search

swuxyj / DeepHash-pytorch

Implementation of Some Deep Hash Algorithms, Including DPSH、DSH、DHN、HashNet、DSDH、DTSH、DFH、GreedyHash、CSQ.
MIT License
495 stars 116 forks source link

请问那两种无监督方式的效果如何呢？ #18

Closed Tramac closed 3 years ago

Tramac commented 3 years ago

请问Unsupervised_BiHalf.py与Unsupervised_GreedyHash.py两种方式的效果有测试么，结果如何呢？

swuxyj commented 3 years ago

你好，这是我刚刚跑的两次结果，可能有一些参数需要调整

[BiHalf Unsupervised] epoch:190, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][191/200][01:47:21] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171319272
[BiHalf Unsupervised][192/200][01:49:35] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169869841
[BiHalf Unsupervised][193/200][01:51:48] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.172836766
[BiHalf Unsupervised][194/200][01:54:02] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.172774910
[BiHalf Unsupervised][195/200][01:56:17] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171555375
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 190.31it/s]
[BiHalf Unsupervised] epoch:195, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][196/200][02:02:04] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169292366
[BiHalf Unsupervised][197/200][02:04:18] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171099631
[BiHalf Unsupervised][198/200][02:06:32] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169411861
[BiHalf Unsupervised][199/200][02:08:46] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.172826315
[BiHalf Unsupervised][200/200][02:10:59] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171521300
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:36<00:00, 270.74it/s]
[BiHalf Unsupervised] epoch:200, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised] epoch:190, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][191/200][01:43:18] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068403235
[GreedyHash Unsupervised][192/200][01:45:32] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.069305371
[GreedyHash Unsupervised][193/200][01:47:48] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068805290
[GreedyHash Unsupervised][194/200][01:50:02] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.067952971
[GreedyHash Unsupervised][195/200][01:52:16] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068188658
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 217.07it/s]
[GreedyHash Unsupervised] epoch:195, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][196/200][01:57:57] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068316468
[GreedyHash Unsupervised][197/200][02:00:12] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068122626
[GreedyHash Unsupervised][198/200][02:02:27] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068907434
[GreedyHash Unsupervised][199/200][02:04:42] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.067760388
[GreedyHash Unsupervised][200/200][02:06:56] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068603883
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:42<00:00, 235.86it/s]
[GreedyHash Unsupervised] epoch:200, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
Tramac commented 3 years ago

