==========
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='resnet152', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet152_Market1501/64bs_lr0.0004_ep120_warm20_seed0/resnet152_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/resnet152_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 0.227 (0.432)	Data 0.001 (0.021)	
Extract Features: [100/128]	Time 0.229 (0.339)	Data 0.000 (0.011)	
Computing jaccard distance...
Jaccard distance computing time cost: 66.13232159614563
Clustering criterion: eps: 0.700
==> Statistics for epoch 0: 765 clusters
Epoch: [0][20/200]	Time 0.691 (1.242)	Data 0.000 (0.054)	Loss 3.136 (3.157)
Epoch: [0][40/200]	Time 0.694 (1.011)	Data 0.000 (0.065)	Loss 2.236 (2.917)
Epoch: [0][60/200]	Time 0.697 (0.933)	Data 0.001 (0.070)	Loss 1.786 (2.714)
Epoch: [0][80/200]	Time 0.698 (0.894)	Data 0.001 (0.070)	Loss 2.131 (2.508)
Epoch: [0][100/200]	Time 0.699 (0.873)	Data 0.001 (0.072)	Loss 1.580 (2.380)
Epoch: [0][120/200]	Time 0.694 (0.858)	Data 0.001 (0.073)	Loss 1.286 (2.286)
Epoch: [0][140/200]	Time 0.698 (0.847)	Data 0.001 (0.074)	Loss 1.522 (2.197)
Epoch: [0][160/200]	Time 0.699 (0.829)	Data 0.000 (0.064)	Loss 1.305 (2.126)
Epoch: [0][180/200]	Time 0.699 (0.824)	Data 0.000 (0.066)	Loss 1.712 (2.071)
Epoch: [0][200/200]	Time 0.699 (0.820)	Data 0.001 (0.067)	Loss 1.507 (2.023)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.258)	Data 0.000 (0.024)	
Extract Features: [100/128]	Time 0.228 (0.249)	Data 0.000 (0.012)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.650578022003174
==> Statistics for epoch 1: 942 clusters
Epoch: [1][20/200]	Time 0.696 (0.753)	Data 0.001 (0.046)	Loss 0.325 (0.398)
Epoch: [1][40/200]	Time 0.690 (0.762)	Data 0.001 (0.059)	Loss 1.784 (0.756)
Epoch: [1][60/200]	Time 0.695 (0.768)	Data 0.001 (0.065)	Loss 1.654 (1.056)
Epoch: [1][80/200]	Time 0.691 (0.750)	Data 0.001 (0.049)	Loss 1.690 (1.230)
Epoch: [1][100/200]	Time 0.690 (0.755)	Data 0.001 (0.055)	Loss 1.469 (1.325)
Epoch: [1][120/200]	Time 0.692 (0.759)	Data 0.001 (0.059)	Loss 1.725 (1.373)
Epoch: [1][140/200]	Time 0.693 (0.749)	Data 0.000 (0.051)	Loss 1.851 (1.414)
Epoch: [1][160/200]	Time 0.693 (0.753)	Data 0.001 (0.054)	Loss 1.617 (1.435)
Epoch: [1][180/200]	Time 0.696 (0.756)	Data 0.001 (0.057)	Loss 1.120 (1.435)
Epoch: [1][200/200]	Time 0.695 (0.750)	Data 0.000 (0.051)	Loss 1.639 (1.455)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.232 (0.260)	Data 0.000 (0.025)	
Extract Features: [100/128]	Time 0.227 (0.248)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.27562189102173
==> Statistics for epoch 2: 975 clusters
Epoch: [2][20/200]	Time 0.693 (0.753)	Data 0.001 (0.052)	Loss 0.366 (0.391)
