72%|███████▏  | 436/602 [01:11<00:27,  6.06it/s]
Validation DataLoader 0:  73%|███████▎  | 437/602 [01:12<00:27,  6.04it/s]
Validation DataLoader 0:  73%|███████▎  | 438/602 [01:12<00:27,  6.05it/s]
Validation DataLoader 0:  73%|███████▎  | 439/602 [01:12<00:26,  6.06it/s]
Validation DataLoader 0:  73%|███████▎  | 440/602 [01:12<00:26,  6.04it/s]
Validation DataLoader 0:  73%|███████▎  | 441/602 [01:12<00:26,  6.05it/s]
Validation DataLoader 0:  73%|███████▎  | 442/602 [01:12<00:26,  6.06it/s]
Validation DataLoader 0:  74%|███████▎  | 443/602 [01:13<00:26,  6.04it/s]
Validation DataLoader 0:  74%|███████▍  | 444/602 [01:13<00:26,  6.05it/s]
Validation DataLoader 0:  74%|███████▍  | 445/602 [01:13<00:25,  6.07it/s]
Validation DataLoader 0:  74%|███████▍  | 446/602 [01:13<00:25,  6.05it/s]
Validation DataLoader 0:  74%|███████▍  | 447/602 [01:13<00:25,  6.06it/s]
Validation DataLoader 0:  74%|███████▍  | 448/602 [01:13<00:25,  6.07it/s]
Validation DataLoader 0:  75%|███████▍  | 449/602 [01:14<00:25,  6.05it/s]
Validation DataLoader 0:  75%|███████▍  | 450/602 [01:14<00:25,  6.06it/s]
Validation DataLoader 0:  75%|███████▍  | 451/602 [01:14<00:24,  6.07it/s]
Validation DataLoader 0:  75%|███████▌  | 452/602 [01:14<00:24,  6.04it/s]
Validation DataLoader 0:  75%|███████▌  | 453/602 [01:14<00:24,  6.06it/s]
Validation DataLoader 0:  75%|███████▌  | 454/602 [01:14<00:24,  6.07it/s]
Validation DataLoader 0:  76%|███████▌  | 455/602 [01:15<00:24,  6.04it/s]
Validation DataLoader 0:  76%|███████▌  | 456/602 [01:15<00:24,  6.05it/s]
Validation DataLoader 0:  76%|███████▌  | 457/602 [01:15<00:23,  6.07it/s]
Validation DataLoader 0:  76%|███████▌  | 458/602 [01:15<00:23,  6.03it/s]
Validation DataLoader 0:  76%|███████▌  | 459/602 [01:15<00:23,  6.05it/s]
Validation DataLoader 0:  76%|███████▋  | 460/602 [01:15<00:23,  6.06it/s]
Validation DataLoader 0:  77%|███████▋  | 461/602 [01:16<00:23,  6.04it/s]
Validation DataLoader 0:  77%|███████▋  | 462/602 [01:16<00:23,  6.05it/s]
Validation DataLoader 0:  77%|███████▋  | 463/602 [01:16<00:22,  6.06it/s]
Validation DataLoader 0:  77%|███████▋  | 464/602 [01:16<00:22,  6.04it/s]
Validation DataLoader 0:  77%|███████▋  | 465/602 [01:16<00:22,  6.06it/s]
Validation DataLoader 0:  77%|███████▋  | 466/602 [01:16<00:22,  6.07it/s]
Validation DataLoader 0:  78%|███████▊  | 467/602 [01:17<00:22,  6.05it/s]
Validation DataLoader 0:  78%|███████▊  | 468/602 [01:17<00:22,  6.06it/s]
Validation DataLoader 0:  78%|███████▊  | 469/602 [01:17<00:21,  6.07it/s]
Validation DataLoader 0:  78%|███████▊  | 470/602 [01:17<00:21,  6.04it/s]
Validation DataLoader 0:  78%|███████▊  | 471/602 [01:17<00:21,  6.05it/s]
Validation DataLoader 0:  78%|███████▊  | 472/602 [01:17<00:21,  6.07it/s]
Validation DataLoader 0:  79%|███████▊  | 473/602 [01:18<00:21,  6.05it/s]
Validation DataLoader 0:  79%|███████▊  | 474/602 [01:18<00:21,  6.06it/s]
Validation DataLoader 0:  79%|███████▉  | 475/602 [01:18<00:20,  6.07it/s]
Validation DataLoader 0:  79%|███████▉  | 476/602 [01:18<00:20,  6.05it/s]
Validation DataLoader 0:  79%|███████▉  | 477/602 [01:18<00:20,  6.06it/s]
Validation DataLoader 0:  79%|███████▉  | 478/602 [01:18<00:20,  6.07it/s]
Validation DataLoader 0:  80%|███████▉  | 479/602 [01:19<00:20,  6.05it/s]
Validation DataLoader 0:  80%|███████▉  | 480/602 [01:19<00:20,  6.06it/s]
Validation DataLoader 0:  80%|███████▉  | 481/602 [01:19<00:19,  6.07it/s]
Validation DataLoader 0:  80%|████████  | 482/602 [01:19<00:19,  6.05it/s]
Validation DataLoader 0:  80%|████████  | 483/602 [01:19<00:19,  6.06it/s]
Validation DataLoader 0:  80%|████████  | 484/602 [01:19<00:19,  6.07it/s]
Validation DataLoader 0:  81%|████████  | 485/602 [01:20<00:19,  6.05it/s]
Validation DataLoader 0:  81%|████████  | 486/602 [01:20<00:19,  6.06it/s]
Validation DataLoader 0:  81%|████████  | 487/602 [01:20<00:18,  6.07it/s]
Validation DataLoader 0:  81%|████████  | 488/602 [01:20<00:18,  6.05it/s]
Validation DataLoader 0:  81%|████████  | 489/602 [01:20<00:18,  6.07it/s]
Validation DataLoader 0:  81%|████████▏ | 490/602 [01:20<00:18,  6.08it/s]
Validation DataLoader 0:  82%|████████▏ | 491/602 [01:21<00:18,  6.06it/s]
Validation DataLoader 0:  82%|████████▏ | 492/602 [01:21<00:18,  6.07it/s]
Validation DataLoader 0:  82%|████████▏ | 493/602 [01:21<00:17,  6.08it/s]
Validation DataLoader 0:  82%|████████▏ | 494/602 [01:21<00:17,  6.06it/s]
Validation DataLoader 0:  82%|████████▏ | 495/602 [01:21<00:17,  6.08it/s]
Validation DataLoader 0:  82%|████████▏ | 496/602 [01:21<00:17,  6.09it/s]
Validation DataLoader 0:  83%|████████▎ | 497/602 [01:21<00:17,  6.07it/s]
Validation DataLoader 0:  83%|████████▎ | 498/602 [01:21<00:17,  6.08it/s]
Validation DataLoader 0:  83%|████████▎ | 499/602 [01:21<00:16,  6.09it/s]
Validation DataLoader 0:  83%|████████▎ | 500/602 [01:22<00:16,  6.07it/s]
Validation DataLoader 0:  83%|████████▎ | 501/602 [01:22<00:16,  6.08it/s]
Validation DataLoader 0:  83%|████████▎ | 502/602 [01:22<00:16,  6.09it/s]
Validation DataLoader 0:  84%|████████▎ | 503/602 [01:22<00:16,  6.08it/s]
Validation DataLoader 0:  84%|████████▎ | 504/602 [01:22<00:16,  6.09it/s]
Validation DataLoader 0:  84%|████████▍ | 505/602 [01:22<00:15,  6.10it/s]
Validation DataLoader 0:  84%|████████▍ | 506/602 [01:23<00:15,  6.08it/s]
Validation DataLoader 0:  84%|████████▍ | 507/602 [01:23<00:15,  6.09it/s]
Validation DataLoader 0:  84%|████████▍ | 508/602 [01:23<00:15,  6.10it/s]
Validation DataLoader 0:  85%|████████▍ | 509/602 [01:23<00:15,  6.08it/s]
Validation DataLoader 0:  85%|████████▍ | 510/602 [01:23<00:15,  6.09it/s]
Validation DataLoader 0:  85%|████████▍ | 511/602 [01:23<00:14,  6.10it/s]
Validation DataLoader 0:  85%|████████▌ | 512/602 [01:24<00:14,  6.08it/s]
Validation DataLoader 0:  85%|████████▌ | 513/602 [01:24<00:14,  6.09it/s]
Validation DataLoader 0:  85%|████████▌ | 514/602 [01:24<00:14,  6.10it/s]
Validation DataLoader 0:  86%|████████▌ | 515/602 [01:24<00:14,  6.08it/s]
Validation DataLoader 0:  86%|████████▌ | 516/602 [01:24<00:14,  6.09it/s]
Validation DataLoader 0:  86%|████████▌ | 517/602 [01:24<00:13,  6.10it/s]
Validation DataLoader 0:  86%|████████▌ | 518/602 [01:25<00:13,  6.09it/s]
Validation DataLoader 0:  86%|████████▌ | 519/602 [01:25<00:13,  6.10it/s]
Validation DataLoader 0:  86%|████████▋ | 520/602 [01:25<00:13,  6.11it/s]
Validation DataLoader 0:  87%|████████▋ | 521/602 [01:25<00:13,  6.09it/s]
Validation DataLoader 0:  87%|████████▋ | 522/602 [01:25<00:13,  6.10it/s]
Validation DataLoader 0:  87%|████████▋ | 523/602 [01:25<00:12,  6.11it/s]
Validation DataLoader 0:  87%|████████▋ | 524/602 [01:25<00:12,  6.10it/s]
Validation DataLoader 0:  87%|████████▋ | 525/602 [01:25<00:12,  6.11it/s]
Validation DataLoader 0:  87%|████████▋ | 526/602 [01:26<00:12,  6.12it/s]
Validation DataLoader 0:  88%|████████▊ | 527/602 [01:26<00:12,  6.10it/s]
Validation DataLoader 0:  88%|████████▊ | 528/602 [01:26<00:12,  6.11it/s]
Validation DataLoader 0:  88%|████████▊ | 529/602 [01:26<00:11,  6.12it/s]
Validation DataLoader 0:  88%|████████▊ | 530/602 [01:26<00:11,  6.10it/s]
Validation DataLoader 0:  88%|████████▊ | 531/602 [01:26<00:11,  6.11it/s]
Validation DataLoader 0:  88%|████████▊ | 532/602 [01:26<00:11,  6.12it/s]
Validation DataLoader 0:  89%|████████▊ | 533/602 [01:27<00:11,  6.10it/s]
Validation DataLoader 0:  89%|████████▊ | 534/602 [01:27<00:11,  6.11it/s]
Validation DataLoader 0:  89%|████████▉ | 535/602 [01:27<00:10,  6.12it/s]
Validation DataLoader 0:  89%|████████▉ | 536/602 [01:27<00:10,  6.11it/s]
Validation DataLoader 0:  89%|████████▉ | 537/602 [01:27<00:10,  6.12it/s]
Validation DataLoader 0:  89%|████████▉ | 538/602 [01:27<00:10,  6.13it/s]
Validation DataLoader 0:  90%|████████▉ | 539/602 [01:28<00:10,  6.11it/s]
Validation DataLoader 0:  90%|████████▉ | 540/602 [01:28<00:10,  6.12it/s]