感谢回复，方便多贴一些epoch的train loss么，想看一下loss有没有收敛，因为我自己训的loss没有收敛的趋势

swuxyj commented 3 years ago

感谢回复，方便多贴一些epoch的train loss么，想看一下loss有没有收敛，因为我自己训的loss没有收敛的趋势

还好没关掉IDE，我这里效果感觉也不算很好，可能是有些细节没实现对

{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}}
Files already downloaded and verified
train_dataset 50000
test_dataset 10000
database_dataset 50000
[GreedyHash Unsupervised][ 1/200][16:25:18] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.086971987
[GreedyHash Unsupervised][ 2/200][16:27:33] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.081444918
[GreedyHash Unsupervised][ 3/200][16:29:48] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.079732886
[GreedyHash Unsupervised][ 4/200][16:32:04] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.078639681
[GreedyHash Unsupervised][ 5/200][16:34:20] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.076589688
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.77it/s]
100%|█████████████████████████████████████████| 782/782 [02:15<00:00,  5.77it/s]
100%|████████████████████████████████████| 10000/10000 [00:56<00:00, 176.45it/s]
[GreedyHash Unsupervised] epoch:5, bit:16, dataset:cifar10-2, MAP:0.358, Best MAP: 0.358
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][ 6/200][16:40:16] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.075557183
[GreedyHash Unsupervised][ 7/200][16:42:29] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.074674132
[GreedyHash Unsupervised][ 8/200][16:44:43] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.074944721
[GreedyHash Unsupervised][ 9/200][16:46:57] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.073286804
[GreedyHash Unsupervised][10/200][16:49:11] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.072839194
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.84it/s]
100%|████████████████████████████████████| 10000/10000 [01:01<00:00, 163.78it/s]
[GreedyHash Unsupervised] epoch:10, bit:16, dataset:cifar10-2, MAP:0.397, Best MAP: 0.397
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][11/200][16:55:06] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.072911030
[GreedyHash Unsupervised][12/200][16:57:20] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.071976886
[GreedyHash Unsupervised][13/200][16:59:34] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.072802947
[GreedyHash Unsupervised][14/200][17:01:48] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.071350123
[GreedyHash Unsupervised][15/200][17:04:02] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.070942914
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [01:04<00:00, 154.84it/s]
[GreedyHash Unsupervised] epoch:15, bit:16, dataset:cifar10-2, MAP:0.404, Best MAP: 0.404
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][16/200][17:10:01] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.071224659
[GreedyHash Unsupervised][17/200][17:12:15] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.070200627
[GreedyHash Unsupervised][18/200][17:14:29] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.070563459
[GreedyHash Unsupervised][19/200][17:16:43] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.070370265
[GreedyHash Unsupervised][20/200][17:18:57] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.070270364
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:58<00:00, 170.03it/s]
[GreedyHash Unsupervised] epoch:20, bit:16, dataset:cifar10-2, MAP:0.408, Best MAP: 0.408
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][21/200][17:24:50] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069599386
[GreedyHash Unsupervised][22/200][17:27:05] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.070187019
[GreedyHash Unsupervised][23/200][17:29:20] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.068669633
[GreedyHash Unsupervised][24/200][17:31:34] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069454180
[GreedyHash Unsupervised][25/200][17:33:48] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069393203
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [01:01<00:00, 162.28it/s]
[GreedyHash Unsupervised] epoch:25, bit:16, dataset:cifar10-2, MAP:0.416, Best MAP: 0.416
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][26/200][17:39:43] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069108784
[GreedyHash Unsupervised][27/200][17:41:59] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069198017
[GreedyHash Unsupervised][28/200][17:44:13] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069431895
[GreedyHash Unsupervised][29/200][17:46:28] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069195433
[GreedyHash Unsupervised][30/200][17:48:42] bit:16, lr:0.000100000, dataset:cifar10-2, train loss:0.069549680
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:59<00:00, 167.26it/s]
[GreedyHash Unsupervised] epoch:30, bit:16, dataset:cifar10-2, MAP:0.417, Best MAP: 0.417
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][31/200][17:54:36] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068288604
[GreedyHash Unsupervised][32/200][17:56:50] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068657745
[GreedyHash Unsupervised][33/200][17:59:04] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068967392
[GreedyHash Unsupervised][34/200][18:01:18] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068450950
[GreedyHash Unsupervised][35/200][18:03:32] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068775797
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:58<00:00, 170.58it/s]
[GreedyHash Unsupervised] epoch:35, bit:16, dataset:cifar10-2, MAP:0.420, Best MAP: 0.420
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][36/200][18:09:25] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.069065871
[GreedyHash Unsupervised][37/200][18:11:40] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.069069995
[GreedyHash Unsupervised][38/200][18:13:54] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.067832918
[GreedyHash Unsupervised][39/200][18:16:08] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068705265
[GreedyHash Unsupervised][40/200][18:18:21] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068581587
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 191.05it/s]
[GreedyHash Unsupervised] epoch:40, bit:16, dataset:cifar10-2, MAP:0.421, Best MAP: 0.421
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][41/200][18:24:08] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.069082240
[GreedyHash Unsupervised][42/200][18:26:24] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068692210
[GreedyHash Unsupervised][43/200][18:28:38] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068323893
[GreedyHash Unsupervised][44/200][18:30:52] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.069435902
[GreedyHash Unsupervised][45/200][18:33:06] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.067835456
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:51<00:00, 195.34it/s]
[GreedyHash Unsupervised] epoch:45, bit:16, dataset:cifar10-2, MAP:0.421, Best MAP: 0.421
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][46/200][18:38:52] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068332585
[GreedyHash Unsupervised][47/200][18:41:07] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068783583
[GreedyHash Unsupervised][48/200][18:43:21] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068441480
[GreedyHash Unsupervised][49/200][18:45:35] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068097797
[GreedyHash Unsupervised][50/200][18:47:49] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068490213
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:42<00:00, 235.29it/s]
[GreedyHash Unsupervised] epoch:50, bit:16, dataset:cifar10-2, MAP:0.424, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][51/200][18:53:26] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068522487
[GreedyHash Unsupervised][52/200][18:55:42] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068207764
[GreedyHash Unsupervised][53/200][18:57:56] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068135931
[GreedyHash Unsupervised][54/200][19:00:09] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068335163
[GreedyHash Unsupervised][55/200][19:02:24] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068260773
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 216.48it/s]
[GreedyHash Unsupervised] epoch:55, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][56/200][19:08:04] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068208010
[GreedyHash Unsupervised][57/200][19:10:20] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.068853707
[GreedyHash Unsupervised][58/200][19:12:34] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.067966615
[GreedyHash Unsupervised][59/200][19:14:48] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.067533841
[GreedyHash Unsupervised][60/200][19:17:02] bit:16, lr:0.000010000, dataset:cifar10-2, train loss:0.067822941
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:43<00:00, 232.40it/s]
[GreedyHash Unsupervised] epoch:60, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][61/200][19:22:39] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068247018
[GreedyHash Unsupervised][62/200][19:24:55] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068330346
[GreedyHash Unsupervised][63/200][19:27:09] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068477684
[GreedyHash Unsupervised][64/200][19:29:24] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068967408
[GreedyHash Unsupervised][65/200][19:31:38] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068012781
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:47<00:00, 212.28it/s]
[GreedyHash Unsupervised] epoch:65, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][66/200][19:37:19] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068274332
[GreedyHash Unsupervised][67/200][19:39:35] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068763141
[GreedyHash Unsupervised][68/200][19:41:49] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068608424
[GreedyHash Unsupervised][69/200][19:44:03] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.069250964
[GreedyHash Unsupervised][70/200][19:46:17] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068508108
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:40<00:00, 246.27it/s]
[GreedyHash Unsupervised] epoch:70, bit:16, dataset:cifar10-2, MAP:0.422, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][71/200][19:51:52] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.067594749
[GreedyHash Unsupervised][72/200][19:54:07] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068524704
[GreedyHash Unsupervised][73/200][19:56:22] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068449215
[GreedyHash Unsupervised][74/200][19:58:36] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068114947
[GreedyHash Unsupervised][75/200][20:00:50] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068340576
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 215.43it/s]
[GreedyHash Unsupervised] epoch:75, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][76/200][20:06:30] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068189252
[GreedyHash Unsupervised][77/200][20:08:46] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068206035
[GreedyHash Unsupervised][78/200][20:11:00] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068516569
[GreedyHash Unsupervised][79/200][20:13:14] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068654019
[GreedyHash Unsupervised][80/200][20:15:29] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068384569
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 218.55it/s]
[GreedyHash Unsupervised] epoch:80, bit:16, dataset:cifar10-2, MAP:0.422, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][81/200][20:21:09] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068563839
[GreedyHash Unsupervised][82/200][20:23:25] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.067778243
[GreedyHash Unsupervised][83/200][20:25:39] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068822997
[GreedyHash Unsupervised][84/200][20:27:53] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.067679787
[GreedyHash Unsupervised][85/200][20:30:08] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068054640
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 217.92it/s]
[GreedyHash Unsupervised] epoch:85, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][86/200][20:35:48] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068755495
[GreedyHash Unsupervised][87/200][20:38:05] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068673159
[GreedyHash Unsupervised][88/200][20:40:19] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068204743
[GreedyHash Unsupervised][89/200][20:42:34] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.067999058
[GreedyHash Unsupervised][90/200][20:44:48] bit:16, lr:0.000001000, dataset:cifar10-2, train loss:0.068513048
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:42<00:00, 233.51it/s]
[GreedyHash Unsupervised] epoch:90, bit:16, dataset:cifar10-2, MAP:0.424, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][91/200][20:50:25] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067915912
[GreedyHash Unsupervised][92/200][20:52:42] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067735995
[GreedyHash Unsupervised][93/200][20:54:56] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068658074
[GreedyHash Unsupervised][94/200][20:57:10] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068343948
[GreedyHash Unsupervised][95/200][20:59:24] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067882208
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 215.93it/s]
[GreedyHash Unsupervised] epoch:95, bit:16, dataset:cifar10-2, MAP:0.424, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][96/200][21:05:04] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068591468
[GreedyHash Unsupervised][97/200][21:07:21] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068827630
[GreedyHash Unsupervised][98/200][21:09:36] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068342773
[GreedyHash Unsupervised][99/200][21:11:50] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068589772
[GreedyHash Unsupervised][100/200][21:14:04] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068755413
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 217.48it/s]
[GreedyHash Unsupervised] epoch:100, bit:16, dataset:cifar10-2, MAP:0.422, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][101/200][21:19:44] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068609942
[GreedyHash Unsupervised][102/200][21:22:00] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.069032875
[GreedyHash Unsupervised][103/200][21:24:14] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067871884
[GreedyHash Unsupervised][104/200][21:26:29] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068308687
[GreedyHash Unsupervised][105/200][21:28:43] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068466161
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 218.15it/s]
[GreedyHash Unsupervised] epoch:105, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][106/200][21:34:23] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068229501
[GreedyHash Unsupervised][107/200][21:36:39] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068070249
[GreedyHash Unsupervised][108/200][21:38:52] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068839744
[GreedyHash Unsupervised][109/200][21:41:06] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067726404
[GreedyHash Unsupervised][110/200][21:43:21] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068487283
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 218.85it/s]
[GreedyHash Unsupervised] epoch:110, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][111/200][21:49:00] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067725275
[GreedyHash Unsupervised][112/200][21:51:16] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068210855
[GreedyHash Unsupervised][113/200][21:53:31] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068473211
[GreedyHash Unsupervised][114/200][21:55:45] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068870538
[GreedyHash Unsupervised][115/200][21:57:58] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068745101
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 216.51it/s]
[GreedyHash Unsupervised] epoch:115, bit:16, dataset:cifar10-2, MAP:0.422, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][116/200][22:03:39] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068130648
[GreedyHash Unsupervised][117/200][22:05:54] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.067687446
[GreedyHash Unsupervised][118/200][22:08:09] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068381593
[GreedyHash Unsupervised][119/200][22:10:23] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.069252968
[GreedyHash Unsupervised][120/200][22:12:36] bit:16, lr:0.000000100, dataset:cifar10-2, train loss:0.068454090
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 217.80it/s]
[GreedyHash Unsupervised] epoch:120, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][121/200][22:18:16] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.067754737
[GreedyHash Unsupervised][122/200][22:20:32] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068431461
[GreedyHash Unsupervised][123/200][22:22:47] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068715992
[GreedyHash Unsupervised][124/200][22:25:01] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068095675
[GreedyHash Unsupervised][125/200][22:27:15] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068372155
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:42<00:00, 233.24it/s]
[GreedyHash Unsupervised] epoch:125, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][126/200][22:32:52] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068977280
[GreedyHash Unsupervised][127/200][22:35:07] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068287498
[GreedyHash Unsupervised][128/200][22:37:22] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.069069975
[GreedyHash Unsupervised][129/200][22:39:36] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068222919
[GreedyHash Unsupervised][130/200][22:41:50] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068120932
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:44<00:00, 223.92it/s]
[GreedyHash Unsupervised] epoch:130, bit:16, dataset:cifar10-2, MAP:0.422, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][131/200][22:47:29] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068622476
[GreedyHash Unsupervised][132/200][22:49:44] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068326924
