==========
Args:Namespace(dataset='msmt17', batch_size=256, workers=4, height=256, width=128, num_instances=8, eps=0.7, eps_gap=0.02, k1=30, k2=6, arch='resnet_ibn50a', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet50_IBN_Market1501/64bs_lr0.0004_ep120_warm20_seed0/resnet50_ibn_120.pth', features=0, dropout=0, momentum=0.2, drop_path_rate=0.3, hw_ratio=2, self_norm=True, conv_stem=False, lr=0.00035, weight_decay=0.0005, optimizer='SGD', epochs=50, iters=200, step_size=20, seed=1, print_freq=20, eval_step=10, temp=0.05, data_dir='/home/ma-user/work/Projects/ReIDNet_Finetune/CContrast/data', logs_dir='log/market2msmt/resnet50_ibn_cion', pooling_type='gem', use_hard=True)
==========
==> Load unlabeled dataset
=> MSMT17 loaded
Dataset statistics:
  ----------------------------------------
  subset   | # ids | # images | # cameras
  ----------------------------------------
  train    |  1041 |    32621 |        15
  query    |  3060 |    11659 |        15
  gallery  |  3060 |    82161 |        15
  ----------------------------------------
pooling_type: gem
checkpoint loaded!
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.529 (0.367)	Data 0.408 (0.077)	
Extract Features: [100/128]	Time 0.117 (0.271)	Data 0.000 (0.066)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.276915311813354
Clustering criterion: eps: 0.700
==> Statistics for epoch 0: 957 clusters
Epoch: [0][20/200]	Time 0.365 (0.766)	Data 0.001 (0.053)	Loss 3.002 (3.009)
Epoch: [0][40/200]	Time 0.364 (0.600)	Data 0.000 (0.061)	Loss 2.283 (2.968)
Epoch: [0][60/200]	Time 0.362 (0.545)	Data 0.001 (0.064)	Loss 1.646 (2.841)
Epoch: [0][80/200]	Time 0.367 (0.500)	Data 0.002 (0.048)	Loss 2.406 (2.664)
Epoch: [0][100/200]	Time 0.371 (0.490)	Data 0.001 (0.055)	Loss 2.263 (2.536)
Epoch: [0][120/200]	Time 0.365 (0.482)	Data 0.001 (0.057)	Loss 1.566 (2.438)
Epoch: [0][140/200]	Time 0.362 (0.466)	Data 0.000 (0.049)	Loss 2.312 (2.363)
Epoch: [0][160/200]	Time 0.365 (0.463)	Data 0.001 (0.052)	Loss 2.231 (2.299)
Epoch: [0][180/200]	Time 0.365 (0.460)	Data 0.001 (0.055)	Loss 2.147 (2.242)
Epoch: [0][200/200]	Time 0.367 (0.451)	Data 0.000 (0.049)	Loss 1.548 (2.205)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.209)	Data 0.000 (0.089)	
Extract Features: [100/128]	Time 0.118 (0.188)	Data 0.000 (0.068)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.51851224899292
==> Statistics for epoch 1: 1003 clusters
Epoch: [1][20/200]	Time 0.366 (0.416)	Data 0.001 (0.049)	Loss 0.459 (0.433)
Epoch: [1][40/200]	Time 0.363 (0.432)	Data 0.001 (0.065)	Loss 1.559 (0.735)
Epoch: [1][60/200]	Time 0.364 (0.411)	Data 0.000 (0.044)	Loss 1.594 (1.103)
Epoch: [1][80/200]	Time 0.366 (0.418)	Data 0.001 (0.051)	Loss 2.310 (1.293)
Epoch: [1][100/200]	Time 0.372 (0.425)	Data 0.001 (0.058)	Loss 1.663 (1.396)
Epoch: [1][120/200]	Time 0.370 (0.415)	Data 0.000 (0.048)	Loss 1.660 (1.464)
Epoch: [1][140/200]	Time 0.370 (0.420)	Data 0.001 (0.053)	Loss 1.577 (1.519)
Epoch: [1][160/200]	Time 0.373 (0.424)	Data 0.001 (0.057)	Loss 1.923 (1.550)
Epoch: [1][180/200]	Time 0.368 (0.418)	Data 0.000 (0.050)	Loss 1.814 (1.576)
Epoch: [1][200/200]	Time 0.372 (0.421)	Data 0.000 (0.054)	Loss 1.767 (1.588)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.283 (0.210)	Data 0.157 (0.088)	
Extract Features: [100/128]	Time 0.118 (0.190)	Data 0.000 (0.069)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.737791776657104
==> Statistics for epoch 2: 1009 clusters
Epoch: [2][20/200]	Time 0.363 (0.421)	Data 0.001 (0.056)	Loss 0.312 (0.435)
Epoch: [2][40/200]	Time 0.366 (0.433)	Data 0.001 (0.069)	Loss 2.291 (0.743)
Epoch: [2][60/200]	Time 0.367 (0.413)	Data 0.000 (0.046)	Loss 2.676 (1.096)
Epoch: [2][80/200]	Time 0.365 (0.422)	Data 0.001 (0.056)	Loss 1.754 (1.301)
Epoch: [2][100/200]	Time 0.378 (0.426)	Data 0.001 (0.060)	Loss 1.747 (1.424)
Epoch: [2][120/200]	Time 0.364 (0.418)	Data 0.000 (0.050)	Loss 1.630 (1.499)
Epoch: [2][140/200]	Time 0.366 (0.422)	Data 0.001 (0.054)	Loss 1.842 (1.551)
Epoch: [2][160/200]	Time 0.366 (0.426)	Data 0.001 (0.058)	Loss 1.567 (1.580)
Epoch: [2][180/200]	Time 0.361 (0.419)	Data 0.000 (0.052)	Loss 1.494 (1.602)
Epoch: [2][200/200]	Time 0.366 (0.422)	Data 0.001 (0.055)	Loss 1.836 (1.617)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.192 (0.208)	Data 0.080 (0.086)	
Extract Features: [100/128]	Time 0.243 (0.193)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.96112895011902
==> Statistics for epoch 3: 996 clusters
Epoch: [3][20/200]	Time 0.368 (0.423)	Data 0.001 (0.054)	Loss 0.456 (0.416)
Epoch: [3][40/200]	Time 0.361 (0.434)	Data 0.001 (0.068)	Loss 1.411 (0.702)
Epoch: [3][60/200]	Time 0.363 (0.411)	Data 0.000 (0.045)	Loss 1.397 (1.089)
Epoch: [3][80/200]	Time 0.364 (0.421)	Data 0.001 (0.054)	Loss 1.749 (1.273)
Epoch: [3][100/200]	Time 0.365 (0.426)	Data 0.001 (0.060)	Loss 2.014 (1.400)
Epoch: [3][120/200]	Time 0.365 (0.416)	Data 0.000 (0.050)	Loss 1.976 (1.459)
Epoch: [3][140/200]	Time 0.365 (0.419)	Data 0.001 (0.053)	Loss 1.721 (1.518)
Epoch: [3][160/200]	Time 0.368 (0.423)	Data 0.001 (0.056)	Loss 1.616 (1.555)
Epoch: [3][180/200]	Time 0.366 (0.417)	Data 0.000 (0.050)	Loss 1.876 (1.585)
Epoch: [3][200/200]	Time 0.366 (0.420)	Data 0.001 (0.053)	Loss 1.347 (1.608)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.211)	Data 0.000 (0.090)	
Extract Features: [100/128]	Time 0.118 (0.190)	Data 0.000 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.472654581069946
==> Statistics for epoch 4: 995 clusters
Epoch: [4][20/200]	Time 0.363 (0.422)	Data 0.001 (0.057)	Loss 0.275 (0.401)
Epoch: [4][40/200]	Time 0.366 (0.432)	Data 0.001 (0.069)	Loss 1.779 (0.689)
Epoch: [4][60/200]	Time 0.362 (0.410)	Data 0.000 (0.046)	Loss 1.879 (1.073)
Epoch: [4][80/200]	Time 0.369 (0.418)	Data 0.003 (0.054)	Loss 2.002 (1.276)
Epoch: [4][100/200]	Time 0.365 (0.424)	Data 0.001 (0.059)	Loss 1.885 (1.368)
Epoch: [4][120/200]	Time 0.364 (0.414)	Data 0.000 (0.049)	Loss 1.792 (1.433)
Epoch: [4][140/200]	Time 0.369 (0.420)	Data 0.001 (0.053)	Loss 1.952 (1.488)
Epoch: [4][160/200]	Time 0.369 (0.424)	Data 0.001 (0.057)	Loss 1.367 (1.518)
Epoch: [4][180/200]	Time 0.364 (0.417)	Data 0.000 (0.051)	Loss 1.932 (1.551)
Epoch: [4][200/200]	Time 0.367 (0.420)	Data 0.001 (0.054)	Loss 1.649 (1.570)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.208 (0.211)	Data 0.090 (0.088)	
Extract Features: [100/128]	Time 0.118 (0.193)	Data 0.000 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.68784856796265
==> Statistics for epoch 5: 993 clusters
Epoch: [5][20/200]	Time 0.365 (0.421)	Data 0.001 (0.056)	Loss 0.371 (0.369)
Epoch: [5][40/200]	Time 0.363 (0.433)	Data 0.001 (0.069)	Loss 1.843 (0.695)
Epoch: [5][60/200]	Time 0.368 (0.410)	Data 0.000 (0.046)	Loss 1.758 (1.066)
Epoch: [5][80/200]	Time 0.367 (0.422)	Data 0.001 (0.056)	Loss 1.340 (1.239)
Epoch: [5][100/200]	Time 0.366 (0.429)	Data 0.001 (0.063)	Loss 2.003 (1.324)
Epoch: [5][120/200]	Time 0.366 (0.419)	Data 0.000 (0.052)	Loss 2.008 (1.402)
Epoch: [5][140/200]	Time 0.362 (0.424)	Data 0.001 (0.057)	Loss 2.165 (1.458)
Epoch: [5][160/200]	Time 0.369 (0.427)	Data 0.001 (0.060)	Loss 1.920 (1.500)
Epoch: [5][180/200]	Time 0.365 (0.420)	Data 0.000 (0.054)	Loss 1.936 (1.519)
Epoch: [5][200/200]	Time 0.372 (0.423)	Data 0.001 (0.057)	Loss 1.740 (1.543)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.119 (0.213)	Data 0.000 (0.091)	
Extract Features: [100/128]	Time 0.118 (0.192)	Data 0.000 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.88687229156494
==> Statistics for epoch 6: 1021 clusters
Epoch: [6][20/200]	Time 0.364 (0.415)	Data 0.001 (0.050)	Loss 0.296 (0.373)
Epoch: [6][40/200]	Time 0.365 (0.433)	Data 0.001 (0.067)	Loss 1.693 (0.636)
Epoch: [6][60/200]	Time 0.362 (0.410)	Data 0.000 (0.045)	Loss 1.839 (1.029)
Epoch: [6][80/200]	Time 0.365 (0.422)	Data 0.001 (0.055)	Loss 2.077 (1.239)
Epoch: [6][100/200]	Time 0.364 (0.429)	Data 0.001 (0.062)	Loss 1.713 (1.331)
Epoch: [6][120/200]	Time 0.364 (0.418)	Data 0.000 (0.052)	Loss 1.444 (1.408)
Epoch: [6][140/200]	Time 0.366 (0.424)	Data 0.001 (0.056)	Loss 1.627 (1.453)
Epoch: [6][160/200]	Time 0.367 (0.427)	Data 0.001 (0.060)	Loss 1.483 (1.494)
Epoch: [6][180/200]	Time 0.365 (0.421)	Data 0.000 (0.053)	Loss 2.035 (1.526)
Epoch: [6][200/200]	Time 0.370 (0.424)	Data 0.001 (0.057)	Loss 1.106 (1.540)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.215 (0.216)	Data 0.097 (0.097)	
Extract Features: [100/128]	Time 0.118 (0.194)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.537171840667725
==> Statistics for epoch 7: 1013 clusters
Epoch: [7][20/200]	Time 0.364 (0.417)	Data 0.001 (0.051)	Loss 0.265 (0.313)
Epoch: [7][40/200]	Time 0.366 (0.434)	Data 0.001 (0.069)	Loss 1.545 (0.615)
Epoch: [7][60/200]	Time 0.362 (0.411)	Data 0.000 (0.046)	Loss 1.522 (0.982)
Epoch: [7][80/200]	Time 0.364 (0.421)	Data 0.001 (0.057)	Loss 1.752 (1.154)
Epoch: [7][100/200]	Time 0.363 (0.427)	Data 0.001 (0.062)	Loss 1.538 (1.262)
Epoch: [7][120/200]	Time 0.367 (0.417)	Data 0.000 (0.052)	Loss 1.579 (1.345)
Epoch: [7][140/200]	Time 0.365 (0.423)	Data 0.001 (0.057)	Loss 1.707 (1.381)
Epoch: [7][160/200]	Time 0.364 (0.427)	Data 0.001 (0.060)	Loss 1.416 (1.415)
Epoch: [7][180/200]	Time 0.370 (0.420)	Data 0.000 (0.054)	Loss 1.800 (1.445)
Epoch: [7][200/200]	Time 0.366 (0.423)	Data 0.001 (0.056)	Loss 1.784 (1.462)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.119 (0.213)	Data 0.000 (0.091)	
Extract Features: [100/128]	Time 0.122 (0.194)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.53044247627258
==> Statistics for epoch 8: 1017 clusters
Epoch: [8][20/200]	Time 0.362 (0.417)	Data 0.001 (0.052)	Loss 0.359 (0.317)
Epoch: [8][40/200]	Time 0.360 (0.432)	Data 0.001 (0.067)	Loss 1.650 (0.613)
Epoch: [8][60/200]	Time 0.360 (0.411)	Data 0.000 (0.045)	Loss 1.618 (0.958)
Epoch: [8][80/200]	Time 0.363 (0.422)	Data 0.001 (0.056)	Loss 1.483 (1.137)
Epoch: [8][100/200]	Time 0.360 (0.428)	Data 0.000 (0.062)	Loss 1.737 (1.251)
Epoch: [8][120/200]	Time 0.366 (0.418)	Data 0.000 (0.052)	Loss 1.795 (1.319)
Epoch: [8][140/200]	Time 0.363 (0.422)	Data 0.001 (0.057)	Loss 1.593 (1.361)
Epoch: [8][160/200]	Time 0.374 (0.426)	Data 0.001 (0.060)	Loss 1.705 (1.402)
Epoch: [8][180/200]	Time 0.366 (0.419)	Data 0.000 (0.054)	Loss 2.013 (1.427)
Epoch: [8][200/200]	Time 0.365 (0.422)	Data 0.001 (0.056)	Loss 1.522 (1.449)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.119 (0.212)	Data 0.000 (0.091)	
Extract Features: [100/128]	Time 0.117 (0.192)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.27349829673767
==> Statistics for epoch 9: 1028 clusters
Epoch: [9][20/200]	Time 0.366 (0.422)	Data 0.001 (0.057)	Loss 0.263 (0.311)
Epoch: [9][40/200]	Time 0.373 (0.436)	Data 0.001 (0.071)	Loss 1.919 (0.566)
Epoch: [9][60/200]	Time 0.363 (0.413)	Data 0.000 (0.048)	Loss 1.697 (0.918)
Epoch: [9][80/200]	Time 0.373 (0.425)	Data 0.001 (0.058)	Loss 1.962 (1.104)
Epoch: [9][100/200]	Time 0.366 (0.432)	Data 0.001 (0.063)	Loss 1.926 (1.212)
Epoch: [9][120/200]	Time 0.362 (0.421)	Data 0.001 (0.053)	Loss 1.489 (1.282)
Epoch: [9][140/200]	Time 0.366 (0.426)	Data 0.001 (0.058)	Loss 1.228 (1.317)
Epoch: [9][160/200]	Time 0.365 (0.419)	Data 0.000 (0.051)	Loss 1.917 (1.358)
Epoch: [9][180/200]	Time 0.367 (0.424)	Data 0.001 (0.056)	Loss 1.407 (1.387)
Epoch: [9][200/200]	Time 0.367 (0.427)	Data 0.001 (0.059)	Loss 1.382 (1.406)
Extract Features: [50/367]	Time 0.169 (0.209)	Data 0.051 (0.088)	
Extract Features: [100/367]	Time 0.116 (0.200)	Data 0.000 (0.078)	
Extract Features: [150/367]	Time 0.117 (0.196)	Data 0.000 (0.074)	
Extract Features: [200/367]	Time 0.119 (0.191)	Data 0.000 (0.070)	
Extract Features: [250/367]	Time 0.119 (0.189)	Data 0.000 (0.068)	
Extract Features: [300/367]	Time 0.120 (0.187)	Data 0.005 (0.066)	
Extract Features: [350/367]	Time 0.122 (0.186)	Data 0.000 (0.065)	
Mean AP: 54.2%

 * Finished epoch   9  model mAP: 54.2%  best: 54.2% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.161 (0.215)	Data 0.043 (0.097)	
Extract Features: [100/128]	Time 0.120 (0.195)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.48395872116089
==> Statistics for epoch 10: 1023 clusters
Epoch: [10][20/200]	Time 0.367 (0.432)	Data 0.001 (0.050)	Loss 0.176 (0.312)
Epoch: [10][40/200]	Time 0.363 (0.446)	Data 0.001 (0.069)	Loss 1.582 (0.586)
Epoch: [10][60/200]	Time 0.359 (0.418)	Data 0.000 (0.046)	Loss 1.838 (0.916)
Epoch: [10][80/200]	Time 0.363 (0.429)	Data 0.001 (0.058)	Loss 1.694 (1.104)
Epoch: [10][100/200]	Time 0.363 (0.432)	Data 0.001 (0.063)	Loss 1.512 (1.201)
Epoch: [10][120/200]	Time 0.363 (0.421)	Data 0.000 (0.053)	Loss 1.630 (1.274)
Epoch: [10][140/200]	Time 0.363 (0.427)	Data 0.000 (0.058)	Loss 1.721 (1.323)
Epoch: [10][160/200]	Time 0.372 (0.430)	Data 0.001 (0.062)	Loss 1.477 (1.371)
Epoch: [10][180/200]	Time 0.362 (0.423)	Data 0.000 (0.055)	Loss 1.748 (1.407)
Epoch: [10][200/200]	Time 0.365 (0.426)	Data 0.001 (0.059)	Loss 2.012 (1.427)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.207)	Data 0.000 (0.086)	
Extract Features: [100/128]	Time 0.120 (0.192)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.73069095611572
==> Statistics for epoch 11: 1033 clusters
Epoch: [11][20/200]	Time 0.363 (0.430)	Data 0.001 (0.058)	Loss 0.289 (0.328)
Epoch: [11][40/200]	Time 0.362 (0.441)	Data 0.001 (0.073)	Loss 1.633 (0.562)
Epoch: [11][60/200]	Time 0.361 (0.415)	Data 0.000 (0.049)	Loss 1.527 (0.907)
Epoch: [11][80/200]	Time 0.366 (0.423)	Data 0.001 (0.058)	Loss 1.413 (1.086)
Epoch: [11][100/200]	Time 0.365 (0.429)	Data 0.001 (0.063)	Loss 1.678 (1.185)
Epoch: [11][120/200]	Time 0.364 (0.418)	Data 0.001 (0.052)	Loss 1.201 (1.258)
Epoch: [11][140/200]	Time 0.365 (0.422)	Data 0.001 (0.056)	Loss 1.234 (1.307)
Epoch: [11][160/200]	Time 0.364 (0.416)	Data 0.000 (0.049)	Loss 1.621 (1.349)
Epoch: [11][180/200]	Time 0.365 (0.420)	Data 0.001 (0.053)	Loss 1.335 (1.374)