Validation DataLoader 0:  90%|████████▉ | 541/602 [01:28<00:09,  6.13it/s]
Validation DataLoader 0:  90%|█████████ | 542/602 [01:28<00:09,  6.11it/s]
Validation DataLoader 0:  90%|█████████ | 543/602 [01:28<00:09,  6.12it/s]
Validation DataLoader 0:  90%|█████████ | 544/602 [01:28<00:09,  6.13it/s]
Validation DataLoader 0:  91%|█████████ | 545/602 [01:29<00:09,  6.11it/s]
Validation DataLoader 0:  91%|█████████ | 546/602 [01:29<00:09,  6.12it/s]
Validation DataLoader 0:  91%|█████████ | 547/602 [01:29<00:08,  6.13it/s]
Validation DataLoader 0:  91%|█████████ | 548/602 [01:29<00:08,  6.11it/s]
Validation DataLoader 0:  91%|█████████ | 549/602 [01:29<00:08,  6.12it/s]
Validation DataLoader 0:  91%|█████████▏| 550/602 [01:29<00:08,  6.14it/s]
Validation DataLoader 0:  92%|█████████▏| 551/602 [01:30<00:08,  6.12it/s]
Validation DataLoader 0:  92%|█████████▏| 552/602 [01:30<00:08,  6.13it/s]
Validation DataLoader 0:  92%|█████████▏| 553/602 [01:30<00:07,  6.14it/s]
Validation DataLoader 0:  92%|█████████▏| 554/602 [01:30<00:07,  6.12it/s]
Validation DataLoader 0:  92%|█████████▏| 555/602 [01:30<00:07,  6.13it/s]
Validation DataLoader 0:  92%|█████████▏| 556/602 [01:30<00:07,  6.14it/s]
Validation DataLoader 0:  93%|█████████▎| 557/602 [01:31<00:07,  6.12it/s]
Validation DataLoader 0:  93%|█████████▎| 558/602 [01:31<00:07,  6.13it/s]
Validation DataLoader 0:  93%|█████████▎| 559/602 [01:31<00:07,  6.14it/s]
Validation DataLoader 0:  93%|█████████▎| 560/602 [01:31<00:06,  6.12it/s]
Validation DataLoader 0:  93%|█████████▎| 561/602 [01:31<00:06,  6.13it/s]
Validation DataLoader 0:  93%|█████████▎| 562/602 [01:31<00:06,  6.14it/s]
Validation DataLoader 0:  94%|█████████▎| 563/602 [01:31<00:06,  6.13it/s]
Validation DataLoader 0:  94%|█████████▎| 564/602 [01:31<00:06,  6.14it/s]
Validation DataLoader 0:  94%|█████████▍| 565/602 [01:31<00:06,  6.15it/s]
Validation DataLoader 0:  94%|█████████▍| 566/602 [01:32<00:05,  6.12it/s]
Validation DataLoader 0:  94%|█████████▍| 567/602 [01:32<00:05,  6.13it/s]
Validation DataLoader 0:  94%|█████████▍| 568/602 [01:32<00:05,  6.14it/s]
Validation DataLoader 0:  95%|█████████▍| 569/602 [01:32<00:05,  6.13it/s]
Validation DataLoader 0:  95%|█████████▍| 570/602 [01:32<00:05,  6.14it/s]
Validation DataLoader 0:  95%|█████████▍| 571/602 [01:32<00:05,  6.15it/s]
Validation DataLoader 0:  95%|█████████▌| 572/602 [01:33<00:04,  6.13it/s]
Validation DataLoader 0:  95%|█████████▌| 573/602 [01:33<00:04,  6.14it/s]
Validation DataLoader 0:  95%|█████████▌| 574/602 [01:33<00:04,  6.15it/s]
Validation DataLoader 0:  96%|█████████▌| 575/602 [01:33<00:04,  6.14it/s]
Validation DataLoader 0:  96%|█████████▌| 576/602 [01:33<00:04,  6.14it/s]
Validation DataLoader 0:  96%|█████████▌| 577/602 [01:33<00:04,  6.15it/s]
Validation DataLoader 0:  96%|█████████▌| 578/602 [01:34<00:03,  6.14it/s]
Validation DataLoader 0:  96%|█████████▌| 579/602 [01:34<00:03,  6.15it/s]
Validation DataLoader 0:  96%|█████████▋| 580/602 [01:34<00:03,  6.16it/s]
Validation DataLoader 0:  97%|█████████▋| 581/602 [01:34<00:03,  6.14it/s]
Validation DataLoader 0:  97%|█████████▋| 582/602 [01:34<00:03,  6.15it/s]
Validation DataLoader 0:  97%|█████████▋| 583/602 [01:34<00:03,  6.16it/s]
Validation DataLoader 0:  97%|█████████▋| 584/602 [01:35<00:02,  6.14it/s]
Validation DataLoader 0:  97%|█████████▋| 585/602 [01:35<00:02,  6.15it/s]
Validation DataLoader 0:  97%|█████████▋| 586/602 [01:35<00:02,  6.16it/s]
Validation DataLoader 0:  98%|█████████▊| 587/602 [01:35<00:02,  6.14it/s]
Validation DataLoader 0:  98%|█████████▊| 588/602 [01:35<00:02,  6.15it/s]
Validation DataLoader 0:  98%|█████████▊| 589/602 [01:35<00:02,  6.16it/s]
Validation DataLoader 0:  98%|█████████▊| 590/602 [01:36<00:01,  6.14it/s]
Validation DataLoader 0:  98%|█████████▊| 591/602 [01:36<00:01,  6.15it/s]
Validation DataLoader 0:  98%|█████████▊| 592/602 [01:36<00:01,  6.16it/s]
Validation DataLoader 0:  99%|█████████▊| 593/602 [01:36<00:01,  6.14it/s]
Validation DataLoader 0:  99%|█████████▊| 594/602 [01:36<00:01,  6.15it/s]
Validation DataLoader 0:  99%|█████████▉| 595/602 [01:36<00:01,  6.16it/s]
Validation DataLoader 0:  99%|█████████▉| 596/602 [01:36<00:00,  6.15it/s]
Validation DataLoader 0:  99%|█████████▉| 597/602 [01:36<00:00,  6.16it/s]
Validation DataLoader 0:  99%|█████████▉| 598/602 [01:36<00:00,  6.17it/s]
Validation DataLoader 0: 100%|█████████▉| 599/602 [01:37<00:00,  6.15it/s]
Validation DataLoader 0: 100%|█████████▉| 600/602 [01:37<00:00,  6.16it/s]
Validation DataLoader 0: 100%|█████████▉| 601/602 [01:37<00:00,  6.17it/s]
Validation DataLoader 0: 100%|██████████| 602/602 [01:37<00:00,  6.16it/s]
Validation DataLoader 0: 100%|██████████| 602/602 [01:37<00:00,  6.16it/s]Restoring states from the checkpoint path at /Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=03-R@1 (val)=0.407.ckpt
Loaded model weights from the checkpoint at /Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/checkpoints/paper_competitors/embcos_formula_seed1/best-epoch=03-R@1 (val)=0.407.ckpt
/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/.venv/lib/python3.11/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:485: Your `test_dataloader`'s sampler has shuffling enabled, it is strongly recommended that you turn shuffling off for val/test dataloaders.
/Users/timrichardson/Documents/projects/personal/blog/rewire.it/workbench/outputs/blog-post-msms-model-selection-20261003-230223/runs/external/msalign-src/.venv/lib/python3.11/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:429: Consider setting `persistent_workers=True` in 'test_dataloader' to speed up the dataloader worker initialization.

────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
     Validate metric           DataLoader 0
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
        R@1 (val)           0.4074343144893646
       R@20 (val)           0.7929602265357971
        R@5 (val)            0.609801709651947
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────

Testing: |          | 0/? [00:00<?, ?it/s]
Testing: |          | 0/? [00:00<?, ?it/s]
Testing DataLoader 0:   0%|          | 0/666 [00:00<?, ?it/s]
Testing DataLoader 0:   0%|          | 1/666 [00:00<00:17, 37.07it/s]
Testing DataLoader 0:   0%|          | 2/666 [00:00<00:16, 40.17it/s]
Testing DataLoader 0:   0%|          | 3/666 [00:00<00:15, 44.14it/s]
Testing DataLoader 0:   1%|          | 4/666 [00:00<01:07,  9.74it/s]
Testing DataLoader 0:   1%|          | 5/666 [00:00<00:56, 11.78it/s]
Testing DataLoader 0:   1%|          | 6/666 [00:00<00:58, 11.27it/s]
Testing DataLoader 0:   1%|          | 7/666 [00:00<01:26,  7.58it/s]
Testing DataLoader 0:   1%|          | 8/666 [00:00<01:17,  8.53it/s]
Testing DataLoader 0:   1%|▏         | 9/666 [00:00<01:12,  9.03it/s]
Testing DataLoader 0:   2%|▏         | 10/666 [00:01<01:39,  6.61it/s]
Testing DataLoader 0:   2%|▏         | 11/666 [00:01<01:30,  7.21it/s]
Testing DataLoader 0:   2%|▏         | 12/666 [00:01<01:23,  7.81it/s]
Testing DataLoader 0:   2%|▏         | 13/666 [00:01<01:39,  6.58it/s]
Testing DataLoader 0:   2%|▏         | 14/666 [00:01<01:32,  7.04it/s]
Testing DataLoader 0:   2%|▏         | 15/666 [00:02<01:26,  7.49it/s]
Testing DataLoader 0:   2%|▏         | 16/666 [00:02<01:42,  6.37it/s]
Testing DataLoader 0:   3%|▎         | 17/666 [00:02<01:36,  6.73it/s]
Testing DataLoader 0:   3%|▎         | 18/666 [00:02<01:32,  6.98it/s]
Testing DataLoader 0:   3%|▎         | 19/666 [00:03<01:42,  6.33it/s]
Testing DataLoader 0:   3%|▎         | 20/666 [00:03<01:37,  6.64it/s]
Testing DataLoader 0:   3%|▎         | 21/666 [00:03<01:34,  6.81it/s]
Testing DataLoader 0:   3%|▎         | 22/666 [00:03<01:41,  6.37it/s]
Testing DataLoader 0:   3%|▎         | 23/666 [00:03<01:37,  6.62it/s]
Testing DataLoader 0:   4%|▎         | 24/666 [00:03<01:33,  6.87it/s]
Testing DataLoader 0:   4%|▍         | 25/666 [00:03<01:42,  6.26it/s]
Testing DataLoader 0:   4%|▍         | 26/666 [00:04<01:38,  6.49it/s]
Testing DataLoader 0:   4%|▍         | 27/666 [00:04<01:37,  6.57it/s]
Testing DataLoader 0:   4%|▍         | 28/666 [00:04<01:42,  6.21it/s]
Testing DataLoader 0:   4%|▍         | 29/666 [00:04<01:39,  6.41it/s]
Testing DataLoader 0:   5%|▍         | 30/666 [00:04<01:37,  6.50it/s]
Testing DataLoader 0:   5%|▍         | 31/666 [00:05<01:43,  6.13it/s]
Testing DataLoader 0:   5%|▍         | 32/666 [00:05<01:40,  6.31it/s]
Testing DataLoader 0:   5%|▍         | 33/666 [00:05<01:37,  6.50it/s]
Testing DataLoader 0:   5%|▌         | 34/666 [00:05<01:42,  6.17it/s]
Testing DataLoader 0:   5%|▌         | 35/666 [00:05<01:39,  6.33it/s]
Testing DataLoader 0:   5%|▌         | 36/666 [00:05<01:38,  6.41it/s]
Testing DataLoader 0:   6%|▌         | 37/666 [00:06<01:42,  6.15it/s]
Testing DataLoader 0:   6%|▌         | 38/666 [00:06<01:39,  6.30it/s]
Testing DataLoader 0:   6%|▌         | 39/666 [00:06<01:37,  6.46it/s]
Testing DataLoader 0:   6%|▌         | 40/666 [00:06<01:41,  6.15it/s]
Testing DataLoader 0:   6%|▌         | 41/666 [00:06<01:39,  6.29it/s]
Testing DataLoader 0:   6%|▋         | 42/666 [00:06<01:37,  6.43it/s]
Testing DataLoader 0:   6%|▋         | 43/666 [00:06<01:40,  6.18it/s]
Testing DataLoader 0:   7%|▋         | 44/666 [00:06<01:38,  6.31it/s]
Testing DataLoader 0:   7%|▋         | 45/666 [00:07<01:39,  6.27it/s]
Testing DataLoader 0:   7%|▋         | 46/666 [00:07<01:47,  5.77it/s]
Testing DataLoader 0:   7%|▋         | 47/666 [00:07<01:45,  5.88it/s]
Testing DataLoader 0:   7%|▋         | 48/666 [00:08<01:43,  5.99it/s]
Testing DataLoader 0:   7%|▋         | 49/666 [00:08<01:49,  5.64it/s]
Testing DataLoader 0:   8%|▊         | 50/666 [00:08<01:47,  5.74it/s]
Testing DataLoader 0:   8%|▊         | 51/666 [00:08<01:45,  5.85it/s]
Testing DataLoader 0:   8%|▊         | 52/666 [00:09<01:48,  5.69it/s]
Testing DataLoader 0:   8%|▊         | 53/666 [00:09<01:45,  5.79it/s]
Testing DataLoader 0:   8%|▊         | 54/666 [00:09<01:43,  5.89it/s]
Testing DataLoader 0:   8%|▊         | 55/666 [00:09<01:47,  5.71it/s]
Testing DataLoader 0:   8%|▊         | 56/666 [00:09<01:45,  5.80it/s]
Testing DataLoader 0:   9%|▊         | 57/666 [00:09<01:43,  5.90it/s]
Testing DataLoader 0:   9%|▊         | 58/666 [00:10<01:46,  5.73it/s]
Testing DataLoader 0:   9%|▉         | 59/666 [00:10<01:44,  5.82it/s]
Testing DataLoader 0:   9%|▉         | 60/666 [00:10<01:42,  5.91it/s]
Testing DataLoader 0:   9%|▉         | 61/666 [00:10<01:46,  5.69it/s]
Testing DataLoader 0:   9%|▉         | 62/666 [00:10<01:44,  5.77it/s]
Testing DataLoader 0:   9%|▉         | 63/666 [00:10<01:42,  5.86it/s]
Testing DataLoader 0:  10%|▉         | 64/666 [00:11<01:45,  5.69it/s]
Testing DataLoader 0:  10%|▉         | 65/666 [00:11<01:44,  5.77it/s]
Testing DataLoader 0:  10%|▉         | 66/666 [00:11<01:42,  5.85it/s]
Testing DataLoader 0:  10%|█         | 67/666 [00:11<01:44,  5.71it/s]
Testing DataLoader 0:  10%|█         | 68/666 [00:11<01:43,  5.79it/s]
Testing DataLoader 0:  10%|█         | 69/666 [00:11<01:41,  5.87it/s]
Testing DataLoader 0:  11%|█         | 70/666 [00:12<01:44,  5.72it/s]
Testing DataLoader 0:  11%|█         | 71/666 [00:12<01:42,  5.79it/s]
Testing DataLoader 0:  11%|█         | 72/666 [00:12<01:41,  5.87it/s]
Testing DataLoader 0:  11%|█         | 73/666 [00:12<01:42,  5.76it/s]
Testing DataLoader 0:  11%|█         | 74/666 [00:12<01:41,  5.83it/s]
Testing DataLoader 0:  11%|█▏        | 75/666 [00:12<01:40,  5.90it/s]
Testing DataLoader 0:  11%|█▏        | 76/666 [00:13<01:41,  5.80it/s]
Testing DataLoader 0:  12%|█▏        | 77/666 [00:13<01:40,  5.87it/s]
Testing DataLoader 0:  12%|█▏        | 78/666 [00:13<01:38,  5.94it/s]
Testing DataLoader 0:  12%|█▏        | 79/666 [00:13<01:41,  5.79it/s]
Testing DataLoader 0:  12%|█▏        | 80/666 [00:13<01:40,  5.86it/s]
Testing DataLoader 0:  12%|█▏        | 81/666 [00:13<01:38,  5.93it/s]
Testing DataLoader 0:  12%|█▏        | 82/666 [00:14<01:40,  5.78it/s]
Testing DataLoader 0:  12%|█▏        | 83/666 [00:14<01:39,  5.85it/s]
Testing DataLoader 0:  13%|█▎        | 84/666 [00:14<01:38,  5.91it/s]
Testing DataLoader 0:  13%|█▎        | 85/666 [00:14<01:40,  5.77it/s]
Testing DataLoader 0:  13%|█▎        | 86/666 [00:14<01:39,  5.83it/s]
Testing DataLoader 0:  13%|█▎        | 87/666 [00:14<01:38,  5.90it/s]
Testing DataLoader 0:  13%|█▎        | 88/666 [00:15<01:40,  5.76it/s]
Testing DataLoader 0:  13%|█▎        | 89/666 [00:15<01:39,  5.82it/s]
Testing DataLoader 0:  14%|█▎        | 90/666 [00:15<01:37,  5.88it/s]
Testing DataLoader 0:  14%|█▎        | 91/666 [00:15<01:40,  5.74it/s]
Testing DataLoader 0:  14%|█▍        | 92/666 [00:15<01:38,  5.80it/s]
Testing DataLoader 0:  14%|█▍        | 93/666 [00:15<01:37,  5.86it/s]
Testing DataLoader 0:  14%|█▍        | 94/666 [00:16<01:39,  5.73it/s]
Testing DataLoader 0:  14%|█▍        | 95/666 [00:16<01:38,  5.79it/s]
Testing DataLoader 0:  14%|█▍        | 96/666 [00:16<01:37,  5.84it/s]
Testing DataLoader 0:  15%|█▍        | 97/666 [00:16<01:39,  5.72it/s]
Testing DataLoader 0:  15%|█▍        | 98/666 [00:16<01:38,  5.77it/s]
Testing DataLoader 0:  15%|█▍        | 99/666 [00:16<01:37,  5.82it/s]
Testing DataLoader 0:  15%|█▌        | 100/666 [00:17<01:38,  5.75it/s]
Testing DataLoader 0:  15%|█▌        | 101/666 [00:17<01:37,  5.80it/s]
Testing DataLoader 0:  15%|█▌        | 102/666 [00:17<01:36,  5.85it/s]
Testing DataLoader 0:  15%|█▌        | 103/666 [00:17<01:38,  5.74it/s]
Testing DataLoader 0:  16%|█▌        | 104/666 [00:17<01:37,  5.79it/s]
Testing DataLoader 0:  16%|█▌        | 105/666 [00:17<01:36,  5.84it/s]
Testing DataLoader 0:  16%|█▌        | 106/666 [00:18<01:37,  5.72it/s]
Testing DataLoader 0:  16%|█▌        | 107/666 [00:18<01:36,  5.77it/s]
Testing DataLoader 0:  16%|█▌        | 108/666 [00:18<01:35,  5.82it/s]
Testing DataLoader 0:  16%|█▋        | 109/666 [00:18<01:36,  5.76it/s]
Testing DataLoader 0:  17%|█▋        | 110/666 [00:18<01:35,  5.81it/s]
Testing DataLoader 0:  17%|█▋        | 111/666 [00:18<01:34,  5.86it/s]
Testing DataLoader 0:  17%|█▋        | 112/666 [00:19<01:35,  5.77it/s]
Testing DataLoader 0:  17%|█▋        | 113/666 [00:19<01:34,  5.82it/s]
Testing DataLoader 0:  17%|█▋        | 114/666 [00:19<01:34,  5.87it/s]
Testing DataLoader 0:  17%|█▋        | 115/666 [00:19<01:35,  5.78it/s]
Testing DataLoader 0:  17%|█▋        | 116/666 [00:19<01:34,  5.83it/s]
Testing DataLoader 0:  18%|█▊        | 117/666 [00:19<01:33,  5.88it/s]
Testing DataLoader 0:  18%|█▊        | 118/666 [00:20<01:34,  5.78it/s]
Testing DataLoader 0:  18%|█▊        | 119/666 [00:20<01:33,  5.82it/s]
Testing DataLoader 0:  18%|█▊        | 120/666 [00:20<01:33,  5.87it/s]
Testing DataLoader 0:  18%|█▊        | 121/666 [00:20<01:34,  5.77it/s]
Testing DataLoader 0:  18%|█▊        | 122/666 [00:20<01:33,  5.82it/s]
Testing DataLoader 0:  18%|█▊        | 123/666 [00:20<01:32,  5.86it/s]
Testing DataLoader 0:  19%|█▊        | 124/666 [00:21<01:33,  5.77it/s]
Testing DataLoader 0:  19%|█▉        | 125/666 [00:21<01:33,  5.81it/s]
Testing DataLoader 0:  19%|█▉        | 126/666 [00:21<01:32,  5.86it/s]
Testing DataLoader 0:  19%|█▉        | 127/666 [00:21<01:33,  5.78it/s]
Testing DataLoader 0:  19%|█▉        | 128/666 [00:21<01:32,  5.82it/s]
Testing DataLoader 0:  19%|█▉        | 129/666 [00:22<01:31,  5.86it/s]
Testing DataLoader 0:  20%|█▉        | 130/666 [00:22<01:32,  5.78it/s]
Testing DataLoader 0:  20%|█▉        | 131/666 [00:22<01:31,  5.83it/s]
Testing DataLoader 0:  20%|█▉        | 132/666 [00:22<01:31,  5.87it/s]
Testing DataLoader 0:  20%|█▉        | 133/666 [00:22<01:31,  5.81it/s]
Testing DataLoader 0:  20%|██        | 134/666 [00:22<01:30,  5.85it/s]
Testing DataLoader 0:  20%|██        | 135/666 [00:22<01:30,  5.89it/s]
Testing DataLoader 0:  20%|██        | 136/666 [00:23<01:31,  5.82it/s]
Testing DataLoader 0:  21%|██        | 137/666 [00:23<01:30,  5.86it/s]
Testing DataLoader 0:  21%|██        | 138/666 [00:23<01:29,  5.90it/s]
Testing DataLoader 0:  21%|██        | 139/666 [00:23<01:30,  5.83it/s]
Testing DataLoader 0:  21%|██        | 140/666 [00:23<01:29,  5.87it/s]
Testing DataLoader 0:  21%|██        | 141/666 [00:23<01:28,  5.91it/s]
Testing DataLoader 0:  21%|██▏       | 142/666 [00:24<01:29,  5.83it/s]
Testing DataLoader 0:  21%|██▏       | 143/666 [00:24<01:29,  5.87it/s]
Testing DataLoader 0:  22%|██▏       | 144/666 [00:24<01:28,  5.91it/s]
Testing DataLoader 0:  22%|██▏       | 145/666 [00:24<01:29,  5.83it/s]
Testing DataLoader 0:  22%|██▏       | 146/666 [00:24<01:28,  5.87it/s]
Testing DataLoader 0:  22%|██▏       | 147/666 [00:24<01:27,  5.90it/s]
Testing DataLoader 0:  22%|██▏       | 148/666 [00:25<01:28,  5.82it/s]
Testing DataLoader 0:  22%|██▏       | 149/666 [00:25<01:28,  5.86it/s]
Testing DataLoader 0:  23%|██▎       | 150/666 [00:25<01:27,  5.90it/s]
Testing DataLoader 0:  23%|██▎       | 151/666 [00:25<01:28,  5.82it/s]
Testing DataLoader 0:  23%|██▎       | 152/666 [00:25<01:27,  5.86it/s]
Testing DataLoader 0:  23%|██▎       | 153/666 [00:25<01:27,  5.90it/s]
Testing DataLoader 0:  23%|██▎       | 154/666 [00:26<01:28,  5.82it/s]
Testing DataLoader 0:  23%|██▎       | 155/666 [00:26<01:27,  5.85it/s]
Testing DataLoader 0:  23%|██▎       | 156/666 [00:26<01:26,  5.89it/s]
Testing DataLoader 0:  24%|██▎       | 157/666 [00:27<01:27,  5.80it/s]
Testing DataLoader 0:  24%|██▎       | 158/666 [00:27<01:27,  5.84it/s]
Testing DataLoader 0:  24%|██▍       | 159/666 [00:27<01:26,  5.87it/s]
Testing DataLoader 0:  24%|██▍       | 160/666 [00:27<01:26,  5.82it/s]
Testing DataLoader 0:  24%|██▍       | 161/666 [00:27<01:26,  5.85it/s]
Testing DataLoader 0:  24%|██▍       | 162/666 [00:27<01:25,  5.89it/s]
Testing DataLoader 0:  24%|██▍       | 163/666 [00:27<01:26,  5.82it/s]
Testing DataLoader 0:  25%|██▍       | 164/666 [00:27<01:25,  5.86it/s]
Testing DataLoader 0:  25%|██▍       | 165/666 [00:28<01:25,  5.89it/s]
Testing DataLoader 0:  25%|██▍       | 166/666 [00:28<01:25,  5.84it/s]
Testing DataLoader 0:  25%|██▌       | 167/666 [00:28<01:24,  5.87it/s]
Testing DataLoader 0:  25%|██▌       | 168/666 [00:28<01:24,  5.91it/s]
Testing DataLoader 0:  25%|██▌       | 169/666 [00:28<01:24,  5.85it/s]
Testing DataLoader 0:  26%|██▌       | 170/666 [00:28<01:24,  5.88it/s]
Testing DataLoader 0:  26%|██▌       | 171/666 [00:28<01:23,  5.91it/s]
Testing DataLoader 0:  26%|██▌       | 172/666 [00:29<01:24,  5.85it/s]
Testing DataLoader 0:  26%|██▌       | 173/666 [00:29<01:23,  5.88it/s]
Testing DataLoader 0:  26%|██▌       | 174/666 [00:29<01:23,  5.92it/s]
Testing DataLoader 0:  26%|██▋       | 175/666 [00:29<01:23,  5.87it/s]
Testing DataLoader 0:  26%|██▋       | 176/666 [00:29<01:23,  5.90it/s]
Testing DataLoader 0:  27%|██▋       | 177/666 [00:29<01:22,  5.93it/s]
Testing DataLoader 0:  27%|██▋       | 178/666 [00:30<01:23,  5.86it/s]
Testing DataLoader 0:  27%|██▋       | 179/666 [00:30<01:22,  5.89it/s]
Testing DataLoader 0:  27%|██▋       | 180/666 [00:30<01:22,  5.92it/s]
Testing DataLoader 0:  27%|██▋       | 181/666 [00:30<01:23,  5.84it/s]
Testing DataLoader 0:  27%|██▋       | 182/666 [00:30<01:22,  5.87it/s]
Testing DataLoader 0:  27%|██▋       | 183/666 [00:31<01:21,  5.90it/s]
Testing DataLoader 0:  28%|██▊       | 184/666 [00:31<01:22,  5.84it/s]
Testing DataLoader 0:  28%|██▊       | 185/666 [00:31<01:21,  5.87it/s]
Testing DataLoader 0:  28%|██▊       | 186/666 [00:31<01:21,  5.90it/s]
Testing DataLoader 0:  28%|██▊       | 187/666 [00:32<01:21,  5.84it/s]
Testing DataLoader 0:  28%|██▊       | 188/666 [00:32<01:21,  5.87it/s]
Testing DataLoader 0:  28%|██▊       | 189/666 [00:32<01:20,  5.90it/s]
Testing DataLoader 0:  29%|██▊       | 190/666 [00:32<01:21,  5.84it/s]
Testing DataLoader 0:  29%|██▊       | 191/666 [00:32<01:20,  5.87it/s]
Testing DataLoader 0:  29%|██▉       | 192/666 [00:32<01:20,  5.89it/s]
Testing DataLoader 0:  29%|██▉       | 193/666 [00:33<01:21,  5.83it/s]
Testing DataLoader 0:  29%|██▉       | 194/666 [00:33<01:20,  5.86it/s]
Testing DataLoader 0:  29%|██▉       | 195/666 [00:33<01:19,  5.89it/s]
Testing DataLoader 0:  29%|██▉       | 196/666 [00:33<01:20,  5.84it/s]
Testing DataLoader 0:  30%|██▉       | 197/666 [00:33<01:19,  5.87it/s]
Testing DataLoader 0:  30%|██▉       | 198/666 [00:33<01:19,  5.90it/s]
Testing DataLoader 0:  30%|██▉       | 199/666 [00:34<01:19,  5.84it/s]
Testing DataLoader 0:  30%|███       | 200/666 [00:34<01:19,  5.87it/s]
Testing DataLoader 0:  30%|███       | 201/666 [00:34<01:18,  5.89it/s]
Testing DataLoader 0:  30%|███       | 202/666 [00:34<01:19,  5.84it/s]
Testing DataLoader 0:  30%|███       | 203/666 [00:34<01:18,  5.86it/s]
Testing DataLoader 0:  31%|███       | 204/666 [00:34<01:18,  5.89it/s]
Testing DataLoader 0:  31%|███       | 205/666 [00:35<01:18,  5.84it/s]
Testing DataLoader 0:  31%|███       | 206/666 [00:35<01:18,  5.87it/s]
Testing DataLoader 0:  31%|███       | 207/666 [00:35<01:17,  5.90it/s]
Testing DataLoader 0:  31%|███       | 208/666 [00:35<01:18,  5.85it/s]
Testing DataLoader 0:  31%|███▏      | 209/666 [00:35<01:17,  5.88it/s]
Testing DataLoader 0:  32%|███▏      | 210/666 [00:35<01:17,  5.90it/s]
Testing DataLoader 0:  32%|███▏      | 211/666 [00:35<01:17,  5.87it/s]
Testing DataLoader 0:  32%|███▏      | 212/666 [00:35<01:17,  5.89it/s]
Testing DataLoader 0:  32%|███▏      | 213/666 [00:35<01:16,  5.92it/s]
Testing DataLoader 0:  32%|███▏      | 214/666 [00:36<01:17,  5.87it/s]
Testing DataLoader 0:  32%|███▏      | 215/666 [00:36<01:16,  5.89it/s]
Testing DataLoader 0:  32%|███▏      | 216/666 [00:36<01:16,  5.92it/s]
Testing DataLoader 0:  33%|███▎      | 217/666 [00:37<01:16,  5.86it/s]
Testing DataLoader 0:  33%|███▎      | 218/666 [00:37<01:16,  5.89it/s]
Testing DataLoader 0:  33%|███▎      | 219/666 [00:37<01:15,  5.91it/s]
Testing DataLoader 0:  33%|███▎      | 220/666 [00:37<01:16,  5.87it/s]
Testing DataLoader 0:  33%|███▎      | 221/666 [00:37<01:15,  5.89it/s]
Testing DataLoader 0:  33%|███▎      | 222/666 [00:37<01:15,  5.91it/s]
Testing DataLoader 0:  33%|███▎      | 223/666 [00:38<01:15,  5.87it/s]
Testing DataLoader 0:  34%|███▎      | 224/666 [00:38<01:15,  5.89it/s]
Testing DataLoader 0:  34%|███▍      | 225/666 [00:38<01:14,  5.91it/s]
Testing DataLoader 0:  34%|███▍      | 226/666 [00:38<01:15,  5.85it/s]
Testing DataLoader 0:  34%|███▍      | 227/666 [00:38<01:14,  5.88it/s]
Testing DataLoader 0:  34%|███▍      | 228/666 [00:38<01:14,  5.90it/s]
Testing DataLoader 0:  34%|███▍      | 229/666 [00:39<01:14,  5.86it/s]
Testing DataLoader 0:  35%|███▍      | 230/666 [00:39<01:14,  5.88it/s]
Testing DataLoader 0:  35%|███▍      | 231/666 [00:39<01:13,  5.90it/s]
Testing DataLoader 0:  35%|███▍      | 232/666 [00:39<01:14,  5.86it/s]
Testing DataLoader 0:  35%|███▍      | 233/666 [00:39<01:13,  5.88it/s]
Testing DataLoader 0:  35%|███▌      | 234/666 [00:39<01:13,  5.90it/s]
Testing DataLoader 0:  35%|███▌      | 235/666 [00:40<01:13,  5.87it/s]
Testing DataLoader 0:  35%|███▌      | 236/666 [00:40<01:13,  5.89it/s]
Testing DataLoader 0:  36%|███▌      | 237/666 [00:40<01:12,  5.91it/s]
Testing DataLoader 0:  36%|███▌      | 238/666 [00:40<01:12,  5.86it/s]
Testing DataLoader 0:  36%|███▌      | 239/666 [00:40<01:12,  5.89it/s]
Testing DataLoader 0:  36%|███▌      | 240/666 [00:40<01:12,  5.91it/s]
Testing DataLoader 0:  36%|███▌      | 241/666 [00:41<01:12,  5.87it/s]
Testing DataLoader 0:  36%|███▋      | 242/666 [00:41<01:11,  5.89it/s]
Testing DataLoader 0:  36%|███▋      | 243/666 [00:41<01:11,  5.92it/s]
Testing DataLoader 0:  37%|███▋      | 244/666 [00:41<01:11,  5.87it/s]
Testing DataLoader 0:  37%|███▋      | 245/666 [00:41<01:11,  5.90it/s]