[GreedyHash Unsupervised][133/200][22:51:59] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.067658676
[GreedyHash Unsupervised][134/200][22:54:13] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068216890
[GreedyHash Unsupervised][135/200][22:56:27] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.067982513
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 215.13it/s]
[GreedyHash Unsupervised] epoch:135, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][136/200][23:02:08] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.067866430
[GreedyHash Unsupervised][137/200][23:04:23] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068729202
[GreedyHash Unsupervised][138/200][23:06:38] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068344468
[GreedyHash Unsupervised][139/200][23:08:53] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068808241
[GreedyHash Unsupervised][140/200][23:11:06] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068362168
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 212.90it/s]
[GreedyHash Unsupervised] epoch:140, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][141/200][23:16:48] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.067980436
[GreedyHash Unsupervised][142/200][23:19:04] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068672417
[GreedyHash Unsupervised][143/200][23:21:20] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068087098
[GreedyHash Unsupervised][144/200][23:23:34] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068525965
[GreedyHash Unsupervised][145/200][23:25:48] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068364377
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:47<00:00, 211.41it/s]
[GreedyHash Unsupervised] epoch:145, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][146/200][23:31:29] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068232640
[GreedyHash Unsupervised][147/200][23:33:45] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068227760
[GreedyHash Unsupervised][148/200][23:36:00] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.067841074
[GreedyHash Unsupervised][149/200][23:38:14] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068781093
[GreedyHash Unsupervised][150/200][23:40:29] bit:16, lr:0.000000010, dataset:cifar10-2, train loss:0.068819650
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:47<00:00, 212.59it/s]
[GreedyHash Unsupervised] epoch:150, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][151/200][23:46:10] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068134912
[GreedyHash Unsupervised][152/200][23:48:24] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068511475
[GreedyHash Unsupervised][153/200][23:50:40] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068667502
[GreedyHash Unsupervised][154/200][23:52:54] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068459364
[GreedyHash Unsupervised][155/200][23:55:08] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.067884858
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 217.41it/s]
[GreedyHash Unsupervised] epoch:155, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][156/200][00:00:48] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068511351
[GreedyHash Unsupervised][157/200][00:03:02] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068099457
[GreedyHash Unsupervised][158/200][00:05:18] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068434633
[GreedyHash Unsupervised][159/200][00:07:32] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.067955661
[GreedyHash Unsupervised][160/200][00:09:46] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068505968
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 217.55it/s]
[GreedyHash Unsupervised] epoch:160, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][161/200][00:15:26] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068767144
[GreedyHash Unsupervised][162/200][00:17:41] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068306874
[GreedyHash Unsupervised][163/200][00:19:57] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068309278
[GreedyHash Unsupervised][164/200][00:22:11] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068852803
[GreedyHash Unsupervised][165/200][00:24:25] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068431030
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:43<00:00, 229.35it/s]
[GreedyHash Unsupervised] epoch:165, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][166/200][00:30:03] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068640526
[GreedyHash Unsupervised][167/200][00:32:18] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.067766831
[GreedyHash Unsupervised][168/200][00:34:33] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068665798
[GreedyHash Unsupervised][169/200][00:36:47] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068723824
[GreedyHash Unsupervised][170/200][00:39:01] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068512879
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 218.57it/s]
[GreedyHash Unsupervised] epoch:170, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][171/200][00:44:42] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068919547
[GreedyHash Unsupervised][172/200][00:46:57] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068150654
[GreedyHash Unsupervised][173/200][00:49:13] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068068548
[GreedyHash Unsupervised][174/200][00:51:27] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068970873
[GreedyHash Unsupervised][175/200][00:53:41] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068554026
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 213.94it/s]
[GreedyHash Unsupervised] epoch:175, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][176/200][00:59:22] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068211066
[GreedyHash Unsupervised][177/200][01:01:37] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068380092
[GreedyHash Unsupervised][178/200][01:03:52] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068609930
[GreedyHash Unsupervised][179/200][01:06:06] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.067485453
[GreedyHash Unsupervised][180/200][01:08:20] bit:16, lr:0.000000001, dataset:cifar10-2, train loss:0.068914622
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 214.02it/s]
[GreedyHash Unsupervised] epoch:180, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][181/200][01:14:01] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068542219
[GreedyHash Unsupervised][182/200][01:16:16] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.069451867
[GreedyHash Unsupervised][183/200][01:18:31] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.067472193
[GreedyHash Unsupervised][184/200][01:20:45] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068591806
[GreedyHash Unsupervised][185/200][01:22:59] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068455448
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 221.44it/s]
[GreedyHash Unsupervised] epoch:185, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][186/200][01:28:39] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068952296
[GreedyHash Unsupervised][187/200][01:30:54] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068350674
[GreedyHash Unsupervised][188/200][01:33:09] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.067960930
[GreedyHash Unsupervised][189/200][01:35:23] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068810971
[GreedyHash Unsupervised][190/200][01:37:38] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068207756
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:45<00:00, 217.59it/s]
[GreedyHash Unsupervised] epoch:190, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][191/200][01:43:18] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068403235
[GreedyHash Unsupervised][192/200][01:45:32] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.069305371
[GreedyHash Unsupervised][193/200][01:47:48] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068805290
[GreedyHash Unsupervised][194/200][01:50:02] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.067952971
[GreedyHash Unsupervised][195/200][01:52:16] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068188658
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:46<00:00, 217.07it/s]
[GreedyHash Unsupervised] epoch:195, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[GreedyHash Unsupervised][196/200][01:57:57] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068316468
[GreedyHash Unsupervised][197/200][02:00:12] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068122626
[GreedyHash Unsupervised][198/200][02:02:27] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068907434
[GreedyHash Unsupervised][199/200][02:04:42] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.067760388
[GreedyHash Unsupervised][200/200][02:06:56] bit:16, lr:0.000000000, dataset:cifar10-2, train loss:0.068603883
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:42<00:00, 235.86it/s]
[GreedyHash Unsupervised] epoch:200, bit:16, dataset:cifar10-2, MAP:0.423, Best MAP: 0.424
{'alpha': 0.1, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[GreedyHash Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.GreedyHashModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=0), 'bit_list': [16], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}

Process finished with exit code 0

{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}}
Files already downloaded and verified
train_dataset 50000
test_dataset 10000
database_dataset 50000
[BiHalf Unsupervised][ 1/200][16:25:01] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.087022262
[BiHalf Unsupervised][ 2/200][16:27:16] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.123172916
[BiHalf Unsupervised][ 3/200][16:29:32] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.131104944
[BiHalf Unsupervised][ 4/200][16:31:47] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.132638569
[BiHalf Unsupervised][ 5/200][16:34:02] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.134509687
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.78it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.79it/s]
100%|████████████████████████████████████| 10000/10000 [01:04<00:00, 154.31it/s]
[BiHalf Unsupervised] epoch:5, bit:64, dataset:cifar10-2, MAP:0.201, Best MAP: 0.201
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][ 6/200][16:40:05] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.136371213
[BiHalf Unsupervised][ 7/200][16:42:19] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.139081755
[BiHalf Unsupervised][ 8/200][16:44:32] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.140467231
[BiHalf Unsupervised][ 9/200][16:46:46] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.144079762
[BiHalf Unsupervised][10/200][16:49:00] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.146111368
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.88it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [01:09<00:00, 142.89it/s]
[BiHalf Unsupervised] epoch:10, bit:64, dataset:cifar10-2, MAP:0.199, Best MAP: 0.201
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][11/200][16:55:03] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.149205325
[BiHalf Unsupervised][12/200][16:57:16] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.150752680
[BiHalf Unsupervised][13/200][16:59:30] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.151674285
[BiHalf Unsupervised][14/200][17:01:44] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.154142388
[BiHalf Unsupervised][15/200][17:03:58] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.154398606
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [01:10<00:00, 142.18it/s]
[BiHalf Unsupervised] epoch:15, bit:64, dataset:cifar10-2, MAP:0.203, Best MAP: 0.203
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][16/200][17:10:02] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.154563008
[BiHalf Unsupervised][17/200][17:12:15] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.155483028
[BiHalf Unsupervised][18/200][17:14:28] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.159410064
[BiHalf Unsupervised][19/200][17:16:42] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.160442888
[BiHalf Unsupervised][20/200][17:18:56] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.160915219
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:12<00:00,  5.89it/s]
100%|████████████████████████████████████| 10000/10000 [01:06<00:00, 149.76it/s]
[BiHalf Unsupervised] epoch:20, bit:64, dataset:cifar10-2, MAP:0.211, Best MAP: 0.211
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][21/200][17:24:56] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.161720591
[BiHalf Unsupervised][22/200][17:27:10] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.162743142
[BiHalf Unsupervised][23/200][17:29:24] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.163320809
[BiHalf Unsupervised][24/200][17:31:37] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.164091209
[BiHalf Unsupervised][25/200][17:33:51] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.165763205
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.87it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [01:10<00:00, 141.80it/s]
[BiHalf Unsupervised] epoch:25, bit:64, dataset:cifar10-2, MAP:0.217, Best MAP: 0.217
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][26/200][17:39:55] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.165477364
[BiHalf Unsupervised][27/200][17:42:09] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.165458770
[BiHalf Unsupervised][28/200][17:44:23] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.166999743
[BiHalf Unsupervised][29/200][17:46:36] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.167017619
[BiHalf Unsupervised][30/200][17:48:50] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.168284995
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [01:08<00:00, 146.28it/s]
[BiHalf Unsupervised] epoch:30, bit:64, dataset:cifar10-2, MAP:0.221, Best MAP: 0.221
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][31/200][17:54:53] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.167746384
[BiHalf Unsupervised][32/200][17:57:06] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.168251880
[BiHalf Unsupervised][33/200][17:59:20] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.168134222
[BiHalf Unsupervised][34/200][18:01:34] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169269744
[BiHalf Unsupervised][35/200][18:03:48] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169616194
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [01:07<00:00, 147.54it/s]
[BiHalf Unsupervised] epoch:35, bit:64, dataset:cifar10-2, MAP:0.222, Best MAP: 0.222
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][36/200][18:09:50] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169836411
[BiHalf Unsupervised][37/200][18:12:04] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.166763194
[BiHalf Unsupervised][38/200][18:14:18] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.167632110
[BiHalf Unsupervised][39/200][18:16:32] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.168248375
[BiHalf Unsupervised][40/200][18:18:46] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169767867
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.84it/s]
100%|████████████████████████████████████| 10000/10000 [01:02<00:00, 159.14it/s]
[BiHalf Unsupervised] epoch:40, bit:64, dataset:cifar10-2, MAP:0.222, Best MAP: 0.222
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][41/200][18:24:43] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169469990
[BiHalf Unsupervised][42/200][18:26:57] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169902897
[BiHalf Unsupervised][43/200][18:29:11] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169574863
[BiHalf Unsupervised][44/200][18:31:25] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.168939093
[BiHalf Unsupervised][45/200][18:33:39] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169985515
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.84it/s]
100%|████████████████████████████████████| 10000/10000 [01:00<00:00, 166.39it/s]
[BiHalf Unsupervised] epoch:45, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][46/200][18:39:34] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.168895377
[BiHalf Unsupervised][47/200][18:41:47] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169201618
[BiHalf Unsupervised][48/200][18:44:01] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169347612
[BiHalf Unsupervised][49/200][18:46:15] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169732978
[BiHalf Unsupervised][50/200][18:48:29] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.170929722
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.81it/s]
100%|████████████████████████████████████| 10000/10000 [00:56<00:00, 178.38it/s]
[BiHalf Unsupervised] epoch:50, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][51/200][18:54:20] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169673784
[BiHalf Unsupervised][52/200][18:56:34] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.171215720
[BiHalf Unsupervised][53/200][18:58:48] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.167951372
[BiHalf Unsupervised][54/200][19:01:01] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.170673698
[BiHalf Unsupervised][55/200][19:03:15] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169259225
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.82it/s]
100%|████████████████████████████████████| 10000/10000 [00:55<00:00, 179.98it/s]
[BiHalf Unsupervised] epoch:55, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][56/200][19:09:05] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.170586847
[BiHalf Unsupervised][57/200][19:11:19] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.169029368
[BiHalf Unsupervised][58/200][19:13:33] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.171738197
[BiHalf Unsupervised][59/200][19:15:47] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.167688646
[BiHalf Unsupervised][60/200][19:18:00] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.170826443
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.82it/s]
100%|████████████████████████████████████| 10000/10000 [00:53<00:00, 187.80it/s]
[BiHalf Unsupervised] epoch:60, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][61/200][19:23:49] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170562721
[BiHalf Unsupervised][62/200][19:26:02] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170385542
[BiHalf Unsupervised][63/200][19:28:16] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169236718
[BiHalf Unsupervised][64/200][19:30:30] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.168630498
[BiHalf Unsupervised][65/200][19:32:43] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170619560
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.87it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.83it/s]
100%|████████████████████████████████████| 10000/10000 [00:54<00:00, 184.20it/s]
[BiHalf Unsupervised] epoch:65, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][66/200][19:38:32] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169722823
[BiHalf Unsupervised][67/200][19:40:46] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170190441
[BiHalf Unsupervised][68/200][19:43:00] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.172224328
[BiHalf Unsupervised][69/200][19:45:14] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.167014810
[BiHalf Unsupervised][70/200][19:47:27] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170543363
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.81it/s]
100%|████████████████████████████████████| 10000/10000 [00:57<00:00, 173.40it/s]
[BiHalf Unsupervised] epoch:70, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][71/200][19:53:20] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.171406943
[BiHalf Unsupervised][72/200][19:55:33] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169873316
[BiHalf Unsupervised][73/200][19:57:48] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.171430457
[BiHalf Unsupervised][74/200][20:00:02] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.173065433
[BiHalf Unsupervised][75/200][20:02:15] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170547824
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.81it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 191.52it/s]
[BiHalf Unsupervised] epoch:75, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][76/200][20:08:03] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.168209178
[BiHalf Unsupervised][77/200][20:10:17] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170289146
[BiHalf Unsupervised][78/200][20:12:31] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170406239
[BiHalf Unsupervised][79/200][20:14:45] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170155211
[BiHalf Unsupervised][80/200][20:16:58] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.166828311
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.83it/s]
100%|████████████████████████████████████| 10000/10000 [00:56<00:00, 176.68it/s]
[BiHalf Unsupervised] epoch:80, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][81/200][20:22:50] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169250150
[BiHalf Unsupervised][82/200][20:25:04] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169797653
[BiHalf Unsupervised][83/200][20:27:18] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.173457040
[BiHalf Unsupervised][84/200][20:29:32] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170883565
[BiHalf Unsupervised][85/200][20:31:46] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170462196
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.82it/s]
100%|████████████████████████████████████| 10000/10000 [00:55<00:00, 179.82it/s]
[BiHalf Unsupervised] epoch:85, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][86/200][20:37:37] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170709428
[BiHalf Unsupervised][87/200][20:39:50] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169278535
[BiHalf Unsupervised][88/200][20:42:04] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.169045246
[BiHalf Unsupervised][89/200][20:44:18] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.168472525
[BiHalf Unsupervised][90/200][20:46:32] bit:64, lr:0.000001000, dataset:cifar10-2, train loss:0.170564332
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.83it/s]
100%|████████████████████████████████████| 10000/10000 [00:55<00:00, 180.26it/s]
[BiHalf Unsupervised] epoch:90, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][91/200][20:52:23] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.169884493
[BiHalf Unsupervised][92/200][20:54:36] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171058214
[BiHalf Unsupervised][93/200][20:56:50] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171922961
[BiHalf Unsupervised][94/200][20:59:04] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171777826
[BiHalf Unsupervised][95/200][21:01:18] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170402915
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.87it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.83it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 189.45it/s]
[BiHalf Unsupervised] epoch:95, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][96/200][21:07:06] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.169250053
[BiHalf Unsupervised][97/200][21:09:19] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171062356
[BiHalf Unsupervised][98/200][21:11:33] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.169781687
[BiHalf Unsupervised][99/200][21:13:47] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171150815
[BiHalf Unsupervised][100/200][21:16:00] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.172658488
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.83it/s]
100%|████████████████████████████████████| 10000/10000 [00:51<00:00, 195.45it/s]
[BiHalf Unsupervised] epoch:100, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][101/200][21:21:46] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170743617
[BiHalf Unsupervised][102/200][21:24:00] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171219061
[BiHalf Unsupervised][103/200][21:26:14] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171010181
[BiHalf Unsupervised][104/200][21:28:28] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170615857
[BiHalf Unsupervised][105/200][21:30:42] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171040586
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.82it/s]
100%|████████████████████████████████████| 10000/10000 [00:57<00:00, 174.39it/s]
[BiHalf Unsupervised] epoch:105, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][106/200][21:36:34] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171469833
[BiHalf Unsupervised][107/200][21:38:48] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.168580705
[BiHalf Unsupervised][108/200][21:41:01] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170766128
[BiHalf Unsupervised][109/200][21:43:15] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170613928
[BiHalf Unsupervised][110/200][21:45:29] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.168248437
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.82it/s]
100%|████████████████████████████████████| 10000/10000 [00:56<00:00, 177.87it/s]
[BiHalf Unsupervised] epoch:110, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][111/200][21:51:20] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.168441541
[BiHalf Unsupervised][112/200][21:53:34] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.169433777
[BiHalf Unsupervised][113/200][21:55:47] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170222484
[BiHalf Unsupervised][114/200][21:58:01] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171178292
[BiHalf Unsupervised][115/200][22:00:15] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170881041
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.83it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.82it/s]
100%|████████████████████████████████████| 10000/10000 [00:53<00:00, 187.76it/s]
[BiHalf Unsupervised] epoch:115, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][116/200][22:06:03] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170970386
[BiHalf Unsupervised][117/200][22:08:17] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.169747570
[BiHalf Unsupervised][118/200][22:10:31] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.169299679
[BiHalf Unsupervised][119/200][22:12:45] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.171258544
[BiHalf Unsupervised][120/200][22:14:58] bit:64, lr:0.000000100, dataset:cifar10-2, train loss:0.170562382
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.82it/s]
100%|█████████████████████████████████████████| 782/782 [02:14<00:00,  5.83it/s]
100%|████████████████████████████████████| 10000/10000 [00:56<00:00, 177.54it/s]
[BiHalf Unsupervised] epoch:120, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][121/200][22:20:50] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.171211918
[BiHalf Unsupervised][122/200][22:23:03] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170325135
[BiHalf Unsupervised][123/200][22:25:17] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170265479
[BiHalf Unsupervised][124/200][22:27:31] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170995128
[BiHalf Unsupervised][125/200][22:29:45] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.171254838
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.75it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.85it/s]
100%|████████████████████████████████████| 10000/10000 [00:57<00:00, 175.25it/s]
[BiHalf Unsupervised] epoch:125, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][126/200][22:35:37] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169053971
[BiHalf Unsupervised][127/200][22:37:50] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169592892
[BiHalf Unsupervised][128/200][22:40:04] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.168909883
[BiHalf Unsupervised][129/200][22:42:18] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.168164616
[BiHalf Unsupervised][130/200][22:44:31] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169030770
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.68it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:54<00:00, 183.74it/s]
[BiHalf Unsupervised] epoch:130, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][131/200][22:50:21] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170680748
[BiHalf Unsupervised][132/200][22:52:34] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170800434
[BiHalf Unsupervised][133/200][22:54:48] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169565057
[BiHalf Unsupervised][134/200][22:57:02] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169288232
[BiHalf Unsupervised][135/200][22:59:16] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169131010
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.67it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:54<00:00, 183.48it/s]
[BiHalf Unsupervised] epoch:135, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][136/200][23:05:06] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169884843
[BiHalf Unsupervised][137/200][23:07:19] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169738742
[BiHalf Unsupervised][138/200][23:09:33] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169849061
[BiHalf Unsupervised][139/200][23:11:47] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169310633
[BiHalf Unsupervised][140/200][23:14:01] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169154467
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.70it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:57<00:00, 174.62it/s]
[BiHalf Unsupervised] epoch:140, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][141/200][23:19:53] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.168446470
[BiHalf Unsupervised][142/200][23:22:07] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170781354
[BiHalf Unsupervised][143/200][23:24:21] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.168881113
[BiHalf Unsupervised][144/200][23:26:35] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170399028
[BiHalf Unsupervised][145/200][23:28:48] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169059005
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.68it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:55<00:00, 181.54it/s]
[BiHalf Unsupervised] epoch:145, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][146/200][23:34:39] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.170089036
[BiHalf Unsupervised][147/200][23:36:52] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.172834086
[BiHalf Unsupervised][148/200][23:39:06] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.169595950
[BiHalf Unsupervised][149/200][23:41:20] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.168953067
[BiHalf Unsupervised][150/200][23:43:34] bit:64, lr:0.000000010, dataset:cifar10-2, train loss:0.171648875
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.72it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:54<00:00, 184.07it/s]
[BiHalf Unsupervised] epoch:150, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][151/200][23:49:23] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.172007499
[BiHalf Unsupervised][152/200][23:51:37] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.172450074
[BiHalf Unsupervised][153/200][23:53:51] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170766164
[BiHalf Unsupervised][154/200][23:56:05] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.171789046
[BiHalf Unsupervised][155/200][23:58:19] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170894051
100%|█████████████████████████████████████████| 157/157 [00:27<00:00,  5.75it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:58<00:00, 172.06it/s]
[BiHalf Unsupervised] epoch:155, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][156/200][00:04:12] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170906443
[BiHalf Unsupervised][157/200][00:06:25] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170921951
[BiHalf Unsupervised][158/200][00:08:39] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169772861
[BiHalf Unsupervised][159/200][00:10:53] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169374790
[BiHalf Unsupervised][160/200][00:13:07] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.172444113
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.82it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:54<00:00, 182.14it/s]
[BiHalf Unsupervised] epoch:160, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][161/200][00:18:58] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169487163
[BiHalf Unsupervised][162/200][00:21:11] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.168562295
[BiHalf Unsupervised][163/200][00:23:25] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170850489
[BiHalf Unsupervised][164/200][00:25:39] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169149568
[BiHalf Unsupervised][165/200][00:27:52] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.167847021
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 190.61it/s]
[BiHalf Unsupervised] epoch:165, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][166/200][00:33:40] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170314863
[BiHalf Unsupervised][167/200][00:35:53] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.168945128
[BiHalf Unsupervised][168/200][00:38:07] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170518310
[BiHalf Unsupervised][169/200][00:40:21] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170235496
[BiHalf Unsupervised][170/200][00:42:34] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169475630
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:53<00:00, 185.76it/s]
[BiHalf Unsupervised] epoch:170, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][171/200][00:48:23] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169451495
[BiHalf Unsupervised][172/200][00:50:36] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170128793
[BiHalf Unsupervised][173/200][00:52:50] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169598758
[BiHalf Unsupervised][174/200][00:55:04] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170031609
[BiHalf Unsupervised][175/200][00:57:17] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170663037
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:55<00:00, 178.69it/s]
[BiHalf Unsupervised] epoch:175, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][176/200][01:03:09] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.169691838
[BiHalf Unsupervised][177/200][01:05:23] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.170641282
[BiHalf Unsupervised][178/200][01:07:36] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.172957161
[BiHalf Unsupervised][179/200][01:09:50] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.172703804
[BiHalf Unsupervised][180/200][01:12:04] bit:64, lr:0.000000001, dataset:cifar10-2, train loss:0.168005221
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.87it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:53<00:00, 188.17it/s]
[BiHalf Unsupervised] epoch:180, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][181/200][01:17:52] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169126885
[BiHalf Unsupervised][182/200][01:20:05] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.168925487
[BiHalf Unsupervised][183/200][01:22:19] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.170437752
[BiHalf Unsupervised][184/200][01:24:33] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169979999
[BiHalf Unsupervised][185/200][01:26:47] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.168602609
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.84it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:54<00:00, 182.92it/s]
[BiHalf Unsupervised] epoch:185, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][186/200][01:32:37] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171825023
[BiHalf Unsupervised][187/200][01:34:50] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169344229
[BiHalf Unsupervised][188/200][01:37:04] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169312734
[BiHalf Unsupervised][189/200][01:39:18] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.168923802
[BiHalf Unsupervised][190/200][01:41:32] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.170517124
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:53<00:00, 186.06it/s]
[BiHalf Unsupervised] epoch:190, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][191/200][01:47:21] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171319272
[BiHalf Unsupervised][192/200][01:49:35] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169869841
[BiHalf Unsupervised][193/200][01:51:48] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.172836766
[BiHalf Unsupervised][194/200][01:54:02] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.172774910
[BiHalf Unsupervised][195/200][01:56:17] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171555375
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.85it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.87it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 190.31it/s]
[BiHalf Unsupervised] epoch:195, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][196/200][02:02:04] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169292366
[BiHalf Unsupervised][197/200][02:04:18] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171099631
[BiHalf Unsupervised][198/200][02:06:32] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.169411861
[BiHalf Unsupervised][199/200][02:08:46] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.172826315
[BiHalf Unsupervised][200/200][02:10:59] bit:64, lr:0.000000000, dataset:cifar10-2, train loss:0.171521300
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.88it/s]
100%|████████████████████████████████████| 10000/10000 [00:36<00:00, 270.74it/s]
[BiHalf Unsupervised] epoch:200, bit:64, dataset:cifar10-2, MAP:0.223, Best MAP: 0.223
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}

Process finished with exit code 0
Tramac commented 3 years ago

谢谢，看起来train loss从0.086左右降到了0.067左右，幅度确实不算大

Tramac commented 3 years ago

话说方便加个微信交流一下嘛，正好最近也在看一些deephash的方法

swuxyj commented 3 years ago

话说方便加个微信交流一下嘛，正好最近也在看一些deephash的方法

可以的，微信号： PyTorch_Coder

swuxyj commented 3 years ago

抱歉，由于之前没仔细看文章，复现出来的BiHalf算法MAP较低，现在经过原文作者提醒，已做了更正，这是最新的日志：

{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}}
Files already downloaded and verified
train_dataset 50000
test_dataset 10000
database_dataset 50000
[BiHalf Unsupervised][ 1/200][09:00:47] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.083776357
[BiHalf Unsupervised][ 2/200][09:05:23] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.121161771
[BiHalf Unsupervised][ 3/200][09:10:00] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.128708757
[BiHalf Unsupervised][ 4/200][09:14:37] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.131950624
[BiHalf Unsupervised][ 5/200][09:19:13] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.132506133
100%|█████████████████████████████████████████| 157/157 [00:55<00:00,  2.83it/s]
100%|█████████████████████████████████████████| 782/782 [03:46<00:00,  3.46it/s]
100%|████████████████████████████████████| 10000/10000 [00:52<00:00, 192.07it/s]
[BiHalf Unsupervised] epoch:5, bit:64, dataset:cifar10-2, MAP:0.548, Best MAP: 0.548
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][ 6/200][09:29:26] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.136071196
[BiHalf Unsupervised][ 7/200][09:32:50] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.137810745
[BiHalf Unsupervised][ 8/200][09:35:04] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.138409827
[BiHalf Unsupervised][ 9/200][09:37:18] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.143489986
[BiHalf Unsupervised][10/200][09:39:32] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.146055311
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.86it/s]
100%|█████████████████████████████████████████| 782/782 [02:13<00:00,  5.86it/s]
100%|████████████████████████████████████| 10000/10000 [00:33<00:00, 297.19it/s]
[BiHalf Unsupervised] epoch:10, bit:64, dataset:cifar10-2, MAP:0.564, Best MAP: 0.564
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][11/200][09:45:00] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.149155811
[BiHalf Unsupervised][12/200][09:47:12] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.152094328
[BiHalf Unsupervised][13/200][09:49:24] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.152906044
[BiHalf Unsupervised][14/200][09:51:36] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.154735048
[BiHalf Unsupervised][15/200][09:53:49] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.156524857
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.96it/s]
100%|█████████████████████████████████████████| 782/782 [02:11<00:00,  5.96it/s]
100%|████████████████████████████████████| 10000/10000 [00:35<00:00, 281.11it/s]
[BiHalf Unsupervised] epoch:15, bit:64, dataset:cifar10-2, MAP:0.575, Best MAP: 0.575
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][16/200][09:59:14] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.158786986
[BiHalf Unsupervised][17/200][10:01:26] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.160365501
[BiHalf Unsupervised][18/200][10:03:38] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.163068954
[BiHalf Unsupervised][19/200][10:05:51] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.164266397
[BiHalf Unsupervised][20/200][10:08:03] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.166818093
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.94it/s]
100%|█████████████████████████████████████████| 782/782 [02:11<00:00,  5.95it/s]
100%|████████████████████████████████████| 10000/10000 [00:33<00:00, 299.00it/s]
[BiHalf Unsupervised] epoch:20, bit:64, dataset:cifar10-2, MAP:0.584, Best MAP: 0.584
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][21/200][10:13:27] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.167804908
[BiHalf Unsupervised][22/200][10:15:39] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.168449978
[BiHalf Unsupervised][23/200][10:17:51] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.169558835
[BiHalf Unsupervised][24/200][10:20:04] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.168348478
[BiHalf Unsupervised][25/200][10:22:16] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.171551456
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.95it/s]
100%|█████████████████████████████████████████| 782/782 [02:11<00:00,  5.96it/s]
100%|████████████████████████████████████| 10000/10000 [00:34<00:00, 287.76it/s]
[BiHalf Unsupervised] epoch:25, bit:64, dataset:cifar10-2, MAP:0.588, Best MAP: 0.588
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][26/200][10:27:41] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.170861172
[BiHalf Unsupervised][27/200][10:29:53] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.173758023
[BiHalf Unsupervised][28/200][10:32:05] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.174744619
[BiHalf Unsupervised][29/200][10:34:18] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.173901173
[BiHalf Unsupervised][30/200][10:36:30] bit:64, lr:0.000100000, dataset:cifar10-2, train loss:0.174011666
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.96it/s]
100%|█████████████████████████████████████████| 782/782 [02:11<00:00,  5.96it/s]
100%|████████████████████████████████████| 10000/10000 [00:34<00:00, 288.43it/s]
[BiHalf Unsupervised] epoch:30, bit:64, dataset:cifar10-2, MAP:0.591, Best MAP: 0.591
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][31/200][10:41:55] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.174081806
[BiHalf Unsupervised][32/200][10:44:07] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.175442401
[BiHalf Unsupervised][33/200][10:46:19] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.175272161
[BiHalf Unsupervised][34/200][10:48:32] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.173932833
[BiHalf Unsupervised][35/200][10:50:44] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.175092362
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.93it/s]
100%|█████████████████████████████████████████| 782/782 [02:11<00:00,  5.94it/s]
100%|████████████████████████████████████| 10000/10000 [00:32<00:00, 305.12it/s]
[BiHalf Unsupervised] epoch:35, bit:64, dataset:cifar10-2, MAP:0.592, Best MAP: 0.592
{'gamma': 6, 'optimizer': {'type': <class 'torch.optim.sgd.SGD'>, 'epoch_lr_decrease': 30, 'optim_params': {'lr': 0.0001, 'weight_decay': 0.0005, 'momentum': 0.9}}, 'info': '[BiHalf Unsupervised]', 'resize_size': 256, 'crop_size': 224, 'batch_size': 64, 'net': <class '__main__.BiHalfModelUnsupervised'>, 'dataset': 'cifar10-2', 'epoch': 200, 'test_map': 5, 'device': device(type='cuda', index=1), 'bit_list': [64], 'topK': 1000, 'n_class': 10, 'data_path': '/dataset/cifar10-2/', 'data': {'train_set': {'list_path': './data/cifar10-2/train.txt', 'batch_size': 64}, 'database': {'list_path': './data/cifar10-2/database.txt', 'batch_size': 64}, 'test': {'list_path': './data/cifar10-2/test.txt', 'batch_size': 64}}, 'num_train': 50000}
[BiHalf Unsupervised][36/200][10:56:08] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.174162896
[BiHalf Unsupervised][37/200][10:58:20] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.175131623
[BiHalf Unsupervised][38/200][11:00:33] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.176157677
[BiHalf Unsupervised][39/200][11:02:45] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.177014620
[BiHalf Unsupervised][40/200][11:04:58] bit:64, lr:0.000010000, dataset:cifar10-2, train loss:0.176051271
100%|█████████████████████████████████████████| 157/157 [00:26<00:00,  5.93it/s]
100%|█████████████████████████████████████████| 782/782 [02:11<00:00,  5.95it/s]
100%|████████████████████████████████████| 10000/10000 [00:34<00:00, 292.30it/s]
[BiHalf Unsupervised] epoch:40, bit:64, dataset:cifar10-2, MAP:0.593, Best MAP: 0.593