Epoch: [11][200/200]	Time 0.368 (0.424)	Data 0.001 (0.057)	Loss 1.311 (1.399)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.216)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.119 (0.197)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.6374990940094
==> Statistics for epoch 12: 1044 clusters
Epoch: [12][20/200]	Time 0.362 (0.420)	Data 0.000 (0.055)	Loss 0.238 (0.314)
Epoch: [12][40/200]	Time 0.360 (0.436)	Data 0.000 (0.072)	Loss 1.574 (0.547)
Epoch: [12][60/200]	Time 0.361 (0.412)	Data 0.000 (0.048)	Loss 1.984 (0.865)
Epoch: [12][80/200]	Time 0.367 (0.422)	Data 0.001 (0.059)	Loss 1.724 (1.025)
Epoch: [12][100/200]	Time 0.362 (0.429)	Data 0.001 (0.065)	Loss 1.561 (1.132)
Epoch: [12][120/200]	Time 0.366 (0.419)	Data 0.001 (0.054)	Loss 1.952 (1.217)
Epoch: [12][140/200]	Time 0.364 (0.424)	Data 0.001 (0.058)	Loss 1.273 (1.253)
Epoch: [12][160/200]	Time 0.363 (0.416)	Data 0.000 (0.051)	Loss 1.758 (1.303)
Epoch: [12][180/200]	Time 0.366 (0.420)	Data 0.001 (0.055)	Loss 1.841 (1.341)
Epoch: [12][200/200]	Time 0.366 (0.424)	Data 0.001 (0.058)	Loss 1.628 (1.361)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.210)	Data 0.000 (0.089)	
Extract Features: [100/128]	Time 0.121 (0.193)	Data 0.000 (0.073)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.242515563964844
==> Statistics for epoch 13: 1021 clusters
Epoch: [13][20/200]	Time 0.365 (0.431)	Data 0.001 (0.056)	Loss 0.230 (0.295)
Epoch: [13][40/200]	Time 0.367 (0.443)	Data 0.001 (0.073)	Loss 1.566 (0.566)
Epoch: [13][60/200]	Time 0.364 (0.417)	Data 0.000 (0.049)	Loss 1.481 (0.876)
Epoch: [13][80/200]	Time 0.362 (0.429)	Data 0.001 (0.061)	Loss 1.835 (1.034)
Epoch: [13][100/200]	Time 0.366 (0.435)	Data 0.001 (0.067)	Loss 1.561 (1.131)
Epoch: [13][120/200]	Time 0.362 (0.423)	Data 0.000 (0.056)	Loss 1.306 (1.205)
Epoch: [13][140/200]	Time 0.371 (0.428)	Data 0.001 (0.061)	Loss 1.567 (1.253)
Epoch: [13][160/200]	Time 0.368 (0.432)	Data 0.001 (0.065)	Loss 1.447 (1.292)
Epoch: [13][180/200]	Time 0.366 (0.425)	Data 0.000 (0.058)	Loss 1.322 (1.324)
Epoch: [13][200/200]	Time 0.374 (0.429)	Data 0.001 (0.061)	Loss 1.296 (1.344)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.218)	Data 0.000 (0.100)	
Extract Features: [100/128]	Time 0.151 (0.196)	Data 0.032 (0.076)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.64910554885864
==> Statistics for epoch 14: 1048 clusters
Epoch: [14][20/200]	Time 0.361 (0.421)	Data 0.001 (0.056)	Loss 0.273 (0.288)
Epoch: [14][40/200]	Time 0.365 (0.443)	Data 0.001 (0.075)	Loss 1.478 (0.513)
Epoch: [14][60/200]	Time 0.366 (0.417)	Data 0.000 (0.050)	Loss 2.090 (0.837)
Epoch: [14][80/200]	Time 0.364 (0.426)	Data 0.001 (0.059)	Loss 1.453 (1.005)
Epoch: [14][100/200]	Time 0.365 (0.432)	Data 0.001 (0.065)	Loss 1.361 (1.097)
Epoch: [14][120/200]	Time 0.363 (0.421)	Data 0.001 (0.054)	Loss 1.564 (1.167)
Epoch: [14][140/200]	Time 0.370 (0.427)	Data 0.001 (0.060)	Loss 1.387 (1.212)
Epoch: [14][160/200]	Time 0.365 (0.420)	Data 0.000 (0.052)	Loss 1.642 (1.252)
Epoch: [14][180/200]	Time 0.366 (0.424)	Data 0.001 (0.056)	Loss 1.773 (1.283)
Epoch: [14][200/200]	Time 0.374 (0.427)	Data 0.001 (0.059)	Loss 1.708 (1.309)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.221)	Data 0.000 (0.100)	
Extract Features: [100/128]	Time 0.120 (0.198)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.779541969299316
==> Statistics for epoch 15: 1055 clusters
Epoch: [15][20/200]	Time 0.365 (0.421)	Data 0.001 (0.055)	Loss 0.266 (0.304)
Epoch: [15][40/200]	Time 0.367 (0.438)	Data 0.001 (0.073)	Loss 1.740 (0.505)
Epoch: [15][60/200]	Time 0.365 (0.414)	Data 0.000 (0.049)	Loss 1.506 (0.816)
Epoch: [15][80/200]	Time 0.367 (0.426)	Data 0.001 (0.059)	Loss 1.410 (0.981)
Epoch: [15][100/200]	Time 0.366 (0.433)	Data 0.001 (0.066)	Loss 1.308 (1.080)
Epoch: [15][120/200]	Time 0.365 (0.422)	Data 0.001 (0.055)	Loss 1.868 (1.158)
Epoch: [15][140/200]	Time 0.366 (0.429)	Data 0.001 (0.060)	Loss 1.483 (1.198)
Epoch: [15][160/200]	Time 0.371 (0.421)	Data 0.000 (0.053)	Loss 1.527 (1.240)
Epoch: [15][180/200]	Time 0.367 (0.425)	Data 0.001 (0.057)	Loss 1.276 (1.269)
Epoch: [15][200/200]	Time 0.368 (0.428)	Data 0.001 (0.060)	Loss 1.390 (1.299)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.212)	Data 0.000 (0.093)	
Extract Features: [100/128]	Time 0.119 (0.192)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.730361223220825
==> Statistics for epoch 16: 1059 clusters
Epoch: [16][20/200]	Time 0.365 (0.420)	Data 0.001 (0.054)	Loss 0.253 (0.295)
Epoch: [16][40/200]	Time 0.366 (0.437)	Data 0.001 (0.071)	Loss 1.325 (0.510)
Epoch: [16][60/200]	Time 0.367 (0.413)	Data 0.000 (0.047)	Loss 1.403 (0.828)
Epoch: [16][80/200]	Time 0.363 (0.422)	Data 0.000 (0.057)	Loss 1.505 (0.996)
Epoch: [16][100/200]	Time 2.102 (0.428)	Data 1.714 (0.062)	Loss 1.493 (1.078)
Epoch: [16][120/200]	Time 0.365 (0.418)	Data 0.001 (0.052)	Loss 1.265 (1.138)
Epoch: [16][140/200]	Time 0.369 (0.422)	Data 0.001 (0.056)	Loss 1.388 (1.196)
Epoch: [16][160/200]	Time 0.367 (0.415)	Data 0.000 (0.049)	Loss 1.401 (1.229)
Epoch: [16][180/200]	Time 0.369 (0.419)	Data 0.001 (0.053)	Loss 1.471 (1.254)
Epoch: [16][200/200]	Time 0.365 (0.423)	Data 0.001 (0.056)	Loss 2.141 (1.283)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.120 (0.218)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.239 (0.197)	Data 0.122 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.64882135391235
==> Statistics for epoch 17: 1007 clusters
Epoch: [17][20/200]	Time 0.363 (0.431)	Data 0.001 (0.059)	Loss 0.380 (0.353)
Epoch: [17][40/200]	Time 0.367 (0.446)	Data 0.001 (0.077)	Loss 1.696 (0.589)
Epoch: [17][60/200]	Time 0.365 (0.420)	Data 0.000 (0.051)	Loss 1.162 (0.865)
Epoch: [17][80/200]	Time 0.368 (0.431)	Data 0.001 (0.062)	Loss 1.589 (0.977)
Epoch: [17][100/200]	Time 0.367 (0.436)	Data 0.001 (0.067)	Loss 1.301 (1.058)
Epoch: [17][120/200]	Time 0.366 (0.424)	Data 0.000 (0.056)	Loss 1.467 (1.109)
Epoch: [17][140/200]	Time 0.368 (0.428)	Data 0.001 (0.060)	Loss 1.683 (1.152)
Epoch: [17][160/200]	Time 0.376 (0.431)	Data 0.001 (0.063)	Loss 1.528 (1.183)
Epoch: [17][180/200]	Time 0.366 (0.424)	Data 0.000 (0.056)	Loss 1.277 (1.209)
Epoch: [17][200/200]	Time 0.370 (0.427)	Data 0.001 (0.059)	Loss 1.341 (1.216)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.121 (0.219)	Data 0.000 (0.097)	
Extract Features: [100/128]	Time 0.119 (0.197)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.18961691856384
==> Statistics for epoch 18: 1035 clusters
Epoch: [18][20/200]	Time 0.366 (0.421)	Data 0.001 (0.054)	Loss 0.279 (0.274)
Epoch: [18][40/200]	Time 0.364 (0.439)	Data 0.001 (0.074)	Loss 1.350 (0.468)
Epoch: [18][60/200]	Time 0.364 (0.414)	Data 0.000 (0.049)	Loss 1.463 (0.804)
Epoch: [18][80/200]	Time 0.365 (0.424)	Data 0.001 (0.059)	Loss 1.305 (0.961)
Epoch: [18][100/200]	Time 0.367 (0.432)	Data 0.001 (0.065)	Loss 1.237 (1.045)
Epoch: [18][120/200]	Time 0.367 (0.421)	Data 0.001 (0.055)	Loss 1.020 (1.103)
Epoch: [18][140/200]	Time 0.365 (0.427)	Data 0.001 (0.060)	Loss 1.714 (1.149)
Epoch: [18][160/200]	Time 0.369 (0.419)	Data 0.000 (0.053)	Loss 1.575 (1.179)