Epoch: [2][40/200]	Time 0.696 (0.771)	Data 0.001 (0.069)	Loss 2.153 (0.680)
Epoch: [2][60/200]	Time 0.691 (0.745)	Data 0.000 (0.046)	Loss 1.521 (1.017)
Epoch: [2][80/200]	Time 0.695 (0.758)	Data 0.001 (0.058)	Loss 1.866 (1.213)
Epoch: [2][100/200]	Time 0.699 (0.761)	Data 0.001 (0.062)	Loss 1.340 (1.279)
Epoch: [2][120/200]	Time 0.701 (0.752)	Data 0.000 (0.052)	Loss 1.362 (1.340)
Epoch: [2][140/200]	Time 0.700 (0.756)	Data 0.001 (0.056)	Loss 1.542 (1.375)
Epoch: [2][160/200]	Time 0.701 (0.759)	Data 0.001 (0.058)	Loss 1.589 (1.409)
Epoch: [2][180/200]	Time 0.700 (0.753)	Data 0.000 (0.052)	Loss 1.778 (1.435)
Epoch: [2][200/200]	Time 0.698 (0.757)	Data 0.001 (0.055)	Loss 1.913 (1.444)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.259)	Data 0.000 (0.025)	
Extract Features: [100/128]	Time 0.229 (0.248)	Data 0.000 (0.012)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.5453565120697
==> Statistics for epoch 3: 1019 clusters
Epoch: [3][20/200]	Time 0.696 (0.751)	Data 0.001 (0.049)	Loss 0.428 (0.359)
Epoch: [3][40/200]	Time 0.692 (0.767)	Data 0.001 (0.066)	Loss 1.726 (0.621)
Epoch: [3][60/200]	Time 0.702 (0.744)	Data 0.000 (0.044)	Loss 1.668 (0.946)
Epoch: [3][80/200]	Time 0.694 (0.751)	Data 0.001 (0.052)	Loss 1.518 (1.148)
Epoch: [3][100/200]	Time 0.699 (0.759)	Data 0.001 (0.059)	Loss 1.913 (1.241)
Epoch: [3][120/200]	Time 0.700 (0.749)	Data 0.000 (0.049)	Loss 1.500 (1.315)
Epoch: [3][140/200]	Time 0.697 (0.754)	Data 0.001 (0.054)	Loss 1.149 (1.357)
Epoch: [3][160/200]	Time 0.696 (0.759)	Data 0.001 (0.057)	Loss 2.394 (1.393)
Epoch: [3][180/200]	Time 0.781 (0.753)	Data 0.000 (0.051)	Loss 1.616 (1.414)
Epoch: [3][200/200]	Time 0.699 (0.757)	Data 0.001 (0.054)	Loss 1.476 (1.433)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.234 (0.262)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.227 (0.249)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.36489725112915
==> Statistics for epoch 4: 976 clusters
Epoch: [4][20/200]	Time 0.687 (0.746)	Data 0.001 (0.055)	Loss 0.264 (0.318)
Epoch: [4][40/200]	Time 0.697 (0.762)	Data 0.001 (0.071)	Loss 1.773 (0.604)
Epoch: [4][60/200]	Time 0.695 (0.739)	Data 0.000 (0.047)	Loss 1.198 (0.908)
Epoch: [4][80/200]	Time 0.693 (0.754)	Data 0.001 (0.056)	Loss 1.263 (1.046)
Epoch: [4][100/200]	Time 0.705 (0.762)	Data 0.001 (0.061)	Loss 1.715 (1.127)
Epoch: [4][120/200]	Time 0.697 (0.753)	Data 0.000 (0.051)	Loss 1.287 (1.187)
Epoch: [4][140/200]	Time 0.695 (0.759)	Data 0.001 (0.056)	Loss 1.779 (1.228)
Epoch: [4][160/200]	Time 0.697 (0.762)	Data 0.001 (0.060)	Loss 1.963 (1.255)
Epoch: [4][180/200]	Time 0.691 (0.756)	Data 0.000 (0.053)	Loss 1.580 (1.283)
Epoch: [4][200/200]	Time 0.698 (0.759)	Data 0.001 (0.056)	Loss 1.310 (1.292)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.261)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.344 (0.250)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.76846218109131
==> Statistics for epoch 5: 988 clusters
Epoch: [5][20/200]	Time 0.696 (0.759)	Data 0.001 (0.052)	Loss 0.255 (0.284)
Epoch: [5][40/200]	Time 0.701 (0.773)	Data 0.001 (0.066)	Loss 1.568 (0.594)
Epoch: [5][60/200]	Time 0.793 (0.751)	Data 0.000 (0.044)	Loss 1.867 (0.900)
Epoch: [5][80/200]	Time 0.696 (0.760)	Data 0.001 (0.054)	Loss 1.389 (1.056)
Epoch: [5][100/200]	Time 0.697 (0.765)	Data 0.001 (0.060)	Loss 1.515 (1.143)
Epoch: [5][120/200]	Time 0.697 (0.755)	Data 0.000 (0.050)	Loss 1.326 (1.210)
Epoch: [5][140/200]	Time 0.696 (0.759)	Data 0.001 (0.054)	Loss 1.536 (1.242)
Epoch: [5][160/200]	Time 0.698 (0.764)	Data 0.001 (0.058)	Loss 2.058 (1.279)
Epoch: [5][180/200]	Time 0.694 (0.756)	Data 0.000 (0.052)	Loss 1.224 (1.297)
Epoch: [5][200/200]	Time 0.694 (0.759)	Data 0.001 (0.055)	Loss 1.396 (1.311)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.264)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.231 (0.252)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.30065059661865
==> Statistics for epoch 6: 1001 clusters
Epoch: [6][20/200]	Time 0.693 (0.755)	Data 0.001 (0.054)	Loss 0.183 (0.282)
Epoch: [6][40/200]	Time 0.701 (0.775)	Data 0.001 (0.071)	Loss 1.467 (0.560)
Epoch: [6][60/200]	Time 0.691 (0.750)	Data 0.000 (0.047)	Loss 1.019 (0.871)
Epoch: [6][80/200]	Time 0.694 (0.759)	Data 0.001 (0.058)	Loss 1.678 (1.038)
Epoch: [6][100/200]	Time 0.699 (0.765)	Data 0.001 (0.063)	Loss 1.649 (1.120)
Epoch: [6][120/200]	Time 0.694 (0.754)	Data 0.000 (0.053)	Loss 1.690 (1.206)
Epoch: [6][140/200]	Time 0.694 (0.757)	Data 0.001 (0.058)	Loss 1.438 (1.254)
Epoch: [6][160/200]	Time 0.698 (0.761)	Data 0.001 (0.061)	Loss 1.575 (1.290)
Epoch: [6][180/200]	Time 0.693 (0.754)	Data 0.000 (0.054)	Loss 1.489 (1.315)
Epoch: [6][200/200]	Time 0.696 (0.757)	Data 0.001 (0.058)	Loss 1.573 (1.329)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.226 (0.263)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.229 (0.251)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.08864188194275
==> Statistics for epoch 7: 1048 clusters
Epoch: [7][20/200]	Time 0.687 (0.747)	Data 0.001 (0.049)	Loss 0.195 (0.318)
Epoch: [7][40/200]	Time 0.700 (0.761)	Data 0.001 (0.062)	Loss 1.637 (0.539)
Epoch: [7][60/200]	Time 0.694 (0.739)	Data 0.000 (0.042)	Loss 1.381 (0.851)
Epoch: [7][80/200]	Time 0.854 (0.750)	Data 0.001 (0.050)	Loss 1.456 (1.013)
Epoch: [7][100/200]	Time 0.697 (0.755)	Data 0.001 (0.056)	Loss 1.989 (1.114)
Epoch: [7][120/200]	Time 0.698 (0.745)	Data 0.001 (0.047)	Loss 1.928 (1.174)
Epoch: [7][140/200]	Time 0.697 (0.750)	Data 0.001 (0.051)	Loss 1.695 (1.214)
Epoch: [7][160/200]	Time 0.698 (0.743)	Data 0.000 (0.045)	Loss 1.042 (1.238)
Epoch: [7][180/200]	Time 0.698 (0.749)	Data 0.001 (0.049)	Loss 1.474 (1.248)
Epoch: [7][200/200]	Time 0.696 (0.753)	Data 0.001 (0.052)	Loss 1.443 (1.269)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.231 (0.261)	Data 0.000 (0.024)	
Extract Features: [100/128]	Time 0.232 (0.248)	Data 0.000 (0.012)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.92478704452515
==> Statistics for epoch 8: 1053 clusters
Epoch: [8][20/200]	Time 0.687 (0.737)	Data 0.001 (0.050)	Loss 0.236 (0.279)
Epoch: [8][40/200]	Time 0.693 (0.753)	Data 0.001 (0.066)	Loss 1.639 (0.516)
Epoch: [8][60/200]	Time 0.688 (0.733)	Data 0.000 (0.044)	Loss 1.330 (0.823)
Epoch: [8][80/200]	Time 0.801 (0.745)	Data 0.000 (0.053)	Loss 1.525 (0.989)
Epoch: [8][100/200]	Time 0.692 (0.752)	Data 0.001 (0.058)	Loss 1.663 (1.085)
Epoch: [8][120/200]	Time 0.697 (0.745)	Data 0.001 (0.048)	Loss 1.570 (1.130)
Epoch: [8][140/200]	Time 0.703 (0.753)	Data 0.001 (0.055)	Loss 1.378 (1.175)
Epoch: [8][160/200]	Time 0.694 (0.747)	Data 0.000 (0.048)	Loss 1.494 (1.207)
Epoch: [8][180/200]	Time 0.689 (0.751)	Data 0.001 (0.051)	Loss 1.384 (1.234)
Epoch: [8][200/200]	Time 0.690 (0.754)	Data 0.001 (0.055)	Loss 1.365 (1.257)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.225 (0.264)	Data 0.000 (0.025)	
Extract Features: [100/128]	Time 0.227 (0.250)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.6115140914917
==> Statistics for epoch 9: 1027 clusters
Epoch: [9][20/200]	Time 0.691 (0.751)	Data 0.001 (0.054)	Loss 0.237 (0.261)
Epoch: [9][40/200]	Time 0.692 (0.759)	Data 0.001 (0.065)	Loss 1.171 (0.457)
Epoch: [9][60/200]	Time 0.691 (0.736)	Data 0.000 (0.043)	Loss 1.441 (0.764)
Epoch: [9][80/200]	Time 0.692 (0.748)	Data 0.001 (0.054)	Loss 1.805 (0.912)
Epoch: [9][100/200]	Time 0.695 (0.756)	Data 0.001 (0.059)	Loss 1.631 (1.016)
Epoch: [9][120/200]	Time 0.698 (0.748)	Data 0.001 (0.049)	Loss 0.972 (1.080)
Epoch: [9][140/200]	Time 0.698 (0.753)	Data 0.003 (0.053)	Loss 1.372 (1.127)
Epoch: [9][160/200]	Time 0.693 (0.747)	Data 0.000 (0.047)	Loss 1.232 (1.161)
Epoch: [9][180/200]	Time 0.838 (0.752)	Data 0.001 (0.051)	Loss 1.617 (1.191)
Epoch: [9][200/200]	Time 0.696 (0.757)	Data 0.001 (0.055)	Loss 1.291 (1.218)
Extract Features: [50/367]	Time 0.226 (0.265)	Data 0.000 (0.030)	
Extract Features: [100/367]	Time 0.230 (0.250)	Data 0.000 (0.015)	
Extract Features: [150/367]	Time 0.227 (0.248)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.318 (0.248)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.234 (0.249)	Data 0.000 (0.007)	
Extract Features: [300/367]	Time 0.227 (0.248)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.233 (0.247)	Data 0.000 (0.005)	
Mean AP: 48.6%

 * Finished epoch   9  model mAP: 48.6%  best: 48.6% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.237 (0.267)	Data 0.001 (0.030)	
Extract Features: [100/128]	Time 0.229 (0.252)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.94245505332947
==> Statistics for epoch 10: 991 clusters
Epoch: [10][20/200]	Time 0.689 (0.771)	Data 0.001 (0.052)	Loss 0.338 (0.442)
Epoch: [10][40/200]	Time 0.690 (0.777)	Data 0.001 (0.069)	Loss 1.421 (0.688)
Epoch: [10][60/200]	Time 0.690 (0.751)	Data 0.000 (0.046)	Loss 1.719 (0.904)
Epoch: [10][80/200]	Time 0.688 (0.758)	Data 0.001 (0.056)	Loss 1.156 (0.990)
Epoch: [10][100/200]	Time 0.689 (0.764)	Data 0.000 (0.064)	Loss 1.307 (1.053)
Epoch: [10][120/200]	Time 0.690 (0.752)	Data 0.000 (0.053)	Loss 1.305 (1.081)
Epoch: [10][140/200]	Time 0.696 (0.757)	Data 0.001 (0.059)	Loss 1.047 (1.106)
Epoch: [10][160/200]	Time 0.692 (0.761)	Data 0.001 (0.062)	Loss 1.305 (1.125)
Epoch: [10][180/200]	Time 0.785 (0.755)	Data 0.000 (0.056)	Loss 1.473 (1.134)
Epoch: [10][200/200]	Time 0.695 (0.758)	Data 0.000 (0.059)	Loss 1.415 (1.143)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.227 (0.265)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.230 (0.251)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.37862825393677
==> Statistics for epoch 11: 1041 clusters
Epoch: [11][20/200]	Time 0.696 (0.741)	Data 0.001 (0.051)	Loss 0.204 (0.238)
Epoch: [11][40/200]	Time 0.705 (0.765)	Data 0.001 (0.068)	Loss 1.209 (0.434)
Epoch: [11][60/200]	Time 0.780 (0.745)	Data 0.000 (0.046)	Loss 1.441 (0.704)
Epoch: [11][80/200]	Time 0.697 (0.754)	Data 0.001 (0.055)	Loss 1.336 (0.842)
Epoch: [11][100/200]	Time 0.695 (0.761)	Data 0.001 (0.061)	Loss 1.597 (0.950)
Epoch: [11][120/200]	Time 0.694 (0.751)	Data 0.000 (0.051)	Loss 1.265 (1.003)
Epoch: [11][140/200]	Time 0.696 (0.757)	Data 0.001 (0.055)	Loss 1.456 (1.043)
Epoch: [11][160/200]	Time 0.693 (0.750)	Data 0.000 (0.048)	Loss 1.194 (1.078)
Epoch: [11][180/200]	Time 0.696 (0.755)	Data 0.001 (0.053)	Loss 1.302 (1.105)
Epoch: [11][200/200]	Time 0.696 (0.759)	Data 0.001 (0.057)	Loss 1.409 (1.126)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.265)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.233 (0.250)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.207749366760254
==> Statistics for epoch 12: 1033 clusters
Epoch: [12][20/200]	Time 0.691 (0.752)	Data 0.001 (0.051)	Loss 0.247 (0.258)
Epoch: [12][40/200]	Time 0.690 (0.773)	Data 0.001 (0.069)	Loss 1.550 (0.477)
Epoch: [12][60/200]	Time 0.693 (0.750)	Data 0.000 (0.047)	Loss 1.039 (0.735)
Epoch: [12][80/200]	Time 0.699 (0.759)	Data 0.001 (0.056)	Loss 1.182 (0.860)
Epoch: [12][100/200]	Time 0.693 (0.768)	Data 0.001 (0.063)	Loss 1.130 (0.947)
Epoch: [12][120/200]	Time 0.694 (0.757)	Data 0.001 (0.053)	Loss 1.269 (1.010)
Epoch: [12][140/200]	Time 0.692 (0.761)	Data 0.001 (0.058)	Loss 1.197 (1.051)
Epoch: [12][160/200]	Time 0.692 (0.753)	Data 0.000 (0.050)	Loss 1.290 (1.084)
Epoch: [12][180/200]	Time 0.700 (0.757)	Data 0.001 (0.055)	Loss 1.317 (1.112)
Epoch: [12][200/200]	Time 0.698 (0.760)	Data 0.001 (0.058)	Loss 0.718 (1.124)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.265)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.230 (0.250)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.23470902442932
==> Statistics for epoch 13: 1055 clusters
Epoch: [13][20/200]	Time 0.692 (0.757)	Data 0.001 (0.056)	Loss 0.163 (0.255)
Epoch: [13][40/200]	Time 0.692 (0.770)	Data 0.000 (0.071)	Loss 1.682 (0.491)
Epoch: [13][60/200]	Time 0.691 (0.745)	Data 0.000 (0.047)	Loss 1.308 (0.730)
Epoch: [13][80/200]	Time 0.693 (0.758)	Data 0.000 (0.060)	Loss 1.159 (0.855)
Epoch: [13][100/200]	Time 0.688 (0.762)	Data 0.001 (0.065)	Loss 1.134 (0.958)
Epoch: [13][120/200]	Time 0.690 (0.751)	Data 0.001 (0.054)	Loss 1.188 (1.009)
Epoch: [13][140/200]	Time 0.695 (0.756)	Data 0.001 (0.059)	Loss 1.241 (1.057)
Epoch: [13][160/200]	Time 0.693 (0.748)	Data 0.000 (0.052)	Loss 1.300 (1.084)
Epoch: [13][180/200]	Time 0.698 (0.753)	Data 0.000 (0.055)	Loss 1.061 (1.098)
Epoch: [13][200/200]	Time 0.694 (0.757)	Data 0.001 (0.058)	Loss 1.148 (1.119)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.228 (0.263)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.238 (0.250)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.894920349121094
==> Statistics for epoch 14: 1058 clusters
Epoch: [14][20/200]	Time 0.686 (0.742)	Data 0.001 (0.054)	Loss 0.365 (0.254)
Epoch: [14][40/200]	Time 0.697 (0.760)	Data 0.001 (0.071)	Loss 1.562 (0.413)
Epoch: [14][60/200]	Time 0.693 (0.738)	Data 0.000 (0.048)	Loss 1.353 (0.712)
Epoch: [14][80/200]	Time 0.699 (0.749)	Data 0.001 (0.057)	Loss 1.064 (0.856)
Epoch: [14][100/200]	Time 2.414 (0.755)	Data 1.694 (0.063)	Loss 1.291 (0.927)
Epoch: [14][120/200]	Time 0.694 (0.747)	Data 0.001 (0.053)	Loss 0.968 (0.981)
Epoch: [14][140/200]	Time 0.696 (0.756)	Data 0.002 (0.059)	Loss 0.930 (1.028)
Epoch: [14][160/200]	Time 0.699 (0.749)	Data 0.000 (0.052)	Loss 0.864 (1.065)
Epoch: [14][180/200]	Time 0.690 (0.752)	Data 0.001 (0.055)	Loss 1.270 (1.088)
Epoch: [14][200/200]	Time 0.704 (0.756)	Data 0.001 (0.058)	Loss 1.197 (1.098)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.261)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.232 (0.249)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.42776679992676
==> Statistