==========
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='resnet50', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet50_Market1501/64bs_lr0.0004_ep120_warm20_seed0/resnet50_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_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
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.266 (0.384)	Data 0.154 (0.075)	
Extract Features: [100/128]	Time 0.124 (0.278)	Data 0.010 (0.064)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.283554553985596
Clustering criterion: eps: 0.700
==> Statistics for epoch 0: 966 clusters
Epoch: [0][20/200]	Time 0.356 (0.773)	Data 0.001 (0.049)	Loss 3.505 (3.263)
Epoch: [0][40/200]	Time 0.353 (0.604)	Data 0.001 (0.065)	Loss 2.601 (3.146)
Epoch: [0][60/200]	Time 0.349 (0.521)	Data 0.000 (0.043)	Loss 3.273 (3.093)
Epoch: [0][80/200]	Time 0.357 (0.499)	Data 0.001 (0.053)	Loss 2.225 (2.906)
Epoch: [0][100/200]	Time 0.355 (0.487)	Data 0.001 (0.057)	Loss 2.003 (2.744)
Epoch: [0][120/200]	Time 0.351 (0.464)	Data 0.000 (0.048)	Loss 2.274 (2.650)
Epoch: [0][140/200]	Time 0.354 (0.460)	Data 0.001 (0.052)	Loss 2.065 (2.549)
Epoch: [0][160/200]	Time 0.351 (0.458)	Data 0.001 (0.056)	Loss 2.448 (2.492)
Epoch: [0][180/200]	Time 0.352 (0.446)	Data 0.000 (0.050)	Loss 2.002 (2.447)
Epoch: [0][200/200]	Time 0.355 (0.445)	Data 0.001 (0.053)	Loss 2.095 (2.398)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.431 (0.218)	Data 0.317 (0.101)	
Extract Features: [100/128]	Time 0.114 (0.196)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.818114042282104
==> Statistics for epoch 1: 1037 clusters
Epoch: [1][20/200]	Time 0.359 (0.412)	Data 0.001 (0.058)	Loss 0.405 (0.483)
Epoch: [1][40/200]	Time 0.356 (0.425)	Data 0.001 (0.071)	Loss 2.122 (0.776)
Epoch: [1][60/200]	Time 0.352 (0.402)	Data 0.000 (0.048)	Loss 1.630 (1.143)
Epoch: [1][80/200]	Time 0.356 (0.411)	Data 0.001 (0.057)	Loss 2.320 (1.359)
Epoch: [1][100/200]	Time 0.359 (0.416)	Data 0.001 (0.062)	Loss 2.134 (1.451)
Epoch: [1][120/200]	Time 0.492 (0.408)	Data 0.001 (0.051)	Loss 2.589 (1.537)
Epoch: [1][140/200]	Time 0.357 (0.412)	Data 0.001 (0.056)	Loss 1.511 (1.565)
Epoch: [1][160/200]	Time 0.353 (0.405)	Data 0.000 (0.049)	Loss 2.209 (1.602)
Epoch: [1][180/200]	Time 0.358 (0.410)	Data 0.001 (0.053)	Loss 2.034 (1.623)
Epoch: [1][200/200]	Time 0.356 (0.413)	Data 0.001 (0.056)	Loss 1.612 (1.642)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.214)	Data 0.000 (0.097)	
Extract Features: [100/128]	Time 0.114 (0.197)	Data 0.000 (0.081)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.028315782547
==> Statistics for epoch 2: 1035 clusters
Epoch: [2][20/200]	Time 0.357 (0.406)	Data 0.001 (0.049)	Loss 0.335 (0.446)
Epoch: [2][40/200]	Time 0.371 (0.421)	Data 0.002 (0.064)	Loss 1.374 (0.726)
Epoch: [2][60/200]	Time 0.351 (0.400)	Data 0.000 (0.043)	Loss 1.874 (1.137)
Epoch: [2][80/200]	Time 0.354 (0.411)	Data 0.001 (0.054)	Loss 1.593 (1.345)
Epoch: [2][100/200]	Time 0.355 (0.416)	Data 0.001 (0.059)	Loss 2.406 (1.460)
Epoch: [2][120/200]	Time 0.488 (0.407)	Data 0.001 (0.050)	Loss 1.994 (1.533)
Epoch: [2][140/200]	Time 0.355 (0.412)	Data 0.001 (0.054)	Loss 1.840 (1.591)
Epoch: [2][160/200]	Time 0.364 (0.405)	Data 0.000 (0.048)	Loss 1.722 (1.615)
Epoch: [2][180/200]	Time 0.359 (0.410)	Data 0.001 (0.052)	Loss 1.756 (1.646)
Epoch: [2][200/200]	Time 0.358 (0.412)	Data 0.001 (0.054)	Loss 2.246 (1.660)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.212)	Data 0.000 (0.094)	
Extract Features: [100/128]	Time 0.114 (0.193)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.40885496139526
==> Statistics for epoch 3: 1044 clusters
Epoch: [3][20/200]	Time 0.355 (0.408)	Data 0.001 (0.050)	Loss 0.579 (0.399)
Epoch: [3][40/200]	Time 0.362 (0.423)	Data 0.001 (0.066)	Loss 1.720 (0.667)
Epoch: [3][60/200]	Time 0.356 (0.401)	Data 0.000 (0.044)	Loss 2.473 (1.058)
Epoch: [3][80/200]	Time 0.359 (0.411)	Data 0.001 (0.054)	Loss 1.686 (1.273)
Epoch: [3][100/200]	Time 0.355 (0.417)	Data 0.001 (0.060)	Loss 1.959 (1.393)
Epoch: [3][120/200]	Time 0.357 (0.407)	Data 0.000 (0.050)	Loss 1.602 (1.473)
Epoch: [3][140/200]	Time 0.354 (0.413)	Data 0.001 (0.055)	Loss 1.876 (1.527)
Epoch: [3][160/200]	Time 0.355 (0.406)	Data 0.000 (0.048)	Loss 2.045 (1.577)
Epoch: [3][180/200]	Time 0.354 (0.409)	Data 0.001 (0.052)	Loss 1.765 (1.607)
Epoch: [3][200/200]	Time 0.361 (0.412)	Data 0.001 (0.055)	Loss 1.711 (1.635)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.217)	Data 0.000 (0.098)	
Extract Features: [100/128]	Time 0.114 (0.196)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.24957346916199
==> Statistics for epoch 4: 1024 clusters
Epoch: [4][20/200]	Time 0.359 (0.413)	Data 0.001 (0.054)	Loss 0.317 (0.418)
Epoch: [4][40/200]	Time 0.357 (0.425)	Data 0.000 (0.067)	Loss 1.601 (0.654)
Epoch: [4][60/200]	Time 0.357 (0.402)	Data 0.000 (0.045)	Loss 1.945 (1.077)
Epoch: [4][80/200]	Time 0.356 (0.412)	Data 0.001 (0.053)	Loss 1.822 (1.259)
Epoch: [4][100/200]	Time 0.356 (0.417)	Data 0.001 (0.059)	Loss 2.019 (1.370)
Epoch: [4][120/200]	Time 0.358 (0.407)	Data 0.001 (0.049)	Loss 1.869 (1.449)
Epoch: [4][140/200]	Time 0.355 (0.412)	Data 0.001 (0.054)	Loss 2.082 (1.504)
Epoch: [4][160/200]	Time 0.357 (0.406)	Data 0.000 (0.047)	Loss 1.525 (1.534)
Epoch: [4][180/200]	Time 0.356 (0.409)	Data 0.001 (0.051)	Loss 2.192 (1.570)
Epoch: [4][200/200]	Time 0.352 (0.412)	Data 0.001 (0.054)	Loss 1.819 (1.585)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.221)	Data 0.000 (0.104)	
Extract Features: [100/128]	Time 0.116 (0.198)	Data 0.000 (0.082)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.19313097000122
==> Statistics for epoch 5: 1028 clusters
Epoch: [5][20/200]	Time 0.357 (0.411)	Data 0.001 (0.054)	Loss 0.347 (0.418)
Epoch: [5][40/200]	Time 0.358 (0.432)	Data 0.001 (0.072)	Loss 2.470 (0.686)
Epoch: [5][60/200]	Time 0.352 (0.407)	Data 0.000 (0.048)	Loss 1.905 (1.075)
Epoch: [5][80/200]	Time 0.356 (0.416)	Data 0.001 (0.058)	Loss 2.011 (1.266)
Epoch: [5][100/200]	Time 0.354 (0.420)	Data 0.001 (0.063)	Loss 1.837 (1.382)
Epoch: [5][120/200]	Time 0.357 (0.409)	Data 0.001 (0.052)	Loss 1.769 (1.453)
Epoch: [5][140/200]	Time 0.356 (0.414)	Data 0.001 (0.057)	Loss 2.576 (1.519)
Epoch: [5][160/200]	Time 0.354 (0.406)	Data 0.000 (0.050)	Loss 1.542 (1.555)
Epoch: [5][180/200]	Time 0.354 (0.411)	Data 0.001 (0.053)	Loss 1.528 (1.586)
Epoch: [5][200/200]	Time 0.360 (0.414)	Data 0.001 (0.057)	Loss 1.770 (1.608)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.219)	Data 0.000 (0.105)	
Extract Features: [100/128]	Time 0.115 (0.197)	Data 0.000 (0.080)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.77540898323059
==> Statistics for epoch 6: 1016 clusters
Epoch: [6][20/200]	Time 0.352 (0.410)	Data 0.001 (0.054)	Loss 0.594 (0.417)
Epoch: [6][40/200]	Time 0.355 (0.425)	Data 0.001 (0.070)	Loss 1.701 (0.725)
Epoch: [6][60/200]	Time 0.350 (0.402)	Data 0.000 (0.047)	Loss 1.931 (1.087)
Epoch: [6][80/200]	Time 0.361 (0.411)	Data 0.001 (0.055)	Loss 1.251 (1.261)
Epoch: [6][100/200]	Time 0.355 (0.417)	Data 0.001 (0.061)	Loss 2.284 (1.381)
Epoch: [6][120/200]	Time 0.357 (0.407)	Data 0.000 (0.051)	Loss 2.409 (1.460)
Epoch: [6][140/200]	Time 0.357 (0.414)	Data 0.001 (0.057)	Loss 1.857 (1.508)
Epoch: [6][160/200]	Time 0.355 (0.417)	Data 0.001 (0.060)	Loss 1.996 (1.544)
Epoch: [6][180/200]	Time 0.357 (0.410)	Data 0.000 (0.053)	Loss 1.645 (1.579)
Epoch: [6][200/200]	Time 0.359 (0.413)	Data 0.001 (0.056)	Loss 1.563 (1.604)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.212)	Data 0.000 (0.094)	
Extract Features: [100/128]	Time 0.113 (0.191)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.242809534072876
==> Statistics for epoch 7: 997 clusters
Epoch: [7][20/200]	Time 0.358 (0.413)	Data 0.001 (0.047)	Loss 0.539 (0.352)
Epoch: [7][40/200]	Time 0.353 (0.422)	Data 0.001 (0.062)	Loss 1.495 (0.625)
Epoch: [7][60/200]	Time 0.350 (0.399)	Data 0.000 (0.041)	Loss 1.512 (0.987)
Epoch: [7][80/200]	Time 0.354 (0.409)	Data 0.001 (0.052)	Loss 1.767 (1.188)
Epoch: [7][100/200]	Time 0.362 (0.416)	Data 0.001 (0.058)	Loss 1.877 (1.293)
Epoch: [7][120/200]	Time 0.354 (0.406)	Data 0.000 (0.049)	Loss 2.010 (1.377)
Epoch: [7][140/200]	Time 0.361 (0.412)	Data 0.001 (0.053)	Loss 1.743 (1.430)
Epoch: [7][160/200]	Time 0.358 (0.415)	Data 0.000 (0.057)	Loss 1.214 (1.464)
Epoch: [7][180/200]	Time 0.355 (0.409)	Data 0.000 (0.051)	Loss 1.908 (1.503)
Epoch: [7][200/200]	Time 0.356 (0.412)	Data 0.001 (0.054)	Loss 1.880 (1.517)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.218)	Data 0.000 (0.103)	
Extract Features: [100/128]	Time 0.117 (0.195)	Data 0.001 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.02547097206116
==> Statistics for epoch 8: 1023 clusters
Epoch: [8][20/200]	Time 0.353 (0.411)	Data 0.001 (0.054)	Loss 0.583 (0.376)
Epoch: [8][40/200]	Time 0.355 (0.428)	Data 0.001 (0.071)	Loss 1.535 (0.648)
Epoch: [8][60/200]	Time 0.353 (0.404)	Data 0.000 (0.047)	Loss 1.499 (1.009)
Epoch: [8][80/200]	Time 0.507 (0.414)	Data 0.001 (0.056)	Loss 1.914 (1.219)
Epoch: [8][100/200]	Time 0.354 (0.418)	Data 0.001 (0.061)	Loss 2.133 (1.341)
Epoch: [8][120/200]	Time 0.358 (0.408)	Data 0.000 (0.051)	Loss 1.851 (1.420)
Epoch: [8][140/200]	Time 0.364 (0.413)	Data 0.003 (0.055)	Loss 2.000 (1.476)
Epoch: [8][160/200]	Time 0.358 (0.416)	Data 0.001 (0.059)	Loss 1.302 (1.506)
Epoch: [8][180/200]	Time 0.358 (0.410)	Data 0.000 (0.052)	Loss 1.914 (1.544)
Epoch: [8][200/200]	Time 0.364 (0.413)	Data 0.001 (0.056)	Loss 2.043 (1.565)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.217)	Data 0.000 (0.099)	
Extract Features: [100/128]	Time 0.114 (0.196)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.96078848838806
==> Statistics for epoch 9: 1032 clusters
Epoch: [9][20/200]	Time 0.356 (0.414)	Data 0.001 (0.056)	Loss 0.436 (0.478)
Epoch: [9][40/200]	Time 0.355 (0.423)	Data 0.001 (0.067)	Loss 1.537 (0.708)
Epoch: [9][60/200]	Time 0.353 (0.401)	Data 0.000 (0.045)	Loss 1.594 (1.029)
Epoch: [9][80/200]	Time 0.356 (0.409)	Data 0.001 (0.052)	Loss 1.448 (1.158)
Epoch: [9][100/200]	Time 0.362 (0.415)	Data 0.001 (0.058)	Loss 1.591 (1.235)
Epoch: [9][120/200]	Time 0.354 (0.405)	Data 0.001 (0.049)	Loss 1.944 (1.298)
Epoch: [9][140/200]	Time 0.353 (0.411)	Data 0.001 (0.053)	Loss 1.285 (1.327)
Epoch: [9][160/200]	Time 0.353 (0.404)	Data 0.000 (0.047)	Loss 1.629 (1.357)
Epoch: [9][180/200]	Time 0.359 (0.408)	Data 0.001 (0.051)	Loss 1.423 (1.376)
Epoch: [9][200/200]	Time 0.356 (0.411)	Data 0.001 (0.054)	Loss 1.560 (1.392)
Extract Features: [50/367]	Time 0.494 (0.220)	Data 0.380 (0.103)	
Extract Features: [100/367]	Time 0.121 (0.203)	Data 0.001 (0.086)	
Extract Features: [150/367]	Time 0.458 (0.198)	Data 0.343 (0.081)	
Extract Features: [200/367]	Time 0.117 (0.194)	Data 0.000 (0.077)	
Extract Features: [250/367]	Time 0.378 (0.192)	Data 0.257 (0.074)	
Extract Features: [300/367]	Time 0.120 (0.190)	Data 0.000 (0.073)	
Extract Features: [350/367]	Time 0.395 (0.189)	Data 0.278 (0.072)	
Mean AP: 44.3%

 * Finished epoch   9  model mAP: 44.3%  best: 44.3% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.218)	Data 0.000 (0.103)	
Extract Features: [100/128]	Time 0.116 (0.198)	Data 0.001 (0.083)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.68933129310608
==> Statistics for epoch 10: 1030 clusters
Epoch: [10][20/200]	Time 0.360 (0.414)	Data 0.001 (0.056)	Loss 0.297 (0.376)
Epoch: [10][40/200]	Time 0.354 (0.432)	Data 0.001 (0.076)	Loss 1.771 (0.650)
Epoch: [10][60/200]	Time 0.355 (0.406)	Data 0.000 (0.051)	Loss 1.611 (0.979)
Epoch: [10][80/200]	Time 0.355 (0.418)	Data 0.001 (0.062)	Loss 1.191 (1.153)
Epoch: [10][100/200]	Time 0.356 (0.423)	Data 0.001 (0.068)	Loss 1.786 (1.276)
Epoch: [10][120/200]	Time 0.354 (0.413)	Data 0.001 (0.057)	Loss 2.103 (1.355)
Epoch: [10][140/200]	Time 0.356 (0.419)	Data 0.001 (0.062)	Loss 1.301 (1.404)
Epoch: [10][160/200]	Time 0.355 (0.411)	Data 0.000 (0.055)	Loss 1.543 (1.443)
Epoch: [10][180/200]	Time 0.353 (0.415)	Data 0.001 (0.059)	Loss 1.428 (1.461)
Epoch: [10][200/200]	Time 0.355 (0.419)	Data 0.001 (0.063)	Loss 1.965 (1.489)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.135 (0.209)	Data 0.021 (0.090)	
Extract Features: [100/128]	Time 0.116 (0.191)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.767176151275635
==> Statistics for epoch 11: 1001 clusters
Epoch: [11][20/200]	Time 0.351 (0.412)	Data 0.001 (0.058)	Loss 0.309 (0.351)
Epoch: [11][40/200]	Time 0.355 (0.427)	Data 0.001 (0.073)	Loss 1.595 (0.609)
Epoch: [11][60/200]	Time 0.349 (0.405)	Data 0.000 (0.049)	Loss 1.624 (0.952)
Epoch: [11][80/200]	Time 0.375 (0.415)	Data 0.001 (0.058)	Loss 2.158 (1.122)
Epoch: [11][100/200]	Time 0.352 (0.421)	Data 0.001 (0.064)	Loss 1.698 (1.239)
Epoch: [11][120/200]	Time 0.350 (0.410)	Data 0.000 (0.054)	Loss 1.829 (1.314)
Epoch: [11][140/200]	Time 0.358 (0.416)	Data 0.001 (0.059)	Loss 2.252 (1.367)
Epoch: [11][160/200]	Time 0.354 (0.420)	Data 0.001 (0.063)	Loss 1.760 (1.403)
Epoch: [11][180/200]	Time 0.352 (0.413)	Data 0.000 (0.056)	Loss 1.435 (1.431)
Epoch: [11][200/200]	Time 0.366 (0.416)	Data 0.001 (0.059)	Loss 1.670 (1.452)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.256 (0.214)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.118 (0.193)	Data 0.000 (0.076)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.923869371414185
==> Statistics for epoch 12: 1014 clusters
Epoch: [12][20/200]	Time 0.350 (0.415)	Data 0.001 (0.056)	Loss 0.350 (0.370)
Epoch: [12][40/200]	Time 0.354 (0.430)	Data 0.001 (0.073)	Loss 1.632 (0.600)
Epoch: [12][60/200]	Time 0.355 (0.407)	Data 0.000 (0.049)	Loss 1.298 (0.937)
Epoch: [12][80/200]	Time 0.356 (0.417)	Data 0.001 (0.059)	Loss 1.621 (1.095)
Epoch: [12][100/200]	Time 0.350 (0.422)	Data 0.001 (0.064)	Loss 1.378 (1.182)
Epoch: [12][120/200]	Time 0.353 (0.411)	Data 0.000 (0.054)	Loss 1.624 (1.238)
Epoch: [12][140/200]	Time 0.358 (0.417)	Data 0.001 (0.060)	Loss 1.699 (1.279)
Epoch: [12][160/200]	Time 0.366 (0.420)	Data 0.001 (0.063)	Loss 1.451 (1.319)
Epoch: [12][180/200]	Time 0.355 (0.413)	Data 0.000 (0.056)	Loss 1.663 (1.352)
Epoch: [12][200/200]	Time 0.354 (0.418)	Data 0.001 (0.060)	Loss 2.094 (1.369)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.214)	Data 0.000 (0.099)	
Extract Features: [100/128]	Time 0.116 (0.192)	Data 0.000 (0.076)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.669326305389404
==> Statistics for epoch 13: 1013 clusters
Epoch: [13][20/200]	Time 0.361 (0.406)	Data 0.001 (0.051)	Loss 0.220 (0.363)
Epoch: [13][40/200]	Time 0.352 (0.426)	Data 0.001 (0.071)	Loss 1.742 (0.642)
Epoch: [13][60/200]	Time 0.353 (0.402)	Data 0.000 (0.047)	Loss 1.586 (0.932)
Epoch: [13][80/200]	Time 0.355 (0.413)	Data 0.001 (0.057)	Loss 1.890 (1.100)
Epoch: [13][100/200]	Time 0.353 (0.420)	Data 0.001 (0.064)	Loss 1.261 (1.214)
Epoch: [13][120/200]	Time 0.356 (0.409)	Data 0.000 (0.054)	Loss 1.474 (1.295)
Epoch: [13][140/200]	Time 0.362 (0.416)	Data 0.001 (0.060)	Loss 2.118 (1.346)
Epoch: [13][160/200]	Time 0.360 (0.420)	Data 0.001 (0.064)	Loss 1.583 (1.378)
Epoch: [13][180/200]	Time 0.352 (0.413)	Data 0.000 (0.057)	Loss 1.836 (1.409)
Epoch: [13][200/200]	Time 0.357 (0.417)	Data 0.001 (0.060)	Loss 1.155 (1.436)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.113 (0.216)	Data 0.000 (0.098)	
Extract Features: [100/128]	Time 0.117 (0.196)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.38001227378845
==> Statistics for epoch 14: 1020 clusters
Epoch: [14][20/200]	Time 0.353 (0.417)	Data 0.000 (0.052)	Loss 0.358 (0.330)
Epoch: [14][40/200]	Time 0.355 (0.434)	Data 0.001 (0.074)	Loss 2.004 (0.617)
Epoch: [14][60/200]	Time 0.352 (0.408)	Data 0.000 (0.049)	Loss 1.723 (0.914)
Epoch: [14][80/200]	Time 0.353 (0.417)	Data 0.001 (0.059)	Loss 1.681 (1.100)
Epoch: [14][100/200]	Time 0.352 (0.424)	Data 0.000 (0.067)	Loss 1.481 (1.216)
Epoch: [14][120/200]	Time 0.352 (0.413)	Data 0.000 (0.056)	Loss 1.799 (1.287)
Epoch: [14][140/200]	Time 0.360 (0.417)	Data 0.000 (0.060)	Loss 1.815 (1.336)
Epoch: [14][160/200]	Time 0.354 (0.421)	Data 0.000 (0.063)	Loss 0.991 (1.357)
Epoch: [14][180/200]	Time 0.356 (0.414)	Data 0.000 (0.056)	Loss 1.721 (1.384)
Epoch: [14][200/200]	Time 0.358 (0.417)	Data 0.000 (0.059)	Loss 1.536 (1.402)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.118 (0.210)	Data 0.000 (0.092)	
Extract Features: [100/128]	Time 0.114 (0.190)	Data 0.000 (0.073)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.94916892051697
==> Statistics for epoch 15: 1012 clusters
Epoch: [15][20/200]	Time 0.353 (0.411)	Data 0.000 (0.049)	Loss 0.261 (0.340)
Epoch: [15][40/200]	Time 0.353 (0.428)	Data 0.001 (0.070)	Loss 1.399 (0.602)
Epoch: [15][60/200]	Time 0.354 (0.404)	Data 0.000 (0.047)	Loss 1.526 (0.936)
Epoch: [15][80/200]	Time 0.361 (0.415)	Data 0.001 (0.058)	Loss 2.028 (1.116)
Epoch: [15][100/200]	Time 0.353 (0.421)	Data 0.001 (0.065)	Loss 1.983 (1.235)
Epoch: [15][120/200]	Time 0.354 (0.410)	Data 0.000 (0.054)	Loss 1.686 (1.305)
Epoch: [15][140/200]	Time 0.366 (0.415)	Data 0.001 (0.059)	Loss 1.780 (1.349)
Epoch: [15][160/200]	Time 0.365 (0.421)	Data 0.003 (0.063)	Loss 1.643 (1.394)
Epoch: [15][180/200]	Time 0.355 (0.414)	Data 0.000 (0.056)	Loss 1.685 (1.426)
Epoch: [15][200/200]	Time 0.360 (0.417)	Data 0.001 (0.059)	Loss 1.815 (1.446)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.214)	Data 0.000 (0.099)	
Extract Features: [100/128]	Time 0.115 (0.195)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.22472310066223
==> Statistics for epoch 16: 1019 clusters
Epoch: [16][20/200]	Time 0.351 (0.414)	Data 0.001 (0.059)	Loss 0.444 (0.331)
Epoch: [16][40/200]	Time 0.357 (0.429)	Data 0.001 (0.074)	Loss 1.850 (0.609)
Epoch: [16][60/200]	Time 0.353 (0.405)	Data 0.000 (0.049)	Loss 1.568 (0.925)
Epoch: [16][80/200]	Time 0.358 (0.415)	Data 0.001 (0.059)	Loss 1.625 (1.086)
Epoch: [16][100/200]	Time 0.359 (0.423)	Data 0.001 (0.067)	Loss 1.655 (1.200)
Epoch: [16][120/200]	Time 0.355 (0.413)	Data 0.000 (0.056)	Loss 1.773 (1.268)
Epoch: [16][140/200]	Time 0.354 (0.418)	Data 0.001 (0.061)	Loss 1.762 (1.319)
Epoch: [16][160/200]	Time 0.361 (0.423)	Data 0.001 (0.066)	Loss 2.026 (1.357)
Epoch: [16][180/200]	Time 0.357 (0.415)	Data 0.000 (0.059)	Loss 1.778 (1.376)
Epoch: [16][200/200]	Time 0.356 (0.419)	Data 0.001 (0.062)	Loss 1.312 (1.398)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.158 (0.209)	Data 0.042 (0.091)	
Extract Features: [100/128]	Time 0.118 (0.191)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.138566732406616
==> Statistics for epoch 17: 1012 clusters
Epoch: [17][20/200]	Time 0.352 (0.408)	Data 0.001 (0.052)	Loss 0.302 (0.346)
Epoch: [17][40/200]	Time 0.354 (0.428)	Data 0.001 (0.072)	Loss 1.665 (0.633)
Epoch: [17][60/200]	Time 0.357 (0.404)	Data 0.000 (0.048)	Loss 1.636 (0.950)
Epoch: [17][80/200]	Time 0.363 (0.415)	Data 0.001 (0.059)	Loss 1.499 (1.080)
Epoch: [17][100/200]	Time 0.355 (0.420)	Data 0.001 (0.064)	Loss 1.658 (1.179)
Epoch: [17][120/200]	Time 0.353 (0.409)	Data 0.000 (0.053)	Loss 1.913 (1.246)
Epoch: [17][140/200]	Time 0.355 (0.415)	Data 0.001 (0.058)	Loss 1.291 (1.292)
Epoch: [17][160/200]	Time 0.353 (0.418)	Data 0.001 (0.062)	Loss 1.606 (1.331)
Epoch: [17][180/200]	Time 0.351 (0.411)	Data 0.000 (0.055)	Loss 1.999 (1.348)
Epoch: [17][200/200]	Time 0.355 (0.415)	Data 0.001 (0.058)	Loss 1.519 (1.373)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.215)	Data 0.000 (0.101)	
Extract Features: [100/128]	Time 0.115 (0.193)	Data 0.000 (0.076)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.93231463432312
==> Statistics for epoch 18: 1013 clusters
Epoch: [18][20/200]	Time 0.351 (0.409)	Data 0.001 (0.053)	Loss 0.282 (0.271)
Epoch: [18][40/200]	Time 0.360 (0.432)	Data 0.001 (0.072)	Loss 1.316 (0.592)
Epoch: [18][60/200]	Time 0.356 (0.406)	Data 0.000 (0.048)	Loss 1.365 (0.915)
Epoch: [18][80/200]	Time 0.361 (0.417)	Data 0.001 (0.059)	Loss 1.537 (1.052)
Epoch: [18][100/200]	Time 0.358 (0.423)	Data 0.001 (0.066)	Loss 1.649 (1.173)
Epoch: [18][120/200]	Time 0.358 (0.412)	Data 0.001 (0.055)	Loss 1.549 (1.248)
Epoch: [18][140/200]	Time 0.355 (0.417)	Data 0.001 (0.060)	Loss 1.806 (1.283)
Epoch: [18][160/200]	Time 0.364 (0.421)	Data 0.001 (0.064)	Loss 1.596 (1.336)
Epoch: [18][180/200]	Time 0.352 (0.415)	Data 0.000 (0.057)	Loss 1.543 (1.361)
Epoch: [18][200/200]	Time 0.356 (0.419)	Data 0.001 (0.061)	Loss 2.063 (1.391)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.216)	Data 0.000 (0.101)	
Extract Features: [100/128]	Time 0.115 (0.193)	Data 0.000 (0.077)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.18015646934509
==> Statistics for epoch 19: 1013 clusters
Epoch: [19][20/200]	Time 0.360 (0.409)	Data 0.001 (0.052)	Loss 0.396 (0.324)
Epoch: [19][40/200]	Time 0.356 (0.432)	Data 0.001 (0.073)	Loss 1.251 (0.595)
Epoch: [19][60/200]	Time 0.353 (0.407)	Data 0.000 (0.049)	Loss 1.620 (0.918)
Epoch: [19][80/200]	Time 0.356 (0.416)	Data 0.001 (0.059)	Loss 1.609 (1.074)
Epoch: [19][100/200]	Time 0.354 (0.423)	Data 0.001 (0.067)	Loss 1.044 (1.157)
Epoch: [19][120/200]	Time 0.355 (0.412)	Data 0.000 (0.056)	Loss 1.414 (1.224)
Epoch: [19][140/200]	Time 0.355 (0.418)	Data 0.001 (0.061)	Loss 1.616 (1.281)
Epoch: [19][160/200]	Time 0.360 (0.422)	Data 0.001 (0.065)	Loss 1.348 (1.314)
Epoch: [19][180/200]	Time 0.354 (0.414)	Data 0.000 (0.058)	Loss 1.454 (1.341)
Epoch: [19][200/200]	Time 0.358 (0.419)	Data 0.001 (0.062)	Loss 1.864 (1.370)
Extract Features: [50/367]	Time 0.238 (0.214)	Data 0.123 (0.099)	
Extract Features: [100/367]	Time 0.117 (0.197)	Data 0.000 (0.081)	
Extract Features: [150/367]	Time 0.237 (0.192)	Data 0.122 (0.077)	
Extract Features: [200/367]	Time 0.117 (0.189)	Data 0.000 (0.072)	
Extract Features: [250/367]	Time 0.123 (0.187)	Data 0.007 (0.071)	
Extract Features: [300/367]	Time 0.120 (0.186)	Data 0.001 (0.069)	
Extract Features: [350/367]	Time 0.204 (0.185)	Data 0.089 (0.069)	
Mean AP: 51.7%

 * Finished epoch  19  model mAP: 51.7%  best: 51.7% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.214)	Data 0.000 (0.100)	
Extract Features: [100/128]	Time 0.120 (0.195)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.25855207443237
==> Statistics for epoch 20: 1033 clusters
Epoch: [20][20/200]	Time 0.350 (0.406)	Data 0.001 (0.051)	Loss 0.244 (0.326)
Epoch: [20][40/200]	Time 0.358 (0.426)	Data 0.001 (0.071)	Loss 1.639 (0.561)
Epoch: [20][60/200]	Time 0.353 (0.404)	Data 0.000 (0.048)	Loss 1.639 (0.885)
Epoch: [20][80/200]	Time 0.359 (0.414)	Data 0.001 (0.058)	Loss 1.795 (1.037)
Epoch: [20][100/200]	Time 0.371 (0.421)	Data 0.001 (0.065)	Loss 1.354 (1.151)
Epoch: [20][120/200]	Time 0.363 (0.410)	Data 0.001 (0.054)	Loss 1.740 (1.203)
Epoch: [20][140/200]	Time 0.355 (0.415)	Data 0.001 (0.060)	Loss 1.448 (1.258)
Epoch: [20][160/200]	Time 0.354 (0.408)	Data 0.000 (0.052)	Loss 1.788 (1.307)
Epoch: [20][180/200]	Time 0.366 (0.412)	Data 0.001 (0.056)	Loss 1.339 (1.335)
Epoch: [20][200/200]	Time 0.357 (0.416)	Data 0.001 (0.060)	Loss 1.912 (1.362)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.526 (0.218)	Data 0.412 (0.100)	
Extract Features: [100/128]	Time 0.114 (0.197)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.49473237991333
==> Statistics for epoch 21: 1031 clusters
Epoch: [21][20/200]	Time 0.354 (0.408)	Data 0.001 (0.051)	Loss 0.333 (0.306)
Epoch: [21][40/200]	Time 0.357 (0.425)	Data 0.001 (0.069)	Loss 1.588 (0.549)
Epoch: [21][60/200]	Time 0.360 (0.402)	Data 0.000 (0.046)	Loss 1.587 (0.859)
Epoch: [21][80/200]	Time 0.361 (0.415)	Data 0.001 (0.058)	Loss 2.013 (1.042)
Epoch: [21][100/200]	Time 0.359 (0.423)	Data 0.001 (0.065)	Loss 1.322 (1.149)
Epoch: [21][120/200]	Time 0.354 (0.412)	Data 0.001 (0.054)	Loss 2.118 (1.218)
Epoch: [21][140/200]	Time 0.356 (0.417)	Data 0.001 (0.060)	Loss 2.075 (1.277)
Epoch: [21][160/200]	Time 0.354 (0.410)	Data 0.000 (0.052)	Loss 1.519 (1.323)
Epoch: [21][180/200]	Time 0.360 (0.414)	Data 0.001 (0.056)	Loss 1.556 (1.351)
Epoch: [21][200/200]	Time 0.354 (0.417)	Data 0.001 (0.060)	Loss 1.391 (1.380)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.112 (0.217)	Data 0.000 (0.098)	
Extract Features: [100/128]	Time 0.118 (0.197)	Data 0.000 (0.076)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.68749761581421
==> Statistics for epoch 22: 1024 clusters
Epoch: [22][20/200]	Time 0.352 (0.406)	Data 0.001 (0.050)	Loss 0.203 (0.301)
Epoch: [22][40/200]	Time 0.363 (0.425)	Data 0.001 (0.070)	Loss 1.474 (0.547)
Epoch: [22][60/200]	Time 0.351 (0.401)	Data 0.000 (0.047)	Loss 1.546 (0.886)
Epoch: [22][80/200]	Time 0.355 (0.412)	Data 0.001 (0.057)	Loss 1.664 (1.042)
Epoch: [22][100/200]	Time 0.356 (0.418)	Data 0.001 (0.063)	Loss 1.256 (1.129)
Epoch: [22][120/200]	Time 0.360 (0.408)	Data 0.001 (0.053)	Loss 1.793 (1.206)
Epoch: [22][140/200]	Time 0.357 (0.413)	Data 0.001 (0.058)	Loss 1.845 (1.248)
Epoch: [22][160/200]	Time 0.352 (0.406)	Data 0.000 (0.051)	Loss 1.279 (1.291)
Epoch: [22][180/200]	Time 0.359 (0.410)	Data 0.001 (0.055)	Loss 1.615 (1.322)
Epoch: [22][200/200]	Time 0.376 (0.414)	Data 0.001 (0.058)	Loss 1.895 (1.342)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.213)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.119 (0.197)	Data 0.000 (0.080)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.44688677787781
==> Statistics for epoch 23: 1031 clusters
Epoch: [23][20/200]	Time 0.354 (0.410)	Data 0.001 (0.053)	Loss 0.282 (0.267)
Epoch: [23][40/200]	Time 0.358 (0.427)	Data 0.001 (0.071)	Loss 1.320 (0.505)
Epoch: [23][60/200]	Time 0.352 (0.403)	Data 0.000 (0.047)	Loss 1.432 (0.860)
Epoch: [23][80/200]	Time 0.351 (0.415)	Data 0.001 (0.059)	Loss 2.191 (1.022)
Epoch: [23][100/200]	Time 0.354 (0.422)	Data 0.001 (0.066)	Loss 1.694 (1.127)
Epoch: [23][120/200]	Time 0.356 (0.411)	Data 0.001 (0.055)	Loss 1.649 (1.192)
Epoch: [23][140/200]	Time 0.356 (0.416)	Data 0.001 (0.060)	Loss 1.683 (1.252)
Epoch: [23][160/200]	Time 0.355 (0.408)	Data 0.000 (0.053)	Loss 1.783 (1.285)
Epoch: [23][180/200]	Time 0.355 (0.413)	Data 0.001 (0.057)	Loss 1.150 (1.310)
Epoch: [23][200/200]	Time 0.360 (0.417)	Data 0.001 (0.060)	Loss 1.628 (1.341)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.223 (0.210)	Data 0.110 (0.090)	
Extract Features: [100/128]	Time 0.114 (0.192)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.49101758003235
==> Statistics for epoch 24: 1015 clusters
Epoch: [24][20/200]	Time 0.355 (0.413)	Data 0.001 (0.050)	Loss 0.267 (0.293)
Epoch: [24][40/200]	Time 0.363 (0.431)	Data 0.001 (0.072)	Loss 1.099 (0.529)
Epoch: [24][60/200]	Time 0.353 (0.405)	Data 0.000 (0.048)	Loss 1.474 (0.865)
Epoch: [24][80/200]	Time 0.358 (0.416)	Data 0.001 (0.059)	Loss 1.346 (1.011)
Epoch: [24][100/200]	Time 0.355 (0.421)	Data 0.001 (0.065)	Loss 1.875 (1.134)
Epoch: [24][120/200]	Time 0.353 (0.410)	Data 0.000 (0.054)	Loss 1.630 (1.202)
Epoch: [24][140/200]	Time 0.363 (0.415)	Data 0.001 (0.059)	Loss 1.868 (1.244)
Epoch: [24][160/200]	Time 0.355 (0.420)	Data 0.001 (0.063)	Loss 1.387 (1.282)
Epoch: [24][180/200]	Time 0.354 (0.413)	Data 0.000 (0.056)	Loss 1.516 (1.306)
Epoch: [24][200/200]	Time 0.354 (0.417)	Data 0.001 (0.059)	Loss 2.051 (1.332)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.243 (0.214)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.116 (0.191)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.62275767326355
==> Statistics for epoch 25: 1022 clusters
Epoch: [25][20/200]	Time 0.357 (0.413)	Data 0.001 (0.057)	Loss 0.260 (0.307)
Epoch: [25][40/200]	Time 0.360 (0.432)	Data 0.001 (0.073)	Loss 1.430 (0.586)
Epoch: [25][60/200]	Time 0.349 (0.406)	Data 0.000 (0.049)	Loss 1.635 (0.862)
Epoch: [25][80/200]	Time 0.355 (0.416)	Data 0.001 (0.059)	Loss 1.580 (1.044)
Epoch: [25][100/200]	Time 0.358 (0.423)	Data 0.001 (0.066)	Loss 1.142 (1.137)
Epoch: [25][120/200]	Time 0.350 (0.412)	Data 0.000 (0.055)	Loss 1.492 (1.221)
Epoch: [25][140/200]	Time 0.353 (0.417)	Data 0.001 (0.060)	Loss 1.664 (1.285)
Epoch: [25][160/200]	Time 0.356 (0.421)	Data 0.001 (0.064)	Loss 1.020 (1.305)
Epoch: [25][180/200]	Time 0.354 (0.414)	Data 0.000 (0.057)	Loss 1.375 (1.336)
Epoch: [25][200/200]	Time 0.357 (0.417)	Data 0.001 (0.060)	Loss 1.300 (1.357)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.214)	Data 0.000 (0.099)	
Extract Features: [100/128]	Time 0.116 (0.194)	Data 0.000 (0.077)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.10548400878906
==> Statistics for epoch 26: 1037 clusters
Epoch: [26][20/200]	Time 0.355 (0.411)	Data 0.001 (0.056)	Loss 0.250 (0.271)
Epoch: [26][40/200]	Time 0.357 (0.429)	Data 0.001 (0.074)	Loss 1.287 (0.530)
Epoch: [26][60/200]	Time 0.357 (0.404)	Data 0.000 (0.049)	Loss 1.281 (0.864)
Epoch: [26][80/200]	Time 0.357 (0.415)	Data 0.001 (0.060)	Loss 1.365 (1.031)
Epoch: [26][100/200]	Time 0.353 (0.422)	Data 0.001 (0.066)	Loss 1.614 (1.138)
Epoch: [26][120/200]	Time 0.355 (0.411)	Data 0.001 (0.055)	Loss 1.369 (1.211)
Epoch: [26][140/200]	Time 0.356 (0.416)	Data 0.001 (0.060)	Loss 1.502 (1.254)
Epoch: [26][160/200]	Time 0.353 (0.409)	Data 0.000 (0.052)	Loss 1.268 (1.296)
Epoch: [26][180/200]	Time 0.358 (0.413)	Data 0.001 (0.057)	Loss 1.855 (1.324)
Epoch: [26][200/200]	Time 0.362 (0.417)	Data 0.001 (0.060)	Loss 1.967 (1.348)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.288 (0.212)	Data 0.168 (0.095)	
Extract Features: [100/128]	Time 0.116 (0.191)	Data 0.000 (0.074)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.51763129234314
==> Statistics for epoch 27: 1038 clusters
Epoch: [27][20/200]	Time 0.355 (0.411)	Data 0.001 (0.055)	Loss 0.361 (0.287)
Epoch: [27][40/200]	Time 0.355 (0.429)	Data 0.001 (0.073)	Loss 1.654 (0.511)
Epoch: [27][60/200]	Time 0.356 (0.404)	Data 0.000 (0.049)	Loss 1.436 (0.855)
Epoch: [27][80/200]	Time 0.354 (0.414)	Data 0.000 (0.059)	Loss 1.227 (1.007)
Epoch: [27][100/200]	Time 0.355 (0.422)	Data 0.001 (0.066)	Loss 1.327 (1.114)
Epoch: [27][120/200]	Time 0.364 (0.412)	Data 0.002 (0.055)	Loss 1.491 (1.179)
Epoch: [27][140/200]	Time 0.356 (0.417)	Data 0.001 (0.060)	Loss 1.511 (1.226)
Epoch: [27][160/200]	Time 0.354 (0.409)	Data 0.000 (0.052)	Loss 1.679 (1.253)
Epoch: [27][180/200]	Time 0.360 (0.413)	Data 0.001 (0.056)	Loss 1.209 (1.295)
Epoch: [27][200/200]	Time 0.355 (0.417)	Data 0.001 (0.060)	Loss 1.354 (1.323)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.117 (0.215)	Data 0.000 (0.097)	
Extract Features: [100/128]	Time 0.125 (0.197)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.63883996009827
==> Statistics for epoch 28: 1024 clusters
Epoch: [28][20/200]	Time 0.356 (0.418)	Data 0.001 (0.059)	Loss 0.318 (0.262)
Epoch: [28][40/200]	Time 0.358 (0.429)	Data 0.002 (0.072)	Loss 1.422 (0.500)
Epoch: [28][60/200]	Time 0.352 (0.405)	Data 0.000 (0.048)	Loss 1.250 (0.820)
Epoch: [28][80/200]	Time 0.353 (0.415)	Data 0.001 (0.058)	Loss 1.485 (1.006)
Epoch: [28][100/200]	Time 0.356 (0.422)	Data 0.001 (0.066)	Loss 1.508 (1.106)
Epoch: [28][120/200]	Time 0.355 (0.412)	Data 0.001 (0.055)	Loss 1.259 (1.169)
Epoch: [28][140/200]	Time 0.356 (0.417)	Data 0.001 (0.060)	Loss 1.059 (1.225)
Epoch: [28][160/200]	Time 0.356 (0.410)	Data 0.000 (0.053)	Loss 1.308 (1.265)
Epoch: [28][180/200]	Time 0.358 (0.414)	Data 0.001 (0.057)	Loss 1.397 (1.295)
Epoch: [28][200/200]	Time 0.357 (0.418)	Data 0.001 (0.060)	Loss 1.747 (1.316)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.210)	Data 0.000 (0.094)	
Extract Features: [100/128]	Time 0.116 (0.195)	Data 0.000 (0.077)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.753411293029785
==> Statistics for epoch 29: 1027 clusters
Epoch: [29][20/200]	Time 0.354 (0.412)	Data 0.001 (0.056)	Loss 0.381 (0.286)
Epoch: [29][40/200]	Time 0.492 (0.429)	Data 0.001 (0.070)	Loss 1.933 (0.539)
Epoch: [29][60/200]	Time 0.353 (0.404)	Data 0.000 (0.047)	Loss 1.548 (0.868)
Epoch: [29][80/200]	Time 0.353 (0.414)	Data 0.001 (0.058)	Loss 1.995 (1.040)
Epoch: [29][100/200]	Time 0.354 (0.422)	Data 0.001 (0.065)	Loss 1.464 (1.129)
Epoch: [29][120/200]	Time 0.359 (0.411)	Data 0.000 (0.054)	Loss 1.958 (1.207)
Epoch: [29][140/200]	Time 0.356 (0.415)	Data 0.001 (0.059)	Loss 1.480 (1.251)
Epoch: [29][160/200]	Time 0.357 (0.408)	Data 0.000 (0.052)	Loss 1.635 (1.280)
Epoch: [29][180/200]	Time 0.356 (0.413)	Data 0.001 (0.056)	Loss 1.931 (1.314)
Epoch: [29][200/200]	Time 0.356 (0.416)	Data 0.001 (0.059)	Loss 1.311 (1.318)
Extract Features: [50/367]	Time 0.426 (0.222)	Data 0.312 (0.106)	
Extract Features: [100/367]	Time 0.118 (0.202)	Data 0.000 (0.085)	
Extract Features: [150/367]	Time 0.115 (0.198)	Data 0.000 (0.081)	
Extract Features: [200/367]	Time 0.117 (0.195)	Data 0.000 (0.077)	
Extract Features: [250/367]	Time 0.119 (0.192)	Data 0.000 (0.075)	
Extract Features: [300/367]	Time 0.117 (0.190)	Data 0.000 (0.072)	
Extract Features: [350/367]	Time 0.124 (0.188)	Data 0.000 (0.071)	
Mean AP: 56.5%

 * Finished epoch  29  model mAP: 56.5%  best: 56.5% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.162 (0.217)	Data 0.049 (0.102)	
Extract Features: [100/128]	Time 0.116 (0.197)	Data 0.000 (0.081)	
Computing jaccard distance...
Jaccard distance computing time cost: 65.70616674423218
==> Statistics for epoch 30: 1021 clusters
Epoch: [30][20/200]	Time 0.352 (0.413)	Data 0.001 (0.059)	Loss 0.166 (0.272)
Epoch: [30][40/200]	Time 0.353 (0.434)	Data 0.001 (0.075)	Loss 1.740 (0.521)
Epoch: [30][60/200]	Time 0.352 (0.407)	Data 0.000 (0.050)	Loss 1.455 (0.811)
Epoch: [30][80/200]	Time 0.356 (0.416)	Data 0.001 (0.059)	Loss 1.350 (0.976)
Epoch: [30][100/200]	Time 0.355 (0.421)	Data 0.001 (0.065)	Loss 1.764 (1.074)
Epoch: [30][120/200]	Time 0.352 (0.410)	Data 0.000 (0.054)	Loss 1.553 (1.141)
Epoch: [30][140/200]	Time 0.356 (0.415)	Data 0.001 (0.058)	Loss 1.304 (1.184)
Epoch: [30][160/200]	Time 0.355 (0.419)	Data 0.001 (0.062)	Loss 1.234 (1.228)
Epoch: [30][180/200]	Time 0.352 (0.412)	Data 0.000 (0.055)	Loss 1.510 (1.259)
Epoch: [30][200/200]	Time 0.356 (0.415)	Data 0.001 (0.059)	Loss 1.346 (1.294)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.216)	Data 0.000 (0.101)	
Extract Features: [100/128]	Time 0.117 (0.195)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.99458980560303
==> Statistics for epoch 31: 1025 clusters
Epoch: [31][20/200]	Time 0.352 (0.406)	Data 0.001 (0.052)	Loss 0.273 (0.270)
Epoch: [31][40/200]	Time 0.356 (0.427)	Data 0.001 (0.073)	Loss 1.684 (0.483)
Epoch: [31][60/200]	Time 0.353 (0.403)	Data 0.000 (0.049)	Loss 1.922 (0.813)
Epoch: [31][80/200]	Time 0.358 (0.418)	Data 0.002 (0.061)	Loss 1.791 (0.972)
Epoch: [31][100/200]	Time 0.355 (0.424)	Data 0.001 (0.067)	Loss 1.370 (1.085)
Epoch: [31][120/200]	Time 0.357 (0.413)	Data 0.001 (0.056)	Loss 1.513 (1.148)
Epoch: [31][140/200]	Time 0.358 (0.418)	Data 0.001 (0.061)	Loss 1.872 (1.219)
Epoch: [31][160/200]	Time 0.355 (0.410)	Data 0.000 (0.053)	Loss 1.513 (1.251)
Epoch: [31][180/200]	Time 0.352 (0.415)	Data 0.001 (0.058)	Loss 1.116 (1.280)
Epoch: [31][200/200]	Time 0.361 (0.419)	Data 0.001 (0.062)	Loss 1.455 (1.301)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.112 (0.227)	Data 0.000 (0.109)	
Extract Features: [100/128]	Time 0.117 (0.201)	Data 0.000 (0.084)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.99739050865173
==> Statistics for epoch 32: 1029 clusters
Epoch: [32][20/200]	Time 0.369 (0.411)	Data 0.001 (0.053)	Loss 0.266 (0.270)
Epoch: [32][40/200]	Time 0.356 (0.431)	Data 0.001 (0.074)	Loss 1.897 (0.502)
Epoch: [32][60/200]	Time 0.355 (0.405)	Data 0.000 (0.050)	Loss 1.578 (0.821)
Epoch: [32][80/200]	Time 0.354 (0.416)	Data 0.001 (0.060)	Loss 1.788 (0.985)
Epoch: [32][100/200]	Time 0.353 (0.422)	Data 0.001 (0.067)	Loss 1.358 (1.103)
Epoch: [32][120/200]	Time 0.356 (0.411)	Data 0.001 (0.056)	Loss 1.530 (1.177)
Epoch: [32][140/200]	Time 0.356 (0.416)	Data 0.001 (0.060)	Loss 1.347 (1.233)
Epoch: [32][160/200]	Time 0.354 (0.409)	Data 0.000 (0.052)	Loss 1.631 (1.266)
Epoch: [32][180/200]	Time 0.357 (0.414)	Data 0.001 (0.057)	Loss 1.074 (1.298)
Epoch: [32][200/200]	Time 0.355 (0.418)	Data 0.001 (0.061)	Loss 1.027 (1.320)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.394 (0.218)	Data 0.282 (0.104)	
Extract Features: [100/128]	Time 0.115 (0.196)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.37106513977051
==> Statistics for epoch 33: 1030 clusters
Epoch: [33][20/200]	Time 0.352 (0.409)	Data 0.001 (0.050)	Loss 0.396 (0.262)
Epoch: [33][40/200]	Time 0.361 (0.427)	Data 0.001 (0.070)	Loss 1.071 (0.503)
Epoch: [33][60/200]	Time 0.352 (0.402)	Data 0.000 (0.047)	Loss 1.560 (0.830)
Epoch: [33][80/200]	Time 0.356 (0.413)	Data 0.001 (0.057)	Loss 1.290 (1.002)
Epoch: [33][100/200]	Time 0.354 (0.419)	Data 0.001 (0.063)	Loss 0.966 (1.092)
Epoch: [33][120/200]	Time 0.355 (0.408)	Data 0.001 (0.053)	Loss 1.475 (1.162)
Epoch: [33][140/200]	Time 0.365 (0.414)	Data 0.001 (0.058)	Loss 1.518 (1.221)
Epoch: [33][160/200]	Time 0.354 (0.407)	Data 0.000 (0.050)	Loss 1.378 (1.250)
Epoch: [33][180/200]	Time 0.354 (0.411)	Data 0.001 (0.055)	Loss 1.364 (1.283)
Epoch: [33][200/200]	Time 0.359 (0.415)	Data 0.001 (0.059)	Loss 1.376 (1.301)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.113 (0.209)	Data 0.000 (0.092)	
Extract Features: [100/128]	Time 0.117 (0.195)	Data 0.001 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.59887623786926
==> Statistics for epoch 34: 1035 clusters
Epoch: [34][20/200]	Time 0.351 (0.422)	Data 0.001 (0.061)	Loss 0.222 (0.269)
Epoch: [34][40/200]	Time 0.356 (0.437)	Data 0.001 (0.079)	Loss 1.866 (0.516)
Epoch: [34][60/200]	Time 0.353 (0.410)	Data 0.000 (0.053)	Loss 1.313 (0.825)
Epoch: [34][80/200]	Time 0.356 (0.418)	Data 0.001 (0.061)	Loss 1.090 (0.984)
Epoch: [34][100/200]	Time 0.356 (0.423)	Data 0.001 (0.066)	Loss 1.576 (1.079)
Epoch: [34][120/200]	Time 0.358 (0.413)	Data 0.001 (0.055)	Loss 1.517 (1.143)
Epoch: [34][140/200]	Time 0.356 (0.419)	Data 0.001 (0.061)	Loss 1.629 (1.208)
Epoch: [34][160/200]	Time 0.355 (0.412)	Data 0.000 (0.054)	Loss 1.496 (1.256)
Epoch: [34][180/200]	Time 0.356 (0.416)	Data 0.001 (0.058)	Loss 1.376 (1.290)
Epoch: [34][200/200]	Time 0.355 (0.419)	Data 0.001 (0.061)	Loss 1.429 (1.318)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.256 (0.225)	Data 0.000 (0.107)	
Extract Features: [100/128]	Time 0.116 (0.197)	Data 0.000 (0.080)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.5722815990448
==> Statistics for epoch 35: 1018 clusters
Epoch: [35][20/200]	Time 0.354 (0.413)	Data 0.001 (0.058)	Loss 0.225 (0.257)
Epoch: [35][40/200]	Time 0.352 (0.429)	Data 0.001 (0.072)	Loss 1.359 (0.489)
Epoch: [35][60/200]	Time 0.354 (0.404)	Data 0.000 (0.048)	Loss 1.590 (0.810)
Epoch: [35][80/200]	Time 0.359 (0.413)	Data 0.001 (0.057)	Loss 1.804 (0.955)
Epoch: [35][100/200]	Time 0.352 (0.419)	Data 0.001 (0.063)	Loss 1.168 (1.069)
Epoch: [35][120/200]	Time 0.352 (0.409)	Data 0.000 (0.053)	Loss 1.296 (1.124)
Epoch: [35][140/200]	Time 0.354 (0.414)	Data 0.001 (0.058)	Loss 1.388 (1.176)
Epoch: [35][160/200]	Time 0.361 (0.417)	Data 0.001 (0.061)	Loss 1.430 (1.216)
Epoch: [35][180/200]	Time 0.356 (0.411)	Data 0.000 (0.055)	Loss 1.048 (1.248)
Epoch: [35][200/200]	Time 0.358 (0.414)	Data 0.001 (0.057)	Loss 1.238 (1.268)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.432 (0.217)	Data 0.319 (0.101)	
Extract Features: [100/128]	Time 0.116 (0.194)	Data 0.000 (0.077)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.73271822929382
==> Statistics for epoch 36: 1025 clusters
Epoch: [36][20/200]	Time 0.354 (0.410)	Data 0.001 (0.053)	Loss 0.220 (0.253)
Epoch: [36][40/200]	Time 0.353 (0.428)	Data 0.001 (0.072)	Loss 1.370 (0.481)
Epoch: [36][60/200]	Time 0.355 (0.404)	Data 0.000 (0.049)	Loss 1.947 (0.835)
Epoch: [36][80/200]	Time 0.354 (0.416)	Data 0.001 (0.060)	Loss 1.517 (0.986)
Epoch: [36][100/200]	Time 0.355 (0.425)	Data 0.001 (0.067)	Loss 1.996 (1.089)
Epoch: [36][120/200]	Time 0.357 (0.413)	Data 0.001 (0.056)	Loss 1.254 (1.150)
Epoch: [36][140/200]	Time 0.360 (0.418)	Data 0.001 (0.061)	Loss 1.325 (1.201)
Epoch: [36][160/200]	Time 0.353 (0.410)	Data 0.000 (0.054)	Loss 1.730 (1.245)
Epoch: [36][180/200]	Time 0.354 (0.414)	Data 0.001 (0.058)	Loss 1.313 (1.273)
Epoch: [36][200/200]	Time 0.357 (0.418)	Data 0.001 (0.061)	Loss 1.718 (1.297)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.121 (0.215)	Data 0.001 (0.097)	
Extract Features: [100/128]	Time 0.119 (0.193)	Data 0.001 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.363882541656494
==> Statistics for epoch 37: 1032 clusters
Epoch: [37][20/200]	Time 0.353 (0.409)	Data 0.001 (0.054)	Loss 0.245 (0.276)