Testing DataLoader 0:  37%|███▋      | 246/666 [00:41<01:10,  5.92it/s]
Testing DataLoader 0:  37%|███▋      | 247/666 [00:41<01:11,  5.89it/s]
Testing DataLoader 0:  37%|███▋      | 248/666 [00:41<01:10,  5.91it/s]
Testing DataLoader 0:  37%|███▋      | 249/666 [00:41<01:10,  5.93it/s]
Testing DataLoader 0:  38%|███▊      | 250/666 [00:42<01:10,  5.89it/s]
Testing DataLoader 0:  38%|███▊      | 251/666 [00:42<01:10,  5.91it/s]
Testing DataLoader 0:  38%|███▊      | 252/666 [00:42<01:09,  5.93it/s]
Testing DataLoader 0:  38%|███▊      | 253/666 [00:42<01:10,  5.89it/s]
Testing DataLoader 0:  38%|███▊      | 254/666 [00:42<01:09,  5.91it/s]
Testing DataLoader 0:  38%|███▊      | 255/666 [00:42<01:09,  5.93it/s]
Testing DataLoader 0:  38%|███▊      | 256/666 [00:43<01:09,  5.88it/s]
Testing DataLoader 0:  39%|███▊      | 257/666 [00:43<01:09,  5.90it/s]
Testing DataLoader 0:  39%|███▊      | 258/666 [00:43<01:08,  5.92it/s]
Testing DataLoader 0:  39%|███▉      | 259/666 [00:44<01:09,  5.88it/s]
Testing DataLoader 0:  39%|███▉      | 260/666 [00:44<01:08,  5.90it/s]
Testing DataLoader 0:  39%|███▉      | 261/666 [00:44<01:08,  5.92it/s]
Testing DataLoader 0:  39%|███▉      | 262/666 [00:44<01:08,  5.88it/s]
Testing DataLoader 0:  39%|███▉      | 263/666 [00:44<01:08,  5.90it/s]
Testing DataLoader 0:  40%|███▉      | 264/666 [00:44<01:07,  5.92it/s]
Testing DataLoader 0:  40%|███▉      | 265/666 [00:45<01:08,  5.88it/s]
Testing DataLoader 0:  40%|███▉      | 266/666 [00:45<01:07,  5.90it/s]
Testing DataLoader 0:  40%|████      | 267/666 [00:45<01:07,  5.92it/s]
Testing DataLoader 0:  40%|████      | 268/666 [00:45<01:07,  5.88it/s]
Testing DataLoader 0:  40%|████      | 269/666 [00:45<01:07,  5.90it/s]
Testing DataLoader 0:  41%|████      | 270/666 [00:45<01:06,  5.92it/s]
Testing DataLoader 0:  41%|████      | 271/666 [00:45<01:06,  5.90it/s]
Testing DataLoader 0:  41%|████      | 272/666 [00:45<01:06,  5.92it/s]
Testing DataLoader 0:  41%|████      | 273/666 [00:45<01:06,  5.94it/s]
Testing DataLoader 0:  41%|████      | 274/666 [00:46<01:06,  5.90it/s]
Testing DataLoader 0:  41%|████▏     | 275/666 [00:46<01:06,  5.92it/s]
Testing DataLoader 0:  41%|████▏     | 276/666 [00:46<01:05,  5.94it/s]
Testing DataLoader 0:  42%|████▏     | 277/666 [00:46<01:05,  5.90it/s]
Testing DataLoader 0:  42%|████▏     | 278/666 [00:46<01:05,  5.92it/s]
Testing DataLoader 0:  42%|████▏     | 279/666 [00:46<01:05,  5.94it/s]
Testing DataLoader 0:  42%|████▏     | 280/666 [00:47<01:05,  5.89it/s]
Testing DataLoader 0:  42%|████▏     | 281/666 [00:47<01:05,  5.91it/s]
Testing DataLoader 0:  42%|████▏     | 282/666 [00:47<01:04,  5.93it/s]
Testing DataLoader 0:  42%|████▏     | 283/666 [00:48<01:05,  5.89it/s]
Testing DataLoader 0:  43%|████▎     | 284/666 [00:48<01:04,  5.91it/s]
Testing DataLoader 0:  43%|████▎     | 285/666 [00:48<01:04,  5.92it/s]
Testing DataLoader 0:  43%|████▎     | 286/666 [00:48<01:04,  5.87it/s]
Testing DataLoader 0:  43%|████▎     | 287/666 [00:48<01:04,  5.89it/s]
Testing DataLoader 0:  43%|████▎     | 288/666 [00:48<01:03,  5.91it/s]
Testing DataLoader 0:  43%|████▎     | 289/666 [00:49<01:04,  5.86it/s]
Testing DataLoader 0:  44%|████▎     | 290/666 [00:49<01:03,  5.88it/s]
Testing DataLoader 0:  44%|████▎     | 291/666 [00:49<01:03,  5.90it/s]
Testing DataLoader 0:  44%|████▍     | 292/666 [00:49<01:03,  5.87it/s]
Testing DataLoader 0:  44%|████▍     | 293/666 [00:49<01:03,  5.89it/s]
Testing DataLoader 0:  44%|████▍     | 294/666 [00:49<01:02,  5.91it/s]
Testing DataLoader 0:  44%|████▍     | 295/666 [00:50<01:03,  5.87it/s]
Testing DataLoader 0:  44%|████▍     | 296/666 [00:50<01:02,  5.89it/s]
Testing DataLoader 0:  45%|████▍     | 297/666 [00:50<01:02,  5.91it/s]
Testing DataLoader 0:  45%|████▍     | 298/666 [00:50<01:02,  5.88it/s]
Testing DataLoader 0:  45%|████▍     | 299/666 [00:50<01:02,  5.89it/s]
Testing DataLoader 0:  45%|████▌     | 300/666 [00:50<01:01,  5.91it/s]
Testing DataLoader 0:  45%|████▌     | 301/666 [00:51<01:02,  5.88it/s]
Testing DataLoader 0:  45%|████▌     | 302/666 [00:51<01:01,  5.90it/s]
Testing DataLoader 0:  45%|████▌     | 303/666 [00:51<01:01,  5.91it/s]
Testing DataLoader 0:  46%|████▌     | 304/666 [00:51<01:01,  5.88it/s]
Testing DataLoader 0:  46%|████▌     | 305/666 [00:51<01:01,  5.90it/s]
Testing DataLoader 0:  46%|████▌     | 306/666 [00:51<01:00,  5.92it/s]
Testing DataLoader 0:  46%|████▌     | 307/666 [00:52<01:01,  5.88it/s]
Testing DataLoader 0:  46%|████▌     | 308/666 [00:52<01:00,  5.90it/s]
Testing DataLoader 0:  46%|████▋     | 309/666 [00:52<01:00,  5.92it/s]
Testing DataLoader 0:  47%|████▋     | 310/666 [00:52<01:00,  5.89it/s]
Testing DataLoader 0:  47%|████▋     | 311/666 [00:52<01:00,  5.91it/s]
Testing DataLoader 0:  47%|████▋     | 312/666 [00:52<00:59,  5.93it/s]
Testing DataLoader 0:  47%|████▋     | 313/666 [00:53<00:59,  5.88it/s]
Testing DataLoader 0:  47%|████▋     | 314/666 [00:53<00:59,  5.90it/s]
Testing DataLoader 0:  47%|████▋     | 315/666 [00:53<00:59,  5.92it/s]
Testing DataLoader 0:  47%|████▋     | 316/666 [00:53<00:59,  5.90it/s]
Testing DataLoader 0:  48%|████▊     | 317/666 [00:53<00:59,  5.91it/s]
Testing DataLoader 0:  48%|████▊     | 318/666 [00:53<00:58,  5.93it/s]
Testing DataLoader 0:  48%|████▊     | 319/666 [00:54<00:58,  5.89it/s]
Testing DataLoader 0:  48%|████▊     | 320/666 [00:54<00:58,  5.91it/s]
Testing DataLoader 0:  48%|████▊     | 321/666 [00:54<00:58,  5.93it/s]
Testing DataLoader 0:  48%|████▊     | 322/666 [00:54<00:58,  5.90it/s]
Testing DataLoader 0:  48%|████▊     | 323/666 [00:54<00:57,  5.91it/s]
Testing DataLoader 0:  49%|████▊     | 324/666 [00:54<00:57,  5.93it/s]
Testing DataLoader 0:  49%|████▉     | 325/666 [00:55<00:57,  5.90it/s]
Testing DataLoader 0:  49%|████▉     | 326/666 [00:55<00:57,  5.91it/s]
Testing DataLoader 0:  49%|████▉     | 327/666 [00:55<00:57,  5.93it/s]
Testing DataLoader 0:  49%|████▉     | 328/666 [00:55<00:57,  5.90it/s]
Testing DataLoader 0:  49%|████▉     | 329/666 [00:55<00:56,  5.92it/s]
Testing DataLoader 0:  50%|████▉     | 330/666 [00:55<00:56,  5.93it/s]
Testing DataLoader 0:  50%|████▉     | 331/666 [00:56<00:56,  5.90it/s]
Testing DataLoader 0:  50%|████▉     | 332/666 [00:56<00:56,  5.92it/s]
Testing DataLoader 0:  50%|█████     | 333/666 [00:56<00:56,  5.94it/s]
Testing DataLoader 0:  50%|█████     | 334/666 [00:56<00:56,  5.91it/s]
Testing DataLoader 0:  50%|█████     | 335/666 [00:56<00:55,  5.92it/s]
Testing DataLoader 0:  50%|█████     | 336/666 [00:56<00:55,  5.94it/s]
Testing DataLoader 0:  51%|█████     | 337/666 [00:57<00:55,  5.91it/s]
Testing DataLoader 0:  51%|█████     | 338/666 [00:57<00:55,  5.92it/s]
Testing DataLoader 0:  51%|█████     | 339/666 [00:57<00:55,  5.94it/s]
Testing DataLoader 0:  51%|█████     | 340/666 [00:57<00:55,  5.90it/s]
Testing DataLoader 0:  51%|█████     | 341/666 [00:57<00:54,  5.92it/s]
Testing DataLoader 0:  51%|█████▏    | 342/666 [00:57<00:54,  5.94it/s]
Testing DataLoader 0:  52%|█████▏    | 343/666 [00:58<00:54,  5.91it/s]
Testing DataLoader 0:  52%|█████▏    | 344/666 [00:58<00:54,  5.93it/s]
Testing DataLoader 0:  52%|█████▏    | 345/666 [00:58<00:54,  5.94it/s]
Testing DataLoader 0:  52%|█████▏    | 346/666 [00:58<00:54,  5.91it/s]
Testing DataLoader 0:  52%|█████▏    | 347/666 [00:58<00:53,  5.93it/s]
Testing DataLoader 0:  52%|█████▏    | 348/666 [00:58<00:53,  5.95it/s]
Testing DataLoader 0:  52%|█████▏    | 349/666 [00:59<00:53,  5.91it/s]
Testing DataLoader 0:  53%|█████▎    | 350/666 [00:59<00:53,  5.92it/s]
Testing DataLoader 0:  53%|█████▎    | 351/666 [00:59<00:53,  5.94it/s]
Testing DataLoader 0:  53%|█████▎    | 352/666 [00:59<00:53,  5.90it/s]
Testing DataLoader 0:  53%|█████▎    | 353/666 [00:59<00:52,  5.92it/s]
Testing DataLoader 0:  53%|█████▎    | 354/666 [00:59<00:52,  5.94it/s]
Testing DataLoader 0:  53%|█████▎    | 355/666 [01:00<00:52,  5.91it/s]
Testing DataLoader 0:  53%|█████▎    | 356/666 [01:00<00:52,  5.92it/s]
Testing DataLoader 0:  54%|█████▎    | 357/666 [01:00<00:52,  5.94it/s]
Testing DataLoader 0:  54%|█████▍    | 358/666 [01:00<00:52,  5.90it/s]
Testing DataLoader 0:  54%|█████▍    | 359/666 [01:00<00:51,  5.91it/s]
Testing DataLoader 0:  54%|█████▍    | 360/666 [01:00<00:51,  5.93it/s]
Testing DataLoader 0:  54%|█████▍    | 361/666 [01:01<00:51,  5.90it/s]
Testing DataLoader 0:  54%|█████▍    | 362/666 [01:01<00:51,  5.91it/s]
Testing DataLoader 0:  55%|█████▍    | 363/666 [01:01<00:51,  5.93it/s]
Testing DataLoader 0:  55%|█████▍    | 364/666 [01:01<00:51,  5.89it/s]
Testing DataLoader 0:  55%|█████▍    | 365/666 [01:01<00:50,  5.91it/s]
Testing DataLoader 0:  55%|█████▍    | 366/666 [01:01<00:50,  5.92it/s]
Testing DataLoader 0:  55%|█████▌    | 367/666 [01:02<00:50,  5.90it/s]
Testing DataLoader 0:  55%|█████▌    | 368/666 [01:02<00:50,  5.91it/s]
Testing DataLoader 0:  55%|█████▌    | 369/666 [01:02<00:50,  5.93it/s]
Testing DataLoader 0:  56%|█████▌    | 370/666 [01:02<00:50,  5.91it/s]
Testing DataLoader 0:  56%|█████▌    | 371/666 [01:02<00:49,  5.92it/s]
Testing DataLoader 0:  56%|█████▌    | 372/666 [01:02<00:49,  5.94it/s]
Testing DataLoader 0:  56%|█████▌    | 373/666 [01:03<00:49,  5.92it/s]
Testing DataLoader 0:  56%|█████▌    | 374/666 [01:03<00:49,  5.93it/s]
Testing DataLoader 0:  56%|█████▋    | 375/666 [01:03<00:48,  5.94it/s]
Testing DataLoader 0:  56%|█████▋    | 376/666 [01:03<00:48,  5.92it/s]
Testing DataLoader 0:  57%|█████▋    | 377/666 [01:03<00:48,  5.94it/s]
Testing DataLoader 0:  57%|█████▋    | 378/666 [01:03<00:48,  5.95it/s]
Testing DataLoader 0:  57%|█████▋    | 379/666 [01:04<00:48,  5.92it/s]
Testing DataLoader 0:  57%|█████▋    | 380/666 [01:04<00:48,  5.94it/s]
Testing DataLoader 0:  57%|█████▋    | 381/666 [01:04<00:47,  5.95it/s]
Testing DataLoader 0:  57%|█████▋    | 382/666 [01:04<00:47,  5.93it/s]
Testing DataLoader 0:  58%|█████▊    | 383/666 [01:04<00:47,  5.94it/s]
Testing DataLoader 0:  58%|█████▊    | 384/666 [01:04<00:47,  5.96it/s]
Testing DataLoader 0:  58%|█████▊    | 385/666 [01:05<00:47,  5.92it/s]
Testing DataLoader 0:  58%|█████▊    | 386/666 [01:05<00:47,  5.94it/s]
Testing DataLoader 0:  58%|█████▊    | 387/666 [01:05<00:46,  5.95it/s]
Testing DataLoader 0:  58%|█████▊    | 388/666 [01:05<00:46,  5.92it/s]
Testing DataLoader 0:  58%|█████▊    | 389/666 [01:05<00:46,  5.94it/s]
Testing DataLoader 0:  59%|█████▊    | 390/666 [01:05<00:46,  5.95it/s]
Testing DataLoader 0:  59%|█████▊    | 391/666 [01:06<00:46,  5.92it/s]
Testing DataLoader 0:  59%|█████▉    | 392/666 [01:06<00:46,  5.93it/s]
Testing DataLoader 0:  59%|█████▉    | 393/666 [01:06<00:45,  5.95it/s]
Testing DataLoader 0:  59%|█████▉    | 394/666 [01:06<00:45,  5.92it/s]
Testing DataLoader 0:  59%|█████▉    | 395/666 [01:06<00:45,  5.93it/s]
Testing DataLoader 0:  59%|█████▉    | 396/666 [01:06<00:45,  5.95it/s]
Testing DataLoader 0:  60%|█████▉    | 397/666 [01:07<00:45,  5.92it/s]
Testing DataLoader 0:  60%|█████▉    | 398/666 [01:07<00:45,  5.93it/s]
Testing DataLoader 0:  60%|█████▉    | 399/666 [01:07<00:44,  5.94it/s]
Testing DataLoader 0:  60%|██████    | 400/666 [01:07<00:44,  5.91it/s]
Testing DataLoader 0:  60%|██████    | 401/666 [01:07<00:44,  5.93it/s]
Testing DataLoader 0:  60%|██████    | 402/666 [01:07<00:44,  5.94it/s]
Testing DataLoader 0:  61%|██████    | 403/666 [01:08<00:44,  5.92it/s]
Testing DataLoader 0:  61%|██████    | 404/666 [01:08<00:44,  5.94it/s]
Testing DataLoader 0:  61%|██████    | 405/666 [01:08<00:43,  5.95it/s]
Testing DataLoader 0:  61%|██████    | 406/666 [01:08<00:43,  5.92it/s]
Testing DataLoader 0:  61%|██████    | 407/666 [01:08<00:43,  5.94it/s]
Testing DataLoader 0:  61%|██████▏   | 408/666 [01:08<00:43,  5.95it/s]
Testing DataLoader 0:  61%|██████▏   | 409/666 [01:08<00:43,  5.93it/s]
Testing DataLoader 0:  62%|██████▏   | 410/666 [01:08<00:43,  5.94it/s]
Testing DataLoader 0:  62%|██████▏   | 411/666 [01:09<00:42,  5.96it/s]
Testing DataLoader 0:  62%|██████▏   | 412/666 [01:09<00:42,  5.93it/s]
Testing DataLoader 0:  62%|██████▏   | 413/666 [01:09<00:42,  5.94it/s]
Testing DataLoader 0:  62%|██████▏   | 414/666 [01:09<00:42,  5.96it/s]
Testing DataLoader 0:  62%|██████▏   | 415/666 [01:10<00:42,  5.93it/s]
Testing DataLoader 0:  62%|██████▏   | 416/666 [01:10<00:42,  5.94it/s]
Testing DataLoader 0:  63%|██████▎   | 417/666 [01:10<00:41,  5.95it/s]
Testing DataLoader 0:  63%|██████▎   | 418/666 [01:10<00:41,  5.92it/s]
Testing DataLoader 0:  63%|██████▎   | 419/666 [01:10<00:41,  5.94it/s]
Testing DataLoader 0:  63%|██████▎   | 420/666 [01:10<00:41,  5.95it/s]
Testing DataLoader 0:  63%|██████▎   | 421/666 [01:11<00:41,  5.93it/s]
Testing DataLoader 0:  63%|██████▎   | 422/666 [01:11<00:41,  5.94it/s]
Testing DataLoader 0:  64%|██████▎   | 423/666 [01:11<00:40,  5.95it/s]
Testing DataLoader 0:  64%|██████▎   | 424/666 [01:11<00:40,  5.93it/s]
Testing DataLoader 0:  64%|██████▍   | 425/666 [01:11<00:40,  5.94it/s]
Testing DataLoader 0:  64%|██████▍   | 426/666 [01:11<00:40,  5.95it/s]
Testing DataLoader 0:  64%|██████▍   | 427/666 [01:12<00:40,  5.92it/s]
Testing DataLoader 0:  64%|██████▍   | 428/666 [01:12<00:40,  5.94it/s]
Testing DataLoader 0:  64%|██████▍   | 429/666 [01:12<00:39,  5.95it/s]
Testing DataLoader 0:  65%|██████▍   | 430/666 [01:12<00:39,  5.92it/s]
Testing DataLoader 0:  65%|██████▍   | 431/666 [01:12<00:39,  5.94it/s]
Testing DataLoader 0:  65%|██████▍   | 432/666 [01:12<00:39,  5.95it/s]
Testing DataLoader 0:  65%|██████▌   | 433/666 [01:13<00:39,  5.92it/s]
Testing DataLoader 0:  65%|██████▌   | 434/666 [01:13<00:39,  5.93it/s]
Testing DataLoader 0:  65%|██████▌   | 435/666 [01:13<00:38,  5.95it/s]
Testing DataLoader 0:  65%|██████▌   | 436/666 [01:13<00:38,  5.93it/s]
Testing DataLoader 0:  66%|██████▌   | 437/666 [01:13<00:38,  5.94it/s]
Testing DataLoader 0:  66%|██████▌   | 438/666 [01:13<00:38,  5.95it/s]
Testing DataLoader 0:  66%|██████▌   | 439/666 [01:14<00:38,  5.93it/s]
Testing DataLoader 0:  66%|██████▌   | 440/666 [01:14<00:38,  5.94it/s]
Testing DataLoader 0:  66%|██████▌   | 441/666 [01:14<00:37,  5.96it/s]
Testing DataLoader 0:  66%|██████▋   | 442/666 [01:14<00:37,  5.92it/s]
Testing DataLoader 0:  67%|██████▋   | 443/666 [01:14<00:37,  5.94it/s]
Testing DataLoader 0:  67%|██████▋   | 444/666 [01:14<00:37,  5.95it/s]
Testing DataLoader 0:  67%|██████▋   | 445/666 [01:15<00:37,  5.92it/s]
Testing DataLoader 0:  67%|██████▋   | 446/666 [01:15<00:37,  5.93it/s]
Testing DataLoader 0:  67%|██████▋   | 447/666 [01:15<00:36,  5.94it/s]
Testing DataLoader 0:  67%|██████▋   | 448/666 [01:15<00:36,  5.92it/s]
Testing DataLoader 0:  67%|██████▋   | 449/666 [01:15<00:36,  5.93it/s]
Testing DataLoader 0:  68%|██████▊   | 450/666 [01:15<00:36,  5.94it/s]
Testing DataLoader 0:  68%|██████▊   | 451/666 [01:16<00:36,  5.92it/s]
Testing DataLoader 0:  68%|██████▊   | 452/666 [01:16<00:36,  5.93it/s]
Testing DataLoader 0:  68%|██████▊   | 453/666 [01:16<00:35,  5.95it/s]
Testing DataLoader 0:  68%|██████▊   | 454/666 [01:16<00:35,  5.93it/s]
Testing DataLoader 0:  68%|██████▊   | 455/666 [01:16<00:35,  5.94it/s]
Testing DataLoader 0:  68%|██████▊   | 456/666 [01:16<00:35,  5.95it/s]
Testing DataLoader 0:  69%|██████▊   | 457/666 [01:17<00:35,  5.92it/s]
Testing DataLoader 0:  69%|██████▉   | 458/666 [01:17<00:35,  5.94it/s]
Testing DataLoader 0:  69%|██████▉   | 459/666 [01:17<00:34,  5.95it/s]
Testing DataLoader 0:  69%|██████▉   | 460/666 [01:17<00:34,  5.92it/s]
Testing DataLoader 0:  69%|██████▉   | 461/666 [01:17<00:34,  5.94it/s]
Testing DataLoader 0:  69%|██████▉   | 462/666 [01:17<00:34,  5.95it/s]
Testing DataLoader 0:  70%|██████▉   | 463/666 [01:18<00:34,  5.93it/s]
Testing DataLoader 0:  70%|██████▉   | 464/666 [01:18<00:34,  5.94it/s]
Testing DataLoader 0:  70%|██████▉   | 465/666 [01:18<00:33,  5.95it/s]
Testing DataLoader 0:  70%|██████▉   | 466/666 [01:18<00:33,  5.93it/s]
Testing DataLoader 0:  70%|███████   | 467/666 [01:18<00:33,  5.94it/s]
Testing DataLoader 0:  70%|███████   | 468/666 [01:18<00:33,  5.95it/s]
Testing DataLoader 0:  70%|███████   | 469/666 [01:19<00:33,  5.93it/s]
Testing DataLoader 0:  71%|███████   | 470/666 [01:19<00:32,  5.94it/s]
Testing DataLoader 0:  71%|███████   | 471/666 [01:19<00:32,  5.95it/s]
Testing DataLoader 0:  71%|███████   | 472/666 [01:19<00:32,  5.93it/s]
Testing DataLoader 0:  71%|███████   | 473/666 [01:19<00:32,  5.94it/s]
Testing DataLoader 0:  71%|███████   | 474/666 [01:19<00:32,  5.95it/s]
Testing DataLoader 0:  71%|███████▏  | 475/666 [01:20<00:32,  5.93it/s]
Testing DataLoader 0:  71%|███████▏  | 476/666 [01:20<00:31,  5.94it/s]
Testing DataLoader 0:  72%|███████▏  | 477/666 [01:20<00:31,  5.95it/s]
Testing DataLoader 0:  72%|███████▏  | 478/666 [01:20<00:31,  5.93it/s]
Testing DataLoader 0:  72%|███████▏  | 479/666 [01:20<00:31,  5.94it/s]
Testing DataLoader 0:  72%|███████▏  | 480/666 [01:20<00:31,  5.95it/s]
Testing DataLoader 0:  72%|███████▏  | 481/666 [01:21<00:31,  5.93it/s]
Testing DataLoader 0:  72%|███████▏  | 482/666 [01:21<00:30,  5.95it/s]
Testing DataLoader 0:  73%|███████▎  | 483/666 [01:21<00:30,  5.96it/s]
Testing DataLoader 0:  73%|███████▎  | 484/666 [01:21<00:30,  5.93it/s]
Testing DataLoader 0:  73%|███████▎  | 485/666 [01:21<00:30,  5.95it/s]
Testing DataLoader 0:  73%|███████▎  | 486/666 [01:21<00:30,  5.96it/s]
Testing DataLoader 0:  73%|███████▎  | 487/666 [01:22<00:30,  5.94it/s]
Testing DataLoader 0:  73%|███████▎  | 488/666 [01:22<00:29,  5.95it/s]
Testing DataLoader 0:  73%|███████▎  | 489/666 [01:22<00:29,  5.96it/s]
Testing DataLoader 0:  74%|███████▎  | 490/666 [01:22<00:29,  5.94it/s]
Testing DataLoader 0:  74%|███████▎  | 491/666 [01:22<00:29,  5.95it/s]
Testing DataLoader 0:  74%|███████▍  | 492/666 [01:22<00:29,  5.96it/s]
Testing DataLoader 0:  74%|███████▍  | 493/666 [01:22<00:29,  5.94it/s]
Testing DataLoader 0:  74%|███████▍  | 494/666 [01:22<00:28,  5.95it/s]
Testing DataLoader 0:  74%|███████▍  | 495/666 [01:22<00:28,  5.97it/s]
Testing DataLoader 0:  74%|███████▍  | 496/666 [01:23<00:28,  5.94it/s]
Testing DataLoader 0:  75%|███████▍  | 497/666 [01:23<00:28,  5.95it/s]
Testing DataLoader 0:  75%|███████▍  | 498/666 [01:23<00:28,  5.96it/s]
Testing DataLoader 0:  75%|███████▍  | 499/666 [01:24<00:28,  5.94it/s]
Testing DataLoader 0:  75%|███████▌  | 500/666 [01:24<00:27,  5.95it/s]
Testing DataLoader 0:  75%|███████▌  | 501/666 [01:24<00:27,  5.96it/s]
Testing DataLoader 0:  75%|███████▌  | 502/666 [01:24<00:27,  5.94it/s]
Testing DataLoader 0:  76%|███████▌  | 503/666 [01:24<00:27,  5.95it/s]
Testing DataLoader 0:  76%|███████▌  | 504/666 [01:24<00:27,  5.96it/s]
Testing DataLoader 0:  76%|███████▌  | 505/666 [01:25<00:27,  5.93it/s]
Testing DataLoader 0:  76%|███████▌  | 506/666 [01:25<00:26,  5.95it/s]
Testing DataLoader 0:  76%|███████▌  | 507/666 [01:25<00:26,  5.96it/s]
Testing DataLoader 0:  76%|███████▋  | 508/666 [01:25<00:26,  5.94it/s]
Testing DataLoader 0:  76%|███████▋  | 509/666 [01:25<00:26,  5.95it/s]
Testing DataLoader 0:  77%|███████▋  | 510/666 [01:25<00:26,  5.96it/s]
Testing DataLoader 0:  77%|███████▋  | 511/666 [01:26<00:26,  5.94it/s]
Testing DataLoader 0:  77%|███████▋  | 512/666 [01:26<00:25,  5.95it/s]
Testing DataLoader 0:  77%|███████▋  | 513/666 [01:26<00:25,  5.96it/s]
Testing DataLoader 0:  77%|███████▋  | 514/666 [01:26<00:25,  5.94it/s]
Testing DataLoader 0:  77%|███████▋  | 515/666 [01:26<00:25,  5.95it/s]
Testing DataLoader 0:  77%|███████▋  | 516/666 [01:26<00:25,  5.96it/s]
Testing DataLoader 0:  78%|███████▊  | 517/666 [01:27<00:25,  5.94it/s]
Testing DataLoader 0:  78%|███████▊  | 518/666 [01:27<00:24,  5.95it/s]
Testing DataLoader 0:  78%|███████▊  | 519/666 [01:27<00:24,  5.96it/s]
Testing DataLoader 0:  78%|███████▊  | 520/666 [01:27<00:24,  5.95it/s]
Testing DataLoader 0:  78%|███████▊  | 521/666 [01:27<00:24,  5.96it/s]
Testing DataLoader 0:  78%|███████▊  | 522/666 [01:27<00:24,  5.97it/s]
Testing DataLoader 0:  79%|███████▊  | 523/666 [01:28<00:24,  5.94it/s]
Testing DataLoader 0:  79%|███████▊  | 524/666 [01:28<00:23,  5.95it/s]
Testing DataLoader 0:  79%|███████▉  | 525/666 [01:28<00:23,  5.96it/s]
Testing DataLoader 0:  79%|███████▉  | 526/666 [01:28<00:23,  5.94it/s]
Testing DataLoader 0:  79%|███████▉  | 527/666 [01:28<00:23,  5.95it/s]
Testing DataLoader 0:  79%|███████▉  | 528/666 [01:28<00:23,  5.96it/s]
Testing DataLoader 0:  79%|███████▉  | 529/666 [01:29<00:23,  5.94it/s]
Testing DataLoader 0:  80%|███████▉  | 530/666 [01:29<00:22,  5.95it/s]
Testing DataLoader 0:  80%|███████▉  | 531/666 [01:29<00:22,  5.96it/s]
Testing DataLoader 0:  80%|███████▉  | 532/666 [01:29<00:22,  5.95it/s]
Testing DataLoader 0:  80%|████████  | 533/666 [01:29<00:22,  5.96it/s]
Testing DataLoader 0:  80%|████████  | 534/666 [01:29<00:22,  5.97it/s]
Testing DataLoader 0:  80%|████████  | 535/666 [01:29<00:22,  5.95it/s]
Testing DataLoader 0:  80%|████████  | 536/666 [01:29<00:21,  5.96it/s]
Testing DataLoader 0:  81%|████████  | 537/666 [01:29<00:21,  5.97it/s]
Testing DataLoader 0:  81%|████████  | 538/666 [01:30<00:21,  5.95it/s]
Testing DataLoader 0:  81%|████████  | 539/666 [01:30<00:21,  5.96it/s]
Testing DataLoader 0:  81%|████████  | 540/666 [01:30<00:21,  5.97it/s]
Testing DataLoader 0:  81%|████████  | 541/666 [01:30<00:21,  5.95it/s]
Testing DataLoader 0:  81%|████████▏ | 542/666 [01:30<00:20,  5.96it/s]
Testing DataLoader 0:  82%|████████▏ | 543/666 [01:31<00:20,  5.97it/s]
Testing DataLoader 0:  82%|████████▏ | 544/666 [01:31<00:20,  5.95it/s]
Testing DataLoader 0:  82%|████████▏ | 545/666 [01:31<00:20,  5.96it/s]
Testing DataLoader 0:  82%|████████▏ | 546/666 [01:31<00:20,  5.97it/s]
Testing DataLoader 0:  82%|████████▏ | 547/666 [01:31<00:19,  5.95it/s]
Testing DataLoader 0:  82%|████████▏ | 548/666 [01:31<00:19,  5.96it/s]
Testing DataLoader 0:  82%|████████▏ | 549/666 [01:31<00:19,  5.97it/s]
Testing DataLoader 0:  83%|████████▎ | 550/666 [01:32<00:19,  5.96it/s]
Testing DataLoader 0:  83%|████████▎ | 551/666 [01:32<00:19,  5.97it/s]
Testing DataLoader 0:  83%|████████▎ | 552/666 [01:32<00:19,  5.98it/s]
Testing DataLoader 0:  83%|████████▎ | 553/666 [01:32<00:18,  5.96it/s]
Testing DataLoader 0:  83%|████████▎ | 554/666 [01:32<00:18,  5.97it/s]
Testing DataLoader 0:  83%|████████▎ | 555/666 [01:32<00:18,  5.98it/s]
Testing DataLoader 0:  83%|████████▎ | 556/666 [01:33<00:18,  5.96it/s]
Testing DataLoader 0:  84%|████████▎ | 557/666 [01:33<00:18,  5.97it/s]
Testing DataLoader 0:  84%|████████▍ | 558/666 [01:33<00:18,  5.98it/s]
Testing DataLoader 0:  84%|████████▍ | 559/666 [01:33<00:17,  5.96it/s]
Testing DataLoader 0:  84%|████████▍ | 560/666 [01:33<00:17,  5.97it/s]
Testing DataLoader 0:  84%|████████▍ | 561/666 [01:33<00:17,  5.98it/s]
Testing DataLoader 0:  84%|████████▍ | 562/666 [01:34<00:17,  5.96it/s]
Testing DataLoader 0:  85%|████████▍ | 563/666 [01:34<00:17,  5.97it/s]
Testing DataLoader 0:  85%|████████▍ | 564/666 [01:34<00:17,  5.98it/s]
Testing DataLoader 0:  85%|████████▍ | 565/666 [01:34<00:16,  5.96it/s]
Testing DataLoader 0:  85%|████████▍ | 566/666 [01:34<00:16,  5.97it/s]
Testing DataLoader 0:  85%|████████▌ | 567/666 [01:34<00:16,  5.98it/s]
Testing DataLoader 0:  85%|████████▌ | 568/666 [01:35<00:16,  5.96it/s]
Testing DataLoader 0:  85%|████████▌ | 569/666 [01:35<00:16,  5.97it/s]
Testing DataLoader 0:  86%|████████▌ | 570/666 [01:35<00:16,  5.98it/s]
Testing DataLoader 0:  86%|████████▌ | 571/666 [01:35<00:15,  5.96it/s]
Testing DataLoader 0:  86%|████████▌ | 572/666 [01:35<00:15,  5.97it/s]
Testing DataLoader 0:  86%|████████▌ | 573/666 [01:35<00:15,  5.98it/s]
Testing DataLoader 0:  86%|████████▌ | 574/666 [01:36<00:15,  5.95it/s]
Testing DataLoader 0:  86%|████████▋ | 575/666 [01:36<00:15,  5.96it/s]
Testing DataLoader 0:  86%|████████▋ | 576/666 [01:36<00:15,  5.97it/s]
Testing DataLoader 0:  87%|████████▋ | 577/666 [01:36<00:14,  5.95it/s]
Testing DataLoader 0:  87%|████████▋ | 578/666 [01:36<00:14,  5.96it/s]
Testing DataLoader 0:  87%|████████▋ | 579/666 [01:36<00:14,  5.97it/s]
Testing DataLoader 0:  87%|████████▋ | 580/666 [01:37<00:14,  5.95it/s]
Testing DataLoader 0:  87%|████████▋ | 581/666 [01:37<00:14,  5.96it/s]
Testing DataLoader 0:  87%|████████▋ | 582/666 [01:37<00:14,  5.97it/s]
Testing DataLoader 0:  88%|████████▊ | 583/666 [01:37<00:13,  5.95it/s]
Testing DataLoader 0:  88%|████████▊ | 584/666 [01:37<00:13,  5.96it/s]
Testing DataLoader 0:  88%|████████▊ | 585/666 [01:37<00:13,  5.97it/s]
Testing DataLoader 0:  88%|████████▊ | 586/666 [01:38<00:13,  5.95it/s]
Testing DataLoader 0:  88%|████████▊ | 587/666 [01:38<00:13,  5.96it/s]
Testing DataLoader 0:  88%|████████▊ | 588/666 [01:38<00:13,  5.97it/s]
Testing DataLoader 0:  88%|████████▊ | 589/666 [01:38<00:12,  5.96it/s]
Testing DataLoader 0:  89%|████████▊ | 590/666 [01:38<00:12,  5.96it/s]
Testing DataLoader 0:  89%|████████▊ | 591/666 [01:38<00:12,  5.97it/s]
Testing DataLoader 0:  89%|████████▉ | 592/666 [01:39<00:12,  5.96it/s]
Testing DataLoader 0:  89%|████████▉ | 593/666 [01:39<00:12,  5.96it/s]
Testing DataLoader 0:  89%|████████▉ | 594/666 [01:39<00:12,  5.97it/s]
Testing DataLoader 0:  89%|████████▉ | 595/666 [01:39<00:11,  5.96it/s]
Testing DataLoader 0:  89%|████████▉ | 596/666 [01:39<00:11,  5.97it/s]
Testing DataLoader 0:  90%|████████▉ | 597/666 [01:39<00:11,  5.98it/s]
Testing DataLoader 0:  90%|████████▉ | 598/666 [01:40<00:11,  5.96it/s]
Testing DataLoader 0:  90%|████████▉ | 599/666 [01:40<00:11,  5.97it/s]
Testing DataLoader 0:  90%|█████████ | 600/666 [01:40<00:11,  5.98it/s]
Testing DataLoader 0:  90%|█████████ | 601/666 [01:40<00:10,  5.96it/s]
Testing DataLoader 0:  90%|█████████ | 602/666 [01:40<00:10,  5.97it/s]
Testing DataLoader 0:  91%|█████████ | 603/666 [01:40<00:10,  5.98it/s]
Testing DataLoader 0:  91%|█████████ | 604/666 [01:41<00:10,  5.96it/s]
Testing DataLoader 0:  91%|█████████ | 605/666 [01:41<00:10,  5.97it/s]
Testing DataLoader 0:  91%|█████████ | 606/666 [01:41<00:10,  5.98it/s]
Testing DataLoader 0:  91%|█████████ | 607/666 [01:41<00:09,  5.96it/s]
Testing DataLoader 0:  91%|█████████▏| 608/666 [01:41<00:09,  5.97it/s]
Testing DataLoader 0:  91%|█████████▏| 609/666 [01:41<00:09,  5.97it/s]
Testing DataLoader 0:  92%|█████████▏| 610/666 [01:42<00:09,  5.95it/s]
Testing DataLoader 0:  92%|█████████▏| 611/666 [01:42<00:09,  5.96it/s]
Testing DataLoader 0:  92%|█████████▏| 612/666 [01:42<00:09,  5.97it/s]
Testing DataLoader 0:  92%|█████████▏| 613/666 [01:42<00:08,  5.95it/s]
Testing DataLoader 0:  92%|█████████▏| 614/666 [01:42<00:08,  5.96it/s]
Testing DataLoader 0:  92%|█████████▏| 615/666 [01:42<00:08,  5.97it/s]
Testing DataLoader 0:  92%|█████████▏| 616/666 [01:43<00:08,  5.96it/s]
Testing DataLoader 0:  93%|█████████▎| 617/666 [01:43<00:08,  5.96it/s]
Testing DataLoader 0:  93%|█████████▎| 618/666 [01:43<00:08,  5.97it/s]
Testing DataLoader 0:  93%|█████████▎| 619/666 [01:43<00:07,  5.96it/s]
Testing DataLoader 0:  93%|█████████▎| 620/666 [01:43<00:07,  5.97it/s]
Testing DataLoader 0:  93%|█████████▎| 621/666 [01:43<00:07,  5.98it/s]
Testing DataLoader 0:  93%|█████████▎| 622/666 [01:44<00:07,  5.96it/s]
Testing DataLoader 0:  94%|█████████▎| 623/666 [01:44<00:07,  5.97it/s]
Testing DataLoader 0:  94%|█████████▎| 624/666 [01:44<00:07,  5.98it/s]
Testing DataLoader 0:  94%|█████████▍| 625/666 [01:44<00:06,  5.96it/s]
Testing DataLoader 0:  94%|█████████▍| 626/666 [01:44<00:06,  5.97it/s]
Testing DataLoader 0:  94%|█████████▍| 627/666 [01:44<00:06,  5.98it/s]
Testing DataLoader 0:  94%|█████████▍| 628/666 [01:45<00:06,  5.96it/s]
Testing DataLoader 0:  94%|█████████▍| 629/666 [01:45<00:06,  5.97it/s]
Testing DataLoader 0:  95%|█████████▍| 630/666 [01:45<00:06,  5.98it/s]
Testing DataLoader 0:  95%|█████████▍| 631/666 [01:45<00:05,  5.97it/s]
Testing DataLoader 0:  95%|█████████▍| 632/666 [01:45<00:05,  5.98it/s]
Testing DataLoader 0:  95%|█████████▌| 633/666 [01:45<00:05,  5.99it/s]
Testing DataLoader 0:  95%|█████████▌| 634/666 [01:46<00:05,  5.97it/s]
Testing DataLoader 0:  95%|█████████▌| 635/666 [01:46<00:05,  5.98it/s]
Testing DataLoader 0:  95%|█████████▌| 636/666 [01:46<00:05,  5.99it/s]
Testing DataLoader 0:  96%|█████████▌| 637/666 [01:46<00:04,  5.97it/s]
Testing DataLoader 0:  96%|█████████▌| 638/666 [01:46<00:04,  5.98it/s]
Testing DataLoader 0:  96%|█████████▌| 639/666 [01:46<00:04,  5.99it/s]
Testing DataLoader 0:  96%|█████████▌| 640/666 [01:47<00:04,  5.97it/s]
Testing DataLoader 0:  96%|█████████▌| 641/666 [01:47<00:04,  5.98it/s]
Testing DataLoader 0:  96%|█████████▋| 642/666 [01:47<00:04,  5.99it/s]
Testing DataLoader 0:  97%|█████████▋| 643/666 [01:47<00:03,  5.97it/s]
Testing DataLoader 0:  97%|█████████▋| 644/666 [01:47<00:03,  5.98it/s]
Testing DataLoader 0:  97%|█████████▋| 645/666 [01:47<00:03,  5.99it/s]
Testing DataLoader 0:  97%|█████████▋| 646/666 [01:48<00:03,  5.97it/s]
Testing DataLoader 0:  97%|█████████▋| 647/666 [01:48<00:03,  5.98it/s]
Testing DataLoader 0:  97%|█████████▋| 648/666 [01:48<00:03,  5.99it/s]
Testing DataLoader 0:  97%|█████████▋| 649/666 [01:48<00:02,  5.97it/s]
Testing DataLoader 0:  98%|█████████▊| 650/666 [01:48<00:02,  5.98it/s]
Testing DataLoader 0:  98%|█████████▊| 651/666 [01:48<00:02,  5.99it/s]
Testing DataLoader 0:  98%|█████████▊| 652/666 [01:49<00:02,  5.97it/s]
Testing DataLoader 0:  98%|█████████▊| 653/666 [01:49<00:02,  5.98it/s]
Testing DataLoader 0:  98%|█████████▊| 654/666 [01:49<00:02,  5.99it/s]
Testing DataLoader 0:  98%|█████████▊| 655/666 [01:49<00:01,  5.97it/s]
Testing DataLoader 0:  98%|█████████▊| 656/666 [01:49<00:01,  5.98it/s]
Testing DataLoader 0:  99%|█████████▊| 657/666 [01:49<00:01,  5.99it/s]
Testing DataLoader 0:  99%|█████████▉| 658/666 [01:50<00:01,  5.97it/s]
Testing DataLoader 0:  99%|█████████▉| 659/666 [01:50<00:01,  5.98it/s]
Testing DataLoader 0:  99%|█████████▉| 660/666 [01:50<00:01,  5.99it/s]
Testing DataLoader 0:  99%|█████████▉| 661/666 [01:50<00:00,  5.97it/s]
Testing DataLoader 0:  99%|█████████▉| 662/666 [01:50<00:00,  5.98it/s]
Testing DataLoader 0: 100%|█████████▉| 663/666 [01:50<00:00,  5.99it/s]
Testing DataLoader 0: 100%|█████████▉| 664/666 [01:51<00:00,  5.97it/s]
Testing DataLoader 0: 100%|█████████▉| 665/666 [01:51<00:00,  5.98it/s]
Testing DataLoader 0: 100%|██████████| 666/666 [01:51<00:00,  5.96it/s]
Testing DataLoader 0: 100%|██████████| 666/666 [01:51<00:00,  5.96it/s]
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
       Test metric             DataLoader 0
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
       R@1 (test)           0.3535875380039215
       R@20 (test)          0.7683132886886597
       R@5 (test)            0.593163013458252
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
    10809.48 real     31782.56 user       302.14 sys
          5773361152  maximum resident set size
                   0  average shared memory size
                   0  average unshared data size
                   0  average unshared stack size
            64155155  page reclaims
               24999  page faults
                   0  swaps
                   0  block input operations
                   0  block output operations
              236208  messages sent
              118104  messages received
                  18  signals received
              359745  voluntary context switches
            27256007  involuntary context switches
       3098021940609  instructions retired
       1316739535020  cycles elapsed
          8252872896  peak memory footprint
exit 0 end 2026-10-05T08:54:24Z