Epoch: [18][180/200]	Time 0.368 (0.425)	Data 0.001 (0.057)	Loss 1.405 (1.201)
Epoch: [18][200/200]	Time 0.374 (0.428)	Data 0.001 (0.060)	Loss 1.649 (1.231)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.120 (0.217)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.124 (0.197)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.23626494407654
==> Statistics for epoch 19: 1043 clusters
Epoch: [19][20/200]	Time 0.365 (0.417)	Data 0.001 (0.051)	Loss 0.154 (0.281)
Epoch: [19][40/200]	Time 0.361 (0.437)	Data 0.001 (0.072)	Loss 1.365 (0.482)
Epoch: [19][60/200]	Time 0.362 (0.415)	Data 0.000 (0.048)	Loss 1.219 (0.784)
Epoch: [19][80/200]	Time 0.369 (0.425)	Data 0.001 (0.058)	Loss 0.960 (0.934)
Epoch: [19][100/200]	Time 0.365 (0.433)	Data 0.001 (0.066)	Loss 1.596 (1.043)
Epoch: [19][120/200]	Time 0.364 (0.422)	Data 0.001 (0.055)	Loss 1.726 (1.100)
Epoch: [19][140/200]	Time 0.366 (0.426)	Data 0.001 (0.059)	Loss 1.294 (1.147)
Epoch: [19][160/200]	Time 0.365 (0.418)	Data 0.000 (0.052)	Loss 1.324 (1.186)
Epoch: [19][180/200]	Time 0.369 (0.423)	Data 0.001 (0.057)	Loss 1.525 (1.210)
Epoch: [19][200/200]	Time 0.367 (0.426)	Data 0.001 (0.060)	Loss 1.540 (1.243)
Extract Features: [50/367]	Time 0.118 (0.215)	Data 0.001 (0.094)	
Extract Features: [100/367]	Time 0.188 (0.198)	Data 0.072 (0.076)	
Extract Features: [150/367]	Time 0.117 (0.192)	Data 0.000 (0.071)	
Extract Features: [200/367]	Time 0.230 (0.189)	Data 0.111 (0.068)	
Extract Features: [250/367]	Time 0.117 (0.188)	Data 0.000 (0.067)	
Extract Features: [300/367]	Time 0.366 (0.186)	Data 0.245 (0.066)	
Extract Features: [350/367]	Time 0.116 (0.185)	Data 0.000 (0.064)	
Mean AP: 58.6%

 * Finished epoch  19  model mAP: 58.6%  best: 58.6% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.214)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.120 (0.192)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.1347279548645
==> Statistics for epoch 20: 1043 clusters
Epoch: [20][20/200]	Time 0.361 (0.421)	Data 0.001 (0.054)	Loss 0.222 (0.256)
Epoch: [20][40/200]	Time 0.361 (0.441)	Data 0.001 (0.076)	Loss 1.198 (0.461)
Epoch: [20][60/200]	Time 0.363 (0.415)	Data 0.000 (0.051)	Loss 1.547 (0.775)
Epoch: [20][80/200]	Time 0.365 (0.427)	Data 0.001 (0.061)	Loss 1.058 (0.934)
Epoch: [20][100/200]	Time 0.364 (0.433)	Data 0.001 (0.067)	Loss 1.350 (1.051)
Epoch: [20][120/200]	Time 0.365 (0.422)	Data 0.001 (0.056)	Loss 1.203 (1.100)
Epoch: [20][140/200]	Time 0.364 (0.427)	Data 0.001 (0.061)	Loss 1.303 (1.138)
Epoch: [20][160/200]	Time 0.364 (0.420)	Data 0.000 (0.053)	Loss 1.269 (1.171)
Epoch: [20][180/200]	Time 0.362 (0.424)	Data 0.001 (0.058)	Loss 1.272 (1.196)
Epoch: [20][200/200]	Time 0.368 (0.428)	Data 0.001 (0.062)	Loss 1.086 (1.209)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.220)	Data 0.000 (0.098)	
Extract Features: [100/128]	Time 0.122 (0.195)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.73014497756958
==> Statistics for epoch 21: 1053 clusters
Epoch: [21][20/200]	Time 0.367 (0.421)	Data 0.001 (0.053)	Loss 0.235 (0.255)
Epoch: [21][40/200]	Time 0.493 (0.441)	Data 0.001 (0.071)	Loss 1.324 (0.450)
Epoch: [21][60/200]	Time 0.363 (0.416)	Data 0.000 (0.048)	Loss 1.773 (0.761)
Epoch: [21][80/200]	Time 0.366 (0.427)	Data 0.001 (0.059)	Loss 1.635 (0.939)
Epoch: [21][100/200]	Time 0.367 (0.432)	Data 0.001 (0.064)	Loss 1.358 (1.016)
Epoch: [21][120/200]	Time 0.369 (0.421)	Data 0.001 (0.053)	Loss 1.238 (1.072)
Epoch: [21][140/200]	Time 0.367 (0.426)	Data 0.001 (0.059)	Loss 1.032 (1.116)
Epoch: [21][160/200]	Time 0.364 (0.419)	Data 0.000 (0.051)	Loss 1.572 (1.149)
Epoch: [21][180/200]	Time 0.363 (0.423)	Data 0.001 (0.055)	Loss 1.515 (1.182)
Epoch: [21][200/200]	Time 0.364 (0.426)	Data 0.001 (0.058)	Loss 1.589 (1.197)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.120 (0.217)	Data 0.000 (0.094)	
Extract Features: [100/128]	Time 0.118 (0.194)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.03773903846741
==> Statistics for epoch 22: 1067 clusters
Epoch: [22][20/200]	Time 0.371 (0.416)	Data 0.001 (0.050)	Loss 0.394 (0.252)
Epoch: [22][40/200]	Time 0.372 (0.440)	Data 0.001 (0.071)	Loss 1.389 (0.438)
Epoch: [22][60/200]	Time 0.365 (0.414)	Data 0.000 (0.048)	Loss 1.587 (0.745)
Epoch: [22][80/200]	Time 0.365 (0.422)	Data 0.001 (0.056)	Loss 1.175 (0.895)
Epoch: [22][100/200]	Time 2.174 (0.429)	Data 1.786 (0.062)	Loss 1.302 (0.984)
Epoch: [22][120/200]	Time 0.364 (0.418)	Data 0.001 (0.052)	Loss 1.429 (1.044)
Epoch: [22][140/200]	Time 0.369 (0.424)	Data 0.001 (0.057)	Loss 1.450 (1.093)
Epoch: [22][160/200]	Time 0.363 (0.416)	Data 0.000 (0.050)	Loss 1.469 (1.143)
Epoch: [22][180/200]	Time 0.369 (0.421)	Data 0.001 (0.055)	Loss 1.405 (1.166)
Epoch: [22][200/200]	Time 0.365 (0.425)	Data 0.001 (0.059)	Loss 1.404 (1.194)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.210)	Data 0.000 (0.088)	
Extract Features: [100/128]	Time 0.159 (0.192)	Data 0.042 (0.069)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.54163670539856
==> Statistics for epoch 23: 1063 clusters
Epoch: [23][20/200]	Time 0.365 (0.429)	Data 0.001 (0.055)	Loss 0.210 (0.246)
Epoch: [23][40/200]	Time 0.368 (0.440)	Data 0.001 (0.070)	Loss 1.100 (0.429)
Epoch: [23][60/200]	Time 0.364 (0.416)	Data 0.000 (0.047)	Loss 1.510 (0.766)
Epoch: [23][80/200]	Time 0.366 (0.425)	Data 0.001 (0.057)	Loss 1.690 (0.922)
Epoch: [23][100/200]	Time 2.230 (0.432)	Data 1.842 (0.064)	Loss 1.541 (1.016)
Epoch: [23][120/200]	Time 0.374 (0.421)	Data 0.001 (0.053)	Loss 1.606 (1.076)
Epoch: [23][140/200]	Time 0.371 (0.427)	Data 0.000 (0.059)	Loss 1.714 (1.133)
Epoch: [23][160/200]	Time 0.368 (0.420)	Data 0.000 (0.052)	Loss 1.181 (1.160)
Epoch: [23][180/200]	Time 0.370 (0.424)	Data 0.001 (0.056)	Loss 1.249 (1.180)
Epoch: [23][200/200]	Time 0.366 (0.428)	Data 0.001 (0.060)	Loss 1.580 (1.213)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.121 (0.216)	Data 0.000 (0.093)	
Extract Features: [100/128]	Time 0.123 (0.195)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.1205153465271
==> Statistics for epoch 24: 1067 clusters
Epoch: [24][20/200]	Time 0.367 (0.427)	Data 0.001 (0.062)	Loss 0.210 (0.259)
Epoch: [24][40/200]	Time 0.362 (0.442)	Data 0.001 (0.076)	Loss 1.191 (0.415)
Epoch: [24][60/200]	Time 0.362 (0.419)	Data 0.000 (0.051)	Loss 1.316 (0.728)
Epoch: [24][80/200]	Time 0.364 (0.428)	Data 0.001 (0.061)	Loss 1.641 (0.903)
Epoch: [24][100/200]	Time 2.211 (0.434)	Data 1.822 (0.067)	Loss 1.374 (0.971)
Epoch: [24][120/200]	Time 0.363 (0.423)	Data 0.001 (0.056)	Loss 1.770 (1.056)
Epoch: [24][140/200]	Time 0.367 (0.427)	Data 0.001 (0.061)	Loss 1.221 (1.097)
Epoch: [24][160/200]	Time 0.517 (0.421)	Data 0.000 (0.053)	Loss 1.175 (1.135)
Epoch: [24][180/200]	Time 0.368 (0.425)	Data 0.001 (0.057)	Loss 1.769 (1.168)
Epoch: [24][200/200]	Time 0.375 (0.429)	Data 0.001 (0.061)	Loss 1.195 (1.196)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.471 (0.216)	Data 0.353 (0.095)	
Extract Features: [100/128]	Time 0.119 (0.194)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.45808529853821
==> Statistics for epoch 25: 1068 clusters
Epoch: [25][20/200]	Time 0.360 (0.423)	Data 0.001 (0.057)	Loss 0.225 (0.277)
Epoch: [25][40/200]	Time 0.369 (0.438)	Data 0.001 (0.073)	Loss 1.302 (0.402)
Epoch: [25][60/200]	Time 0.363 (0.414)	Data 0.000 (0.049)	Loss 1.613 (0.731)
Epoch: [25][80/200]	Time 0.366 (0.424)	Data 0.001 (0.059)	Loss 1.189 (0.884)
Epoch: [25][100/200]	Time 2.120 (0.432)	Data 1.730 (0.064)	Loss 1.379 (0.973)
Epoch: [25][120/200]	Time 0.366 (0.421)	Data 0.001 (0.054)	Loss 1.713 (1.047)
Epoch: [25][140/200]	Time 0.368 (0.425)	Data 0.001 (0.058)	Loss 1.496 (1.101)
Epoch: [25][160/200]	Time 0.365 (0.419)	Data 0.000 (0.051)	Loss 1.783 (1.137)
Epoch: [25][180/200]	Time 0.365 (0.422)	Data 0.001 (0.054)	Loss 1.214 (1.160)
Epoch: [25][200/200]	Time 0.366 (0.426)	Data 0.001 (0.058)	Loss 1.834 (1.184)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.213)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.119 (0.192)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.28943848609924
==> Statistics for epoch 26: 1067 clusters
Epoch: [26][20/200]	Time 0.363 (0.425)	Data 0.001 (0.052)	Loss 0.189 (0.242)
Epoch: [26][40/200]	Time 0.365 (0.439)	Data 0.001 (0.069)	Loss 1.322 (0.442)
Epoch: [26][60/200]	Time 0.364 (0.415)	Data 0.000 (0.046)	Loss 1.486 (0.743)
Epoch: [26][80/200]	Time 0.363 (0.426)	Data 0.001 (0.059)	Loss 1.624 (0.886)
Epoch: [26][100/200]	Time 2.100 (0.432)	Data 1.713 (0.064)	Loss 1.868 (0.999)
Epoch: [26][120/200]	Time 0.371 (0.422)	Data 0.001 (0.054)	Loss 1.440 (1.076)
Epoch: [26][140/200]	Time 0.362 (0.426)	Data 0.001 (0.058)	Loss 1.271 (1.132)
Epoch: [26][160/200]	Time 0.366 (0.419)	Data 0.000 (0.051)	Loss 1.554 (1.163)
Epoch: [26][180/200]	Time 0.368 (0.424)	Data 0.001 (0.056)	Loss 1.683 (1.191)
Epoch: [26][200/200]	Time 0.364 (0.428)	Data 0.001 (0.060)	Loss 1.550 (1.206)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.209)	Data 0.000 (0.091)	
Extract Features: [100/128]	Time 0.121 (0.194)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.25462770462036
==> Statistics for epoch 27: 1059 clusters
Epoch: [27][20/200]	Time 0.368 (0.424)	Data 0.001 (0.059)	Loss 0.238 (0.245)
Epoch: [27][40/200]	Time 0.364 (0.442)	Data 0.001 (0.078)	Loss 1.175 (0.410)
Epoch: [27][60/200]	Time 0.366 (0.417)	Data 0.000 (0.052)	Loss 1.174 (0.720)
Epoch: [27][80/200]	Time 0.367 (0.430)	Data 0.001 (0.063)	Loss 0.885 (0.871)
Epoch: [27][100/200]	Time 2.227 (0.436)	Data 1.838 (0.069)	Loss 1.636 (0.983)
Epoch: [27][120/200]	Time 0.368 (0.424)	Data 0.001 (0.058)	Loss 1.317 (1.046)
Epoch: [27][140/200]	Time 0.364 (0.430)	Data 0.001 (0.063)	Loss 1.849 (1.093)
Epoch: [27][160/200]	Time 0.364 (0.422)	Data 0.000 (0.055)	Loss 1.370 (1.128)
Epoch: [27][180/200]	Time 0.368 (0.425)	Data 0.001 (0.058)	Loss 1.122 (1.156)
Epoch: [27][200/200]	Time 0.366 (0.429)	Data 0.001 (0.062)	Loss 1.444 (1.181)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.260 (0.213)	Data 0.143 (0.091)	
Extract Features: [100/128]	Time 0.119 (0.193)	Data 0.000 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.88459920883179
==> Statistics for epoch 28: 1062 clusters
Epoch: [28][20/200]	Time 0.361 (0.416)	Data 0.001 (0.051)	Loss 0.218 (0.249)
Epoch: [28][40/200]	Time 0.364 (0.440)	Data 0.001 (0.072)	Loss 1.133 (0.433)
Epoch: [28][60/200]	Time 0.364 (0.415)	Data 0.000 (0.048)	Loss 1.444 (0.732)
Epoch: [28][80/200]	Time 0.366 (0.425)	Data 0.001 (0.058)	Loss 1.502 (0.869)
Epoch: [28][100/200]	Time 2.158 (0.431)	Data 1.778 (0.064)	Loss 1.023 (0.969)
Epoch: [28][120/200]	Time 0.367 (0.420)	Data 0.001 (0.054)	Loss 1.101 (1.042)
Epoch: [28][140/200]	Time 0.367 (0.426)	Data 0.001 (0.059)	Loss 1.340 (1.089)
Epoch: [28][160/200]	Time 0.364 (0.418)	Data 0.000 (0.052)	Loss 1.146 (1.136)
Epoch: [28][180/200]	Time 0.364 (0.423)	Data 0.001 (0.056)	Loss 1.571 (1.163)
Epoch: [28][200/200]	Time 0.366 (0.426)	Data 0.001 (0.060)	Loss 1.226 (1.182)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.218)	Data 0.000 (0.097)	
Extract Features: [100/128]	Time 0.124 (0.195)	Data 0.000 (0.073)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.54994297027588
==> Statistics for epoch 29: 1059 clusters
Epoch: [29][20/200]	Time 0.363 (0.428)	Data 0.001 (0.054)	Loss 0.238 (0.238)
Epoch: [29][40/200]	Time 0.368 (0.440)	Data 0.001 (0.070)	Loss 1.494 (0.426)
Epoch: [29][60/200]	Time 0.373 (0.417)	Data 0.000 (0.047)	Loss 0.897 (0.737)
Epoch: [29][80/200]	Time 0.367 (0.427)	Data 0.001 (0.058)	Loss 1.245 (0.904)
Epoch: [29][100/200]	Time 2.106 (0.432)	Data 1.711 (0.064)	Loss 1.239 (0.996)
Epoch: [29][120/200]	Time 0.366 (0.423)	Data 0.001 (0.053)	Loss 1.222 (1.064)
Epoch: [29][140/200]	Time 0.367 (0.429)	Data 0.001 (0.060)	Loss 1.439 (1.116)
Epoch: [29][160/200]	Time 0.369 (0.421)	Data 0.000 (0.052)	Loss 1.397 (1.149)
Epoch: [29][180/200]	Time 0.368 (0.424)	Data 0.000 (0.056)	Loss 1.374 (1.177)
Epoch: [29][200/200]	Time 0.366 (0.428)	Data 0.001 (0.059)	Loss 1.611 (1.202)
Extract Features: [50/367]	Time 0.280 (0.212)	Data 0.163 (0.090)	
Extract Features: [100/367]	Time 0.186 (0.195)	Data 0.068 (0.074)	
Extract Features: [150/367]	Time 0.297 (0.191)	Data 0.173 (0.070)	
Extract Features: [200/367]	Time 0.250 (0.188)	Data 0.130 (0.067)	
Extract Features: [250/367]	Time 0.120 (0.188)	Data 0.000 (0.067)	
Extract Features: [300/367]	Time 0.428 (0.187)	Data 0.306 (0.066)	
Extract Features: [350/367]	Time 0.120 (0.186)	Data 0.000 (0.064)	
Mean AP: 61.9%

 * Finished epoch  29  model mAP: 61.9%  best: 61.9% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.218)	Data 0.000 (0.099)	
Extract Features: [100/128]	Time 0.120 (0.195)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.70884108543396
==> Statistics for epoch 30: 1058 clusters
Epoch: [30][20/200]	Time 0.367 (0.426)	Data 0.001 (0.061)	Loss 0.325 (0.232)
Epoch: [30][40/200]	Time 0.366 (0.434)	Data 0.001 (0.070)	Loss 1.407 (0.454)
Epoch: [30][60/200]	Time 0.365 (0.411)	Data 0.000 (0.047)	Loss 1.576 (0.746)
Epoch: [30][80/200]	Time 0.368 (0.422)	Data 0.001 (0.057)	Loss 1.374 (0.901)
Epoch: [30][100/200]	Time 2.216 (0.431)	Data 1.827 (0.064)	Loss 1.467 (0.986)
Epoch: [30][120/200]	Time 0.366 (0.421)	Data 0.001 (0.054)	Loss 1.402 (1.058)
Epoch: [30][140/200]	Time 0.389 (0.424)	Data 0.001 (0.058)	Loss 1.203 (1.100)
Epoch: [30][160/200]	Time 0.365 (0.417)	Data 0.000 (0.051)	Loss 1.533 (1.123)
Epoch: [30][180/200]	Time 0.366 (0.422)	Data 0.001 (0.055)	Loss 1.278 (1.151)
Epoch: [30][200/200]	Time 0.368 (0.426)	Data 0.001 (0.059)	Loss 1.013 (1.175)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.207 (0.214)	Data 0.088 (0.093)	
Extract Features: [100/128]	Time 0.116 (0.192)	Data 0.000 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.27864599227905
==> Statistics for epoch 31: 1076 clusters
Epoch: [31][20/200]	Time 0.363 (0.418)	Data 0.001 (0.053)	Loss 0.202 (0.268)
Epoch: [31][40/200]	Time 0.365 (0.437)	Data 0.001 (0.072)	Loss 0.983 (0.414)
Epoch: [31][60/200]	Time 0.367 (0.415)	Data 0.000 (0.048)	Loss 1.053 (0.725)
Epoch: [31][80/200]	Time 0.369 (0.427)	Data 0.001 (0.060)	Loss 1.390 (0.877)
Epoch: [31][100/200]	Time 2.156 (0.433)	Data 1.760 (0.066)	Loss 1.445 (0.973)
Epoch: [31][120/200]	Time 0.364 (0.423)	Data 0.000 (0.055)	Loss 1.809 (1.051)
Epoch: [31][140/200]	Time 0.372 (0.428)	Data 0.000 (0.060)	Loss 1.132 (1.094)
Epoch: [31][160/200]	Time 0.362 (0.420)	Data 0.000 (0.052)	Loss 1.684 (1.133)
Epoch: [31][180/200]	Time 0.368 (0.424)	Data 0.001 (0.056)	Loss 1.262 (1.152)
Epoch: [31][200/200]	Time 0.366 (0.427)	Data 0.000 (0.060)	Loss 1.394 (1.172)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.119 (0.210)	Data 0.000 (0.088)	
Extract Features: [100/128]	Time 0.303 (0.193)	Data 0.000 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.135730028152466
==> Statistics for epoch 32: 1061 clusters
Epoch: [32][20/200]	Time 0.364 (0.420)	Data 0.001 (0.055)	Loss 0.137 (0.224)
Epoch: [32][40/200]	Time 0.368 (0.439)	Data 0.001 (0.071)	Loss 1.465 (0.400)
Epoch: [32][60/200]	Time 0.366 (0.415)	Data 0.000 (0.047)	Loss 1.264 (0.707)
Epoch: [32][80/200]	Time 0.365 (0.426)	Data 0.001 (0.059)	Loss 1.652 (0.860)
Epoch: [32][100/200]	Time 2.247 (0.433)	Data 1.858 (0.066)	Loss 1.512 (0.963)
Epoch: [32][120/200]	Time 0.367 (0.423)	Data 0.002 (0.055)	Loss 1.244 (1.023)
Epoch: [32][140/200]	Time 0.364 (0.428)	Data 0.001 (0.060)	Loss 1.058 (1.063)
Epoch: [32][160/200]	Time 0.361 (0.420)	Data 0.000 (0.053)	Loss 1.353 (1.098)
Epoch: [32][180/200]	Time 0.364 (0.425)	Data 0.000 (0.057)	Loss 1.298 (1.138)
Epoch: [32][200/200]	Time 0.365 (0.428)	Data 0.000 (0.061)	Loss 1.446 (1.164)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.120 (0.216)	Data 0.000 (0.093)	
Extract Features: [100/128]	Time 0.117 (0.199)	Data 0.000 (0.076)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.254494190216064
==> Statistics for epoch 33: 1067 clusters
Epoch: [33][20/200]	Time 0.364 (0.417)	Data 0.001 (0.052)	Loss 0.252 (0.252)
Epoch: [33][40/200]	Time 0.366 (0.432)	Data 0.001 (0.067)	Loss 1.395 (0.423)
Epoch: [33][60/200]	Time 0.362 (0.412)	Data 0.000 (0.045)	Loss 1.372 (0.737)
Epoch: [33][80/200]	Time 0.364 (0.423)	Data 0.001 (0.056)	Loss 1.413 (0.877)
Epoch: [33][100/200]	Time 2.049 (0.429)	Data 1.663 (0.062)	Loss 0.954 (0.971)
Epoch: [33][120/200]	Time 0.363 (0.418)	Data 0.001 (0.052)	Loss 1.311 (1.039)
Epoch: [33][140/200]	Time 0.366 (0.423)	Data 0.001 (0.056)	Loss 1.174 (1.091)
Epoch: [33][160/200]	Time 0.368 (0.416)	Data 0.000 (0.049)	Loss 1.809 (1.122)
Epoch: [33][180/200]	Time 0.364 (0.421)	Data 0.001 (0.054)	Loss 1.389 (1.154)
Epoch: [33][200/200]	Time 0.369 (0.423)	Data 0.001 (0.056)	Loss 1.497 (1.169)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.216)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.119 (0.193)	Data 0.000 (0.071)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.65109419822693
==> Statistics for epoch 34: 1065 clusters
Epoch: [34][20/200]	Time 0.361 (0.418)	Data 0.001 (0.054)	Loss 0.404 (0.244)
Epoch: [34][40/200]	Time 0.362 (0.442)	Data 0.001 (0.075)	Loss 1.551 (0.433)
Epoch: [34][60/200]	Time 0.363 (0.416)	Data 0.001 (0.050)	Loss 1.356 (0.756)
Epoch: [34][80/200]	Time 0.373 (0.427)	Data 0.001 (0.060)	Loss 1.301 (0.922)
Epoch: [34][100/200]	Time 2.009 (0.431)	Data 1.621 (0.065)	Loss 1.285 (0.988)
Epoch: [34][120/200]	Time 0.363 (0.420)	Data 0.001 (0.054)	Loss 1.411 (1.046)
Epoch: [34][140/200]	Time 0.363 (0.424)	Data 0.001 (0.058)	Loss 1.197 (1.092)
Epoch: [34][160/200]	Time 0.364 (0.417)	Data 0.000 (0.051)	Loss 1.695 (1.125)
Epoch: [34][180/200]	Time 0.378 (0.421)	Data 0.001 (0.055)	Loss 1.412 (1.148)
Epoch: [34][200/200]	Time 0.371 (0.424)	Data 0.001 (0.058)	Loss 1.472 (1.164)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.572 (0.217)	Data 0.450 (0.094)	
Extract Features: [100/128]	Time 0.120 (0.196)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.15971493721008
==> Statistics for epoch 35: 1053 clusters
Epoch: [35][20/200]	Time 0.365 (0.423)	Data 0.001 (0.058)	Loss 0.218 (0.224)
Epoch: [35][40/200]	Time 0.369 (0.439)	Data 0.001 (0.074)	Loss 1.422 (0.436)
Epoch: [35][60/200]	Time 0.364 (0.417)	Data 0.000 (0.050)	Loss 1.464 (0.737)
Epoch: [35][80/200]	Time 0.371 (0.428)	Data 0.001 (0.060)	Loss 1.180 (0.879)
Epoch: [35][100/200]	Time 0.367 (0.433)	Data 0.001 (0.065)	Loss 1.794 (0.978)
Epoch: [35][120/200]	Time 0.365 (0.422)	Data 0.001 (0.055)	Loss 1.440 (1.022)
Epoch: [35][140/200]	Time 0.372 (0.428)	Data 0.001 (0.061)	Loss 1.184 (1.064)
Epoch: [35][160/200]	Time 0.363 (0.421)	Data 0.000 (0.053)	Loss 1.668 (1.107)
Epoch: [35][180/200]	Time 0.362 (0.425)	Data 0.000 (0.057)	Loss 1.621 (1.132)
Epoch: [35][200/200]	Time 0.373 (0.429)	Data 0.001 (0.061)	Loss 1.317 (1.152)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.271 (0.214)	Data 0.155 (0.091)	
Extract Features: [100/128]	Time 0.116 (0.195)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.75919771194458
==> Statistics for epoch 36: 1065 clusters
Epoch: [36][20/200]	Time 0.365 (0.422)	Data 0.001 (0.055)	Loss 0.150 (0.226)
Epoch: [36][40/200]	Time 0.362 (0.437)	Data 0.001 (0.071)	Loss 1.191 (0.397)
Epoch: [36][60/200]	Time 0.363 (0.413)	Data 0.000 (0.047)	Loss 1.280 (0.689)
Epoch: [36][80/200]	Time 0.366 (0.423)	Data 0.001 (0.057)	Loss 1.030 (0.833)
Epoch: [36][100/200]	Time 2.205 (0.432)	Data 1.816 (0.064)	Loss 1.165 (0.932)
Epoch: [36][120/200]	Time 0.365 (0.421)	Data 0.000 (0.053)	Loss 1.286 (1.002)
Epoch: [36][140/200]	Time 0.366 (0.425)	Data 0.000 (0.057)	Loss 0.932 (1.050)
Epoch: [36][160/200]	Time 0.362 (0.419)	Data 0.000 (0.050)	Loss 1.635 (1.098)
Epoch: [36][180/200]	Time 0.368 (0.423)	Data 0.001 (0.054)	Loss 1.609 (1.129)
Epoch: [36][200/200]	Time 0.368 (0.427)	Data 0.000 (0.059)	Loss 1.096 (1.151)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.214)	Data 0.000 (0.093)	
Extract Features: [100/128]	Time 0.124 (0.194)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.76071810722351
==> Statistics for epoch 37: 1069 clusters
Epoch: [37][20/200]	Time 0.361 (0.415)	Data 0.001 (0.050)	Loss 0.172 (0.243)
Epoch: [37][40/200]	Time 0.364 (0.436)	Data 0.001 (0.070)	Loss 1.416 (0.434)
Epoch: [37][60/200]	Time 0.395 (0.414)	Data 0.000 (0.047)	Loss 0.979 (0.720)
Epoch: [37][80/200]	Time 0.365 (0.427)	Data 0.001 (0.058)	Loss 1.383 (0.891)
Epoch: [37][100/200]	Time 2.176 (0.433)	Data 1.787 (0.065)	Loss 1.259 (0.973)
Epoch: [37][120/200]	Time 0.367 (0.422)	Data 0.001 (0.054)	Loss 2.054 (1.035)
Epoch: [37][140/200]	Time 0.368 (0.428)	Data 0.000 (0.059)	Loss 1.467 (1.093)
Epoch: [37][160/200]	Time 0.366 (0.421)	Data 0.000 (0.052)	Loss 1.230 (1.124)
Epoch: [37][180/200]	Time 0.366 (0.425)	Data 0.000 (0.057)	Loss 1.272 (1.149)
Epoch: [37][200/200]	Time 0.368 (0.428)	Data 0.001 (0.060)	Loss 1.598 (1.173)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.120 (0.211)	Data 0.000 (0.089)	
Extract Features: [100/128]	Time 0.140 (0.191)	Data 0.024 (0.070)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.34649705886841
==> Statistics for epoch 38: 1085 clusters
Epoch: [38][20/200]	Time 0.364 (0.424)	Data 0.001 (0.057)	Loss 0.216 (0.242)
Epoch: [38][40/200]	Time 0.366 (0.437)	Data 0.001 (0.070)	Loss 1.415 (0.438)
Epoch: [38][60/200]	Time 0.364 (0.413)	Data 0.000 (0.047)	Loss 1.545 (0.737)
Epoch: [38][80/200]	Time 0.368 (0.425)	Data 0.001 (0.057)	Loss 1.101 (0.867)
Epoch: [38][100/200]	Time 2.042 (0.430)	Data 1.655 (0.062)	Loss 1.147 (0.964)
Epoch: [38][120/200]	Time 0.373 (0.419)	Data 0.001 (0.052)	Loss 1.296 (1.045)
Epoch: [38][140/200]	Time 0.369 (0.424)	Data 0.001 (0.056)	Loss 1.411 (1.087)
Epoch: [38][160/200]	Time 0.365 (0.417)	Data 0.000 (0.050)	Loss 1.176 (1.134)
Epoch: [38][180/200]	Time 0.366 (0.421)	Data 0.001 (0.054)	Loss 1.369 (1.150)
Epoch: [38][200/200]	Time 0.368 (0.425)	Data 0.001 (0.058)	Loss 1.176 (1.173)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.265 (0.215)	Data 0.148 (0.094)	
Extract Features: [100/128]	Time 0.116 (0.195)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.45996117591858
==> Statistics for epoch 39: 1065 clusters
Epoch: [39][20/200]	Time 0.367 (0.424)	Data 0.001 (0.057)	Loss 0.170 (0.231)
Epoch: [39][40/200]	Time 0.370 (0.436)	Data 0.001 (0.069)	Loss 1.275 (0.427)
Epoch: [39][60/200]	Time 0.372 (0.414)	Data 0.000 (0.046)	Loss 1.086 (0.732)
Epoch: [39][80/200]	Time 0.367 (0.425)	Data 0.001 (0.056)	Loss 1.231 (0.885)
Epoch: [39][100/200]	Time 2.015 (0.430)	Data 1.627 (0.062)	Loss 1.417 (0.963)
Epoch: [39][120/200]	Time 0.367 (0.419)	Data 0.001 (0.051)	Loss 1.589 (1.031)
Epoch: [39][140/200]	Time 0.365 (0.426)	Data 0.001 (0.058)	Loss 1.485 (1.070)
Epoch: [39][160/200]	Time 0.366 (0.418)	Data 0.000 (0.050)	Loss 1.187 (1.110)
Epoch: [39][180/200]	Time 0.366 (0.425)	Data 0.001 (0.056)	Loss 1.338 (1.134)
Epoch: [39][200/200]	Time 0.367 (0.428)	Data 0.001 (0.060)	Loss 1.199 (1.152)
Extract Features: [50/367]	Time 0.121 (0.212)	Data 0.000 (0.090)	
Extract Features: [100/367]	Time 0.120 (0.195)	Data 0.000 (0.073)	
Extract Features: [150/367]	Time 0.116 (0.190)	Data 0.000 (0.069)	
Extract Features: [200/367]	Time 0.126 (0.188)	Data 0.000 (0.067)	
Extract Features: [250/367]	Time 0.285 (0.187)	Data 0.000 (0.065)	
Extract Features: [300/367]	Time 0.122 (0.187)	Data 0.000 (0.064)	
Extract Features: [350/367]	Time 0.199 (0.187)	Data 0.079 (0.064)	
Mean AP: 62.4%

 * Finished epoch  39  model mAP: 62.4%  best: 62.4% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.219)	Data 0.000 (0.097)	
Extract Features: [100/128]	Time 0.117 (0.195)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.97528958320618
==> Statistics for epoch 40: 1084 clusters
Epoch: [40][20/200]	Time 0.362 (0.423)	Data 0.001 (0.051)	Loss 0.243 (0.244)
Epoch: [40][40/200]	Time 0.367 (0.441)	Data 0.001 (0.073)	Loss 1.148 (0.410)
Epoch: [40][60/200]	Time 0.375 (0.416)	Data 0.000 (0.049)	Loss 0.978 (0.705)
Epoch: [40][80/200]	Time 0.371 (0.426)	Data 0.001 (0.059)	Loss 1.230 (0.866)
Epoch: [40][100/200]	Time 2.165 (0.432)	Data 1.775 (0.065)	Loss 1.294 (0.952)
Epoch: [40][120/200]	Time 0.364 (0.421)	Data 0.001 (0.055)	Loss 1.365 (1.018)
Epoch: [40][140/200]	Time 0.371 (0.426)	Data 0.001 (0.059)	Loss 1.156 (1.071)
Epoch: [40][160/200]	Time 0.371 (0.419)	Data 0.000 (0.052)	Loss 1.417 (1.093)
Epoch: [40][180/200]	Time 0.364 (0.423)	Data 0.001 (0.057)	Loss 1.450 (1.122)
Epoch: [40][200/200]	Time 0.491 (0.428)	Data 0.001 (0.061)	Loss 1.073 (1.143)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.120 (0.213)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.116 (0.195)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.85842537879944
==> Statistics for epoch 41: 1072 clusters
Epoch: [41][20/200]	Time 0.367 (0.423)	Data 0.001 (0.057)	Loss 0.194 (0.230)
Epoch: [41][40/200]	Time 0.364 (0.442)	Data 0.001 (0.073)	Loss 1.534 (0.408)
Epoch: [41][60/200]	Time 0.366 (0.417)	Data 0.000 (0.049)	Loss 1.431 (0.715)
Epoch: [41][80/200]	Time 0.365 (0.427)	Data 0.001 (0.059)	Loss 1.514 (0.880)
Epoch: [41][100/200]	Time 2.253 (0.433)	Data 1.867 (0.066)	Loss 1.216 (0.975)
Epoch: [41][120/200]	Time 0.363 (0.422)	Data 0.001 (0.055)	Loss 1.330 (1.035)
Epoch: [41][140/200]	Time 0.361 (0.427)	Data 0.001 (0.060)	Loss 1.862 (1.075)
Epoch: [41][160/200]	Time 0.364 (0.419)	Data 0.000 (0.053)	Loss 1.113 (1.107)
Epoch: [41][180/200]	Time 0.363 (0.424)	Data 0.001 (0.057)	Loss 0.949 (1.127)
Epoch: [41][200/200]	Time 0.365 (0.427)	Data 0.001 (0.060)	Loss 1.670 (1.148)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.474 (0.218)	Data 0.357 (0.097)	
Extract Features: [100/128]	Time 0.118 (0.195)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.96304488182068
==> Statistics for epoch 42: 1078 clusters
Epoch: [42][20/200]	Time 0.360 (0.423)	Data 0.001 (0.058)	Loss 0.202 (0.230)
Epoch: [42][40/200]	Time 0.365 (0.438)	Data 0.001 (0.074)	Loss 1.099 (0.389)
Epoch: [42][60/200]	Time 0.365 (0.414)	Data 0.000 (0.050)	Loss 1.107 (0.704)
Epoch: [42][80/200]	Time 0.368 (0.423)	Data 0.001 (0.058)	Loss 1.705 (0.884)
Epoch: [42][100/200]	Time 2.187 (0.431)	Data 1.797 (0.065)	Loss 1.508 (0.983)
Epoch: [42][120/200]	Time 0.366 (0.421)	Data 0.001 (0.054)	Loss 1.173 (1.053)
Epoch: [42][140/200]	Time 0.369 (0.425)	Data 0.001 (0.058)	Loss 1.174 (1.093)
Epoch: [42][160/200]	Time 0.366 (0.418)	Data 0.000 (0.051)	Loss 1.463 (1.133)
Epoch: [42][180/200]	Time 0.367 (0.422)	Data 0.001 (0.056)	Loss 1.239 (1.157)
Epoch: [42][200/200]	Time 0.364 (0.426)	Data 0.001 (0.058)	Loss 1.143 (1.179)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.124 (0.216)	Data 0.000 (0.093)	
Extract Features: [100/128]	Time 0.116 (0.193)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.89049172401428
==> Statistics for epoch 43: 1076 clusters
Epoch: [43][20/200]	Time 0.368 (0.422)	Data 0.001 (0.056)	Loss 0.248 (0.241)
Epoch: [43][40/200]	Time 0.367 (0.440)	Data 0.001 (0.075)	Loss 1.151 (0.432)
Epoch: [43][60/200]	Time 0.362 (0.418)	Data 0.000 (0.050)	Loss 1.108 (0.711)
Epoch: [43][80/200]	Time 0.374 (0.429)	Data 0.001 (0.061)	Loss 1.733 (0.876)
Epoch: [43][100/200]	Time 2.370 (0.436)	Data 1.975 (0.069)	Loss 1.300 (0.974)
Epoch: [43][120/200]	Time 0.370 (0.425)	Data 0.001 (0.058)	Loss 1.252 (1.046)
Epoch: [43][140/200]	Time 0.366 (0.431)	Data 0.001 (0.062)	Loss 1.487 (1.094)
Epoch: [43][160/200]	Time 0.365 (0.423)	Data 0.000 (0.055)	Loss 1.232 (1.125)
Epoch: [43][180/200]	Time 0.518 (0.428)	Data 0.001 (0.059)	Loss 1.556 (1.153)
Epoch: [43][200/200]	Time 0.371 (0.432)	Data 0.003 (0.063)	Loss 0.820 (1.168)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.239 (0.217)	Data 0.122 (0.099)	
Extract Features: [100/128]	Time 0.120 (0.196)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.77777600288391
==> Statistics for epoch 44: 1076 clusters
Epoch: [44][20/200]	Time 0.365 (0.418)	Data 0.001 (0.054)	Loss 0.176 (0.235)
Epoch: [44][40/200]	Time 0.364 (0.432)	Data 0.001 (0.067)	Loss 1.218 (0.406)
Epoch: [44][60/200]	Time 0.362 (0.409)	Data 0.000 (0.045)	Loss 0.917 (0.710)
Epoch: [44][80/200]	Time 0.365 (0.419)	Data 0.001 (0.054)	Loss 1.199 (0.860)
Epoch: [44][100/200]	Time 2.109 (0.425)	Data 1.719 (0.061)	Loss 1.405 (0.935)
Epoch: [44][120/200]	Time 0.377 (0.417)	Data 0.001 (0.051)	Loss 1.478 (1.008)
Epoch: [44][140/200]	Time 0.372 (0.422)	Data 0.001 (0.056)	Loss 1.860 (1.053)
Epoch: [44][160/200]	Time 0.362 (0.415)	Data 0.000 (0.049)	Loss 1.290 (1.096)
Epoch: [44][180/200]	Time 0.368 (0.420)	Data 0.001 (0.054)	Loss 1.206 (1.128)
Epoch: [44][200/200]	Time 0.366 (0.425)	Data 0.001 (0.058)	Loss 1.572 (1.155)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.269 (0.216)	Data 0.000 (0.093)	
Extract Features: [100/128]	Time 0.125 (0.195)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.132524490356445
==> Statistics for epoch 45: 1066 clusters
Epoch: [45][20/200]	Time 0.366 (0.429)	Data 0.001 (0.058)	Loss 0.213 (0.235)
Epoch: [45][40/200]	Time 0.365 (0.438)	Data 0.001 (0.070)	Loss 1.011 (0.384)
Epoch: [45][60/200]	Time 0.366 (0.414)	Data 0.000 (0.047)	Loss 1.129 (0.670)
Epoch: [45][80/200]	Time 0.368 (0.426)	Data 0.001 (0.059)	Loss 1.578 (0.846)
Epoch: [45][100/200]	Time 2.174 (0.432)	Data 1.775 (0.065)	Loss 1.070 (0.938)
Epoch: [45][120/200]	Time 0.362 (0.421)	Data 0.001 (0.054)	Loss 1.208 (1.012)
Epoch: [45][140/200]	Time 0.373 (0.427)	Data 0.001 (0.059)	Loss 1.189 (1.054)
Epoch: [45][160/200]	Time 0.363 (0.419)	Data 0.000 (0.052)	Loss 1.005 (1.093)
Epoch: [45][180/200]	Time 0.368 (0.423)	Data 0.001 (0.056)	Loss 1.589 (1.121)
Epoch: [45][200/200]	Time 0.365 (0.426)	Data 0.001 (0.059)	Loss 1.224 (1.139)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.122 (0.221)	Data 0.000 (0.098)	
Extract Features: [100/128]	Time 0.117 (0.196)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.235082149505615
==> Statistics for epoch 46: 1076 clusters
Epoch: [46][20/200]	Time 0.366 (0.424)	Data 0.001 (0.058)	Loss 0.190 (0.241)
Epoch: [46][40/200]	Time 0.363 (0.440)	Data 0.001 (0.074)	Loss 1.261 (0.399)
Epoch: [46][60/200]	Time 0.361 (0.415)	Data 0.000 (0.050)	Loss 1.183 (0.710)
Epoch: [46][80/200]	Time 0.366 (0.427)	Data 0.001 (0.061)	Loss 1.141 (0.874)
Epoch: [46][100/200]	Time 2.157 (0.434)	Data 1.778 (0.067)	Loss 1.456 (0.980)
Epoch: [46][120/200]	Time 0.369 (0.423)	Data 0.001 (0.056)	Loss 1.621 (1.046)
Epoch: [46][140/200]	Time 0.365 (0.428)	Data 0.001 (0.062)	Loss 1.506 (1.094)
Epoch: [46][160/200]	Time 0.365 (0.421)	Data 0.000 (0.054)	Loss 1.574 (1.118)
Epoch: [46][180/200]	Time 0.371 (0.425)	Data 0.001 (0.059)	Loss 1.505 (1.145)
Epoch: [46][200/200]	Time 0.377 (0.429)	Data 0.001 (0.062)	Loss 1.055 (1.164)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.119 (0.218)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.119 (0.196)	Data 0.000 (0.073)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.42084741592407
==> Statistics for epoch 47: 1078 clusters
Epoch: [47][20/200]	Time 0.365 (0.429)	Data 0.001 (0.057)	Loss 0.186 (0.229)
Epoch: [47][40/200]	Time 0.365 (0.445)	Data 0.001 (0.076)	Loss 1.617 (0.411)
Epoch: [47][60/200]	Time 0.362 (0.419)	Data 0.000 (0.051)	Loss 1.399 (0.685)
Epoch: [47][80/200]	Time 0.366 (0.430)	Data 0.001 (0.062)	Loss 1.805 (0.827)
Epoch: [47][100/200]	Time 2.174 (0.435)	Data 1.784 (0.068)	Loss 1.677 (0.930)
Epoch: [47][120/200]	Time 0.370 (0.425)	Data 0.001 (0.056)	Loss 1.481 (0.994)
Epoch: [47][140/200]	Time 0.367 (0.430)	Data 0.001 (0.062)	Loss 1.017 (1.027)
Epoch: [47][160/200]	Time 0.378 (0.422)	Data 0.000 (0.054)	Loss 1.275 (1.062)
Epoch: [47][180/200]	Time 0.366 (0.426)	Data 0.001 (0.058)	Loss 1.181 (1.095)
Epoch: [47][200/200]	Time 0.365 (0.430)	Data 0.001 (0.062)	Loss 1.241 (1.125)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.217)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.120 (0.195)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.34566116333008
==> Statistics for epoch 48: 1085 clusters
Epoch: [48][20/200]	Time 0.362 (0.419)	Data 0.001 (0.055)	Loss 0.249 (0.235)
Epoch: [48][40/200]	Time 0.365 (0.438)	Data 0.001 (0.074)	Loss 1.032 (0.418)
Epoch: [48][60/200]	Time 0.366 (0.414)	Data 0.000 (0.050)	Loss 1.245 (0.716)
Epoch: [48][80/200]	Time 0.363 (0.426)	Data 0.001 (0.059)	Loss 1.185 (0.848)
Epoch: [48][100/200]	Time 2.074 (0.431)	Data 1.683 (0.064)	Loss 1.865 (0.979)
Epoch: [48][120/200]	Time 0.370 (0.420)	Data 0.001 (0.054)	Loss 0.910 (1.043)
Epoch: [48][140/200]	Time 0.366 (0.427)	Data 0.001 (0.060)	Loss 1.552 (1.085)
Epoch: [48][160/200]	Time 0.368 (0.419)	Data 0.000 (0.052)	Loss 0.963 (1.121)
Epoch: [48][180/200]	Time 0.365 (0.424)	Data 0.001 (0.057)	Loss 1.066 (1.150)
Epoch: [48][200/200]	Time 0.365 (0.427)	Data 0.001 (0.061)	Loss 0.987 (1.174)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.123 (0.214)	Data 0.000 (0.091)	
Extract Features: [100/128]	Time 0.119 (0.195)	Data 0.000 (0.072)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.648887157440186
==> Statistics for epoch 49: 1083 clusters
Epoch: [49][20/200]	Time 0.361 (0.426)	Data 0.001 (0.060)	Loss 0.141 (0.241)
Epoch: [49][40/200]	Time 0.365 (0.442)	Data 0.001 (0.075)	Loss 1.322 (0.431)
Epoch: [49][60/200]	Time 0.365 (0.418)	Data 0.000 (0.050)	Loss 1.051 (0.746)
Epoch: [49][80/200]	Time 0.365 (0.428)	Data 0.001 (0.061)	Loss 1.144 (0.896)
Epoch: [49][100/200]	Time 2.087 (0.434)	Data 1.701 (0.066)	Loss 1.793 (1.004)
Epoch: [49][120/200]	Time 0.367 (0.422)	Data 0.001 (0.055)	Loss 1.476 (1.074)
Epoch: [49][140/200]	Time 0.370 (0.429)	Data 0.001 (0.061)	Loss 1.505 (1.117)
Epoch: [49][160/200]	Time 0.364 (0.421)	Data 0.000 (0.053)	Loss 1.333 (1.135)
Epoch: [49][180/200]	Time 0.368 (0.425)	Data 0.001 (0.057)	Loss 1.043 (1.163)
Epoch: [49][200/200]	Time 0.368 (0.428)	Data 0.001 (0.060)	Loss 1.670 (1.181)
Extract Features: [50/367]	Time 0.116 (0.210)	Data 0.000 (0.090)	
Extract Features: [100/367]	Time 0.119 (0.195)	Data 0.000 (0.074)	
Extract Features: [150/367]	Time 0.118 (0.192)	Data 0.000 (0.070)	
Extract Features: [200/367]	Time 0.123 (0.189)	Data 0.000 (0.068)	
Extract Features: [250/367]	Time 0.121 (0.189)	Data 0.000 (0.067)	
Extract Features: [300/367]	Time 0.118 (0.187)	Data 0.000 (0.065)	
Extract Features: [350/367]	Time 0.124 (0.186)	Data 0.000 (0.065)	
Mean AP: 62.7%

 * Finished epoch  49  model mAP: 62.7%  best: 62.7% *

==> Test with the best model:
=> Loaded checkpoint 'log/market2msmt/resnet50_ibn_cion/model_best.pth.tar'
Extract Features: [50/367]	Time 0.118 (0.208)	Data 0.001 (0.089)	
Extract Features: [100/367]	Time 0.440 (0.202)	Data 0.324 (0.082)	
Extract Features: [150/367]	Time 0.118 (0.197)	Data 0.000 (0.077)	
Extract Features: [200/367]	Time 0.349 (0.194)	Data 0.230 (0.073)	
Extract Features: [250/367]	Time 0.118 (0.191)	Data 0.000 (0.069)	
Extract Features: [300/367]	Time 0.372 (0.190)	Data 0.253 (0.069)	
Extract Features: [350/367]	Time 0.117 (0.189)	Data 0.000 (0.068)	
Mean AP: 62.7%
CMC Scores:
  top-1          84.3%
  top-5          91.2%
  top-10         93.0%
Total running time:  2:59:35.696759