Epoch: [37][40/200]	Time 0.364 (0.424)	Data 0.001 (0.069)	Loss 1.638 (0.505)
Epoch: [37][60/200]	Time 0.354 (0.402)	Data 0.000 (0.046)	Loss 1.526 (0.808)
Epoch: [37][80/200]	Time 0.354 (0.412)	Data 0.001 (0.056)	Loss 1.626 (0.991)
Epoch: [37][100/200]	Time 0.355 (0.419)	Data 0.001 (0.063)	Loss 1.618 (1.087)
Epoch: [37][120/200]	Time 0.354 (0.410)	Data 0.000 (0.052)	Loss 1.622 (1.154)
Epoch: [37][140/200]	Time 0.359 (0.414)	Data 0.000 (0.057)	Loss 1.666 (1.200)
Epoch: [37][160/200]	Time 0.354 (0.407)	Data 0.000 (0.050)	Loss 1.336 (1.229)
Epoch: [37][180/200]	Time 0.354 (0.412)	Data 0.001 (0.055)	Loss 1.915 (1.260)
Epoch: [37][200/200]	Time 0.354 (0.415)	Data 0.000 (0.058)	Loss 1.680 (1.280)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.221)	Data 0.000 (0.103)	
Extract Features: [100/128]	Time 0.238 (0.199)	Data 0.000 (0.081)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.878499031066895
==> Statistics for epoch 38: 1022 clusters
Epoch: [38][20/200]	Time 0.356 (0.413)	Data 0.001 (0.056)	Loss 0.282 (0.288)
Epoch: [38][40/200]	Time 0.363 (0.429)	Data 0.001 (0.073)	Loss 1.652 (0.522)
Epoch: [38][60/200]	Time 0.355 (0.405)	Data 0.000 (0.049)	Loss 1.546 (0.820)
Epoch: [38][80/200]	Time 0.357 (0.415)	Data 0.001 (0.059)	Loss 1.569 (0.981)
Epoch: [38][100/200]	Time 0.362 (0.423)	Data 0.001 (0.067)	Loss 1.386 (1.089)
Epoch: [38][120/200]	Time 0.356 (0.412)	Data 0.000 (0.056)	Loss 1.599 (1.164)
Epoch: [38][140/200]	Time 0.356 (0.418)	Data 0.001 (0.061)	Loss 1.632 (1.195)
Epoch: [38][160/200]	Time 0.356 (0.421)	Data 0.001 (0.064)	Loss 1.306 (1.246)
Epoch: [38][180/200]	Time 0.359 (0.414)	Data 0.001 (0.057)	Loss 1.678 (1.277)
Epoch: [38][200/200]	Time 0.359 (0.418)	Data 0.001 (0.061)	Loss 1.741 (1.300)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.217)	Data 0.000 (0.099)	
Extract Features: [100/128]	Time 0.116 (0.195)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.15025186538696
==> Statistics for epoch 39: 1043 clusters
Epoch: [39][20/200]	Time 0.367 (0.417)	Data 0.001 (0.052)	Loss 0.222 (0.273)
Epoch: [39][40/200]	Time 0.363 (0.436)	Data 0.001 (0.075)	Loss 1.485 (0.497)
Epoch: [39][60/200]	Time 0.352 (0.410)	Data 0.000 (0.050)	Loss 1.409 (0.824)
Epoch: [39][80/200]	Time 0.354 (0.419)	Data 0.001 (0.060)	Loss 2.419 (0.989)
Epoch: [39][100/200]	Time 0.367 (0.426)	Data 0.001 (0.067)	Loss 1.523 (1.084)
Epoch: [39][120/200]	Time 0.359 (0.416)	Data 0.001 (0.056)	Loss 1.337 (1.148)
Epoch: [39][140/200]	Time 0.359 (0.421)	Data 0.001 (0.061)	Loss 1.230 (1.205)
Epoch: [39][160/200]	Time 0.356 (0.413)	Data 0.000 (0.053)	Loss 1.108 (1.247)
Epoch: [39][180/200]	Time 0.359 (0.417)	Data 0.001 (0.057)	Loss 1.543 (1.269)
Epoch: [39][200/200]	Time 0.359 (0.420)	Data 0.001 (0.061)	Loss 1.369 (1.301)
Extract Features: [50/367]	Time 0.195 (0.214)	Data 0.081 (0.097)	
Extract Features: [100/367]	Time 0.116 (0.199)	Data 0.000 (0.082)	
Extract Features: [150/367]	Time 0.117 (0.194)	Data 0.000 (0.077)	
Extract Features: [200/367]	Time 0.117 (0.191)	Data 0.000 (0.074)	
Extract Features: [250/367]	Time 0.117 (0.190)	Data 0.000 (0.072)	
Extract Features: [300/367]	Time 0.117 (0.189)	Data 0.000 (0.071)	
Extract Features: [350/367]	Time 0.116 (0.188)	Data 0.000 (0.071)	
Mean AP: 57.3%

 * Finished epoch  39  model mAP: 57.3%  best: 57.3% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.218)	Data 0.000 (0.100)	
Extract Features: [100/128]	Time 0.113 (0.197)	Data 0.000 (0.081)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.874858140945435
==> Statistics for epoch 40: 1048 clusters
Epoch: [40][20/200]	Time 0.353 (0.412)	Data 0.001 (0.057)	Loss 0.290 (0.271)
Epoch: [40][40/200]	Time 0.356 (0.430)	Data 0.001 (0.072)	Loss 1.161 (0.504)
Epoch: [40][60/200]	Time 0.356 (0.405)	Data 0.000 (0.048)	Loss 2.111 (0.826)
Epoch: [40][80/200]	Time 0.355 (0.416)	Data 0.001 (0.060)	Loss 0.822 (0.978)
Epoch: [40][100/200]	Time 0.355 (0.422)	Data 0.001 (0.066)	Loss 1.192 (1.077)
Epoch: [40][120/200]	Time 0.352 (0.411)	Data 0.001 (0.055)	Loss 1.523 (1.147)
Epoch: [40][140/200]	Time 0.355 (0.416)	Data 0.001 (0.060)	Loss 1.619 (1.202)
Epoch: [40][160/200]	Time 0.354 (0.409)	Data 0.000 (0.053)	Loss 1.841 (1.237)
Epoch: [40][180/200]	Time 0.354 (0.414)	Data 0.001 (0.057)	Loss 1.294 (1.269)
Epoch: [40][200/200]	Time 0.355 (0.417)	Data 0.001 (0.060)	Loss 1.878 (1.294)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.202 (0.211)	Data 0.090 (0.097)	
Extract Features: [100/128]	Time 0.115 (0.192)	Data 0.000 (0.075)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.528205156326294
==> Statistics for epoch 41: 1040 clusters
Epoch: [41][20/200]	Time 0.352 (0.408)	Data 0.001 (0.054)	Loss 0.134 (0.272)
Epoch: [41][40/200]	Time 0.352 (0.427)	Data 0.001 (0.072)	Loss 1.604 (0.514)
Epoch: [41][60/200]	Time 0.351 (0.402)	Data 0.000 (0.048)	Loss 1.343 (0.825)
Epoch: [41][80/200]	Time 0.353 (0.414)	Data 0.001 (0.058)	Loss 1.546 (0.985)
Epoch: [41][100/200]	Time 0.370 (0.419)	Data 0.001 (0.063)	Loss 1.452 (1.083)
Epoch: [41][120/200]	Time 0.355 (0.409)	Data 0.001 (0.053)	Loss 1.327 (1.146)
Epoch: [41][140/200]	Time 0.358 (0.415)	Data 0.002 (0.058)	Loss 1.330 (1.203)
Epoch: [41][160/200]	Time 0.353 (0.408)	Data 0.000 (0.051)	Loss 1.123 (1.241)
Epoch: [41][180/200]	Time 0.355 (0.412)	Data 0.001 (0.056)	Loss 1.500 (1.267)
Epoch: [41][200/200]	Time 0.360 (0.415)	Data 0.001 (0.059)	Loss 1.382 (1.288)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.217)	Data 0.000 (0.100)	
Extract Features: [100/128]	Time 0.117 (0.196)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.787688970565796
==> Statistics for epoch 42: 1049 clusters
Epoch: [42][20/200]	Time 0.354 (0.410)	Data 0.001 (0.054)	Loss 0.235 (0.250)
Epoch: [42][40/200]	Time 0.357 (0.425)	Data 0.001 (0.069)	Loss 1.790 (0.498)
Epoch: [42][60/200]	Time 0.354 (0.402)	Data 0.000 (0.046)	Loss 1.593 (0.820)
Epoch: [42][80/200]	Time 0.354 (0.413)	Data 0.001 (0.057)	Loss 1.560 (0.978)
Epoch: [42][100/200]	Time 0.355 (0.420)	Data 0.001 (0.064)	Loss 1.179 (1.072)
Epoch: [42][120/200]	Time 0.376 (0.411)	Data 0.001 (0.053)	Loss 1.369 (1.147)
Epoch: [42][140/200]	Time 0.352 (0.417)	Data 0.001 (0.059)	Loss 1.234 (1.196)
Epoch: [42][160/200]	Time 0.355 (0.409)	Data 0.000 (0.052)	Loss 1.511 (1.242)
Epoch: [42][180/200]	Time 0.357 (0.414)	Data 0.001 (0.057)	Loss 1.586 (1.275)
Epoch: [42][200/200]	Time 0.357 (0.417)	Data 0.001 (0.060)	Loss 2.117 (1.305)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.323 (0.215)	Data 0.209 (0.097)	
Extract Features: [100/128]	Time 0.117 (0.195)	Data 0.000 (0.077)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.321080446243286
==> Statistics for epoch 43: 1053 clusters
Epoch: [43][20/200]	Time 0.352 (0.415)	Data 0.001 (0.054)	Loss 0.318 (0.269)
Epoch: [43][40/200]	Time 0.351 (0.429)	Data 0.001 (0.071)	Loss 1.215 (0.476)
Epoch: [43][60/200]	Time 0.352 (0.404)	Data 0.000 (0.048)	Loss 1.186 (0.775)
Epoch: [43][80/200]	Time 0.357 (0.415)	Data 0.001 (0.059)	Loss 1.414 (0.968)
Epoch: [43][100/200]	Time 0.349 (0.421)	Data 0.001 (0.065)	Loss 1.445 (1.074)
Epoch: [43][120/200]	Time 0.354 (0.411)	Data 0.001 (0.054)	Loss 1.489 (1.142)
Epoch: [43][140/200]	Time 0.353 (0.417)	Data 0.001 (0.060)	Loss 1.780 (1.185)
Epoch: [43][160/200]	Time 0.354 (0.410)	Data 0.000 (0.053)	Loss 1.472 (1.228)
Epoch: [43][180/200]	Time 0.357 (0.415)	Data 0.001 (0.057)	Loss 1.631 (1.262)
Epoch: [43][200/200]	Time 0.370 (0.418)	Data 0.001 (0.061)	Loss 1.589 (1.283)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.228)	Data 0.000 (0.110)	
Extract Features: [100/128]	Time 0.117 (0.203)	Data 0.000 (0.086)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.14496827125549
==> Statistics for epoch 44: 1037 clusters
Epoch: [44][20/200]	Time 0.353 (0.408)	Data 0.001 (0.049)	Loss 0.219 (0.274)
Epoch: [44][40/200]	Time 0.357 (0.427)	Data 0.001 (0.069)	Loss 1.489 (0.494)
Epoch: [44][60/200]	Time 0.353 (0.405)	Data 0.000 (0.046)	Loss 1.842 (0.812)
Epoch: [44][80/200]	Time 0.367 (0.416)	Data 0.001 (0.057)	Loss 1.458 (0.976)
Epoch: [44][100/200]	Time 0.357 (0.422)	Data 0.001 (0.064)	Loss 1.339 (1.073)
Epoch: [44][120/200]	Time 0.358 (0.411)	Data 0.001 (0.054)	Loss 1.815 (1.141)
Epoch: [44][140/200]	Time 0.358 (0.417)	Data 0.001 (0.060)	Loss 1.537 (1.184)
Epoch: [44][160/200]	Time 0.356 (0.410)	Data 0.000 (0.052)	Loss 1.203 (1.218)
Epoch: [44][180/200]	Time 0.355 (0.414)	Data 0.001 (0.056)	Loss 1.966 (1.253)
Epoch: [44][200/200]	Time 0.355 (0.418)	Data 0.001 (0.060)	Loss 1.658 (1.276)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.216)	Data 0.000 (0.101)	
Extract Features: [100/128]	Time 0.127 (0.194)	Data 0.000 (0.077)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.66800904273987
==> Statistics for epoch 45: 1044 clusters
Epoch: [45][20/200]	Time 0.350 (0.409)	Data 0.001 (0.054)	Loss 0.315 (0.281)
Epoch: [45][40/200]	Time 0.353 (0.423)	Data 0.001 (0.070)	Loss 2.040 (0.516)
Epoch: [45][60/200]	Time 0.352 (0.400)	Data 0.000 (0.047)	Loss 1.022 (0.807)
Epoch: [45][80/200]	Time 0.355 (0.411)	Data 0.001 (0.057)	Loss 1.369 (0.982)
Epoch: [45][100/200]	Time 0.353 (0.417)	Data 0.001 (0.063)	Loss 1.606 (1.086)
Epoch: [45][120/200]	Time 0.356 (0.407)	Data 0.001 (0.053)	Loss 1.419 (1.145)
Epoch: [45][140/200]	Time 0.358 (0.414)	Data 0.001 (0.058)	Loss 1.323 (1.205)
Epoch: [45][160/200]	Time 0.354 (0.407)	Data 0.000 (0.051)	Loss 1.427 (1.229)
Epoch: [45][180/200]	Time 0.357 (0.411)	Data 0.001 (0.055)	Loss 1.018 (1.258)
Epoch: [45][200/200]	Time 0.355 (0.415)	Data 0.001 (0.059)	Loss 1.461 (1.279)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.214)	Data 0.000 (0.097)	
Extract Features: [100/128]	Time 0.115 (0.196)	Data 0.000 (0.080)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.808117389678955
==> Statistics for epoch 46: 1040 clusters
Epoch: [46][20/200]	Time 0.375 (0.416)	Data 0.001 (0.054)	Loss 0.213 (0.267)
Epoch: [46][40/200]	Time 0.355 (0.433)	Data 0.001 (0.074)	Loss 2.005 (0.487)
Epoch: [46][60/200]	Time 0.354 (0.407)	Data 0.000 (0.049)	Loss 1.547 (0.790)
Epoch: [46][80/200]	Time 0.357 (0.419)	Data 0.001 (0.062)	Loss 1.353 (0.952)
Epoch: [46][100/200]	Time 0.356 (0.426)	Data 0.001 (0.069)	Loss 1.292 (1.052)
Epoch: [46][120/200]	Time 0.356 (0.415)	Data 0.001 (0.058)	Loss 1.033 (1.110)
Epoch: [46][140/200]	Time 0.358 (0.419)	Data 0.001 (0.062)	Loss 1.944 (1.158)
Epoch: [46][160/200]	Time 0.352 (0.411)	Data 0.000 (0.054)	Loss 1.839 (1.202)
Epoch: [46][180/200]	Time 0.358 (0.417)	Data 0.001 (0.059)	Loss 1.235 (1.228)
Epoch: [46][200/200]	Time 0.356 (0.421)	Data 0.001 (0.063)	Loss 1.252 (1.256)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.116 (0.216)	Data 0.000 (0.102)	
Extract Features: [100/128]	Time 0.117 (0.195)	Data 0.000 (0.079)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.643903970718384
==> Statistics for epoch 47: 1042 clusters
Epoch: [47][20/200]	Time 0.355 (0.411)	Data 0.001 (0.055)	Loss 0.236 (0.250)
Epoch: [47][40/200]	Time 0.352 (0.425)	Data 0.001 (0.071)	Loss 1.567 (0.481)
Epoch: [47][60/200]	Time 0.350 (0.402)	Data 0.000 (0.048)	Loss 1.136 (0.778)
Epoch: [47][80/200]	Time 0.354 (0.412)	Data 0.001 (0.057)	Loss 1.642 (0.944)
Epoch: [47][100/200]	Time 0.353 (0.419)	Data 0.001 (0.064)	Loss 1.452 (1.047)
Epoch: [47][120/200]	Time 0.354 (0.409)	Data 0.001 (0.053)	Loss 1.492 (1.117)
Epoch: [47][140/200]	Time 0.358 (0.414)	Data 0.001 (0.059)	Loss 1.455 (1.165)
Epoch: [47][160/200]	Time 0.356 (0.407)	Data 0.000 (0.051)	Loss 0.921 (1.206)
Epoch: [47][180/200]	Time 0.363 (0.412)	Data 0.001 (0.056)	Loss 1.296 (1.237)
Epoch: [47][200/200]	Time 0.362 (0.416)	Data 0.001 (0.060)	Loss 1.284 (1.265)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.115 (0.212)	Data 0.000 (0.095)	
Extract Features: [100/128]	Time 0.115 (0.197)	Data 0.000 (0.080)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.39188742637634
==> Statistics for epoch 48: 1041 clusters
Epoch: [48][20/200]	Time 0.355 (0.416)	Data 0.001 (0.059)	Loss 0.175 (0.262)
Epoch: [48][40/200]	Time 0.354 (0.435)	Data 0.001 (0.078)	Loss 1.488 (0.496)
Epoch: [48][60/200]	Time 0.351 (0.408)	Data 0.000 (0.052)	Loss 1.358 (0.798)
Epoch: [48][80/200]	Time 0.354 (0.417)	Data 0.001 (0.062)	Loss 1.576 (0.972)
Epoch: [48][100/200]	Time 0.353 (0.425)	Data 0.001 (0.068)	Loss 1.448 (1.079)
Epoch: [48][120/200]	Time 0.354 (0.414)	Data 0.001 (0.057)	Loss 1.301 (1.162)
Epoch: [48][140/200]	Time 0.356 (0.418)	Data 0.001 (0.061)	Loss 1.620 (1.203)
Epoch: [48][160/200]	Time 0.355 (0.411)	Data 0.000 (0.054)	Loss 1.506 (1.236)
Epoch: [48][180/200]	Time 0.358 (0.415)	Data 0.001 (0.058)	Loss 1.859 (1.268)
Epoch: [48][200/200]	Time 0.356 (0.418)	Data 0.001 (0.061)	Loss 1.484 (1.298)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.114 (0.215)	Data 0.000 (0.096)	
Extract Features: [100/128]	Time 0.115 (0.195)	Data 0.000 (0.078)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.73207330703735
==> Statistics for epoch 49: 1060 clusters
Epoch: [49][20/200]	Time 0.354 (0.409)	Data 0.001 (0.050)	Loss 0.250 (0.262)
Epoch: [49][40/200]	Time 0.357 (0.427)	Data 0.002 (0.069)	Loss 1.465 (0.439)
Epoch: [49][60/200]	Time 0.353 (0.403)	Data 0.000 (0.046)	Loss 1.562 (0.780)
Epoch: [49][80/200]	Time 0.351 (0.413)	Data 0.001 (0.057)	Loss 1.647 (0.973)
Epoch: [49][100/200]	Time 2.184 (0.419)	Data 1.796 (0.063)	Loss 1.495 (1.064)
Epoch: [49][120/200]	Time 0.355 (0.409)	Data 0.001 (0.053)	Loss 1.795 (1.145)
Epoch: [49][140/200]	Time 0.358 (0.415)	Data 0.001 (0.058)	Loss 1.003 (1.193)
Epoch: [49][160/200]	Time 0.356 (0.409)	Data 0.000 (0.051)	Loss 1.678 (1.236)
Epoch: [49][180/200]	Time 0.357 (0.414)	Data 0.001 (0.056)	Loss 1.602 (1.274)
Epoch: [49][200/200]	Time 0.356 (0.417)	Data 0.001 (0.059)	Loss 1.973 (1.294)
Extract Features: [50/367]	Time 0.114 (0.218)	Data 0.000 (0.102)	
Extract Features: [100/367]	Time 0.115 (0.199)	Data 0.000 (0.082)	
Extract Features: [150/367]	Time 0.117 (0.195)	Data 0.000 (0.078)	
Extract Features: [200/367]	Time 0.115 (0.191)	Data 0.000 (0.074)	
Extract Features: [250/367]	Time 0.122 (0.189)	Data 0.000 (0.072)	
Extract Features: [300/367]	Time 0.117 (0.188)	Data 0.000 (0.070)	
Extract Features: [350/367]	Time 0.118 (0.187)	Data 0.000 (0.069)	
Mean AP: 57.4%

 * Finished epoch  49  model mAP: 57.4%  best: 57.4% *

==> Test with the best model:
=> Loaded checkpoint 'log/market2msmt/resnet50_cion/model_best.pth.tar'
Extract Features: [50/367]	Time 0.112 (0.211)	Data 0.000 (0.094)	
Extract Features: [100/367]	Time 0.116 (0.201)	Data 0.000 (0.084)	
Extract Features: [150/367]	Time 0.263 (0.196)	Data 0.148 (0.080)	
Extract Features: [200/367]	Time 0.116 (0.193)	Data 0.000 (0.076)	
Extract Features: [250/367]	Time 0.362 (0.192)	Data 0.246 (0.074)	
Extract Features: [300/367]	Time 0.115 (0.189)	Data 0.000 (0.073)	
Extract Features: [350/367]	Time 0.391 (0.189)	Data 0.278 (0.072)	
Mean AP: 57.4%
CMC Scores:
  top-1          81.3%
  top-5          89.2%
  top-10         91.4%
Total running time:  2:58:47.389984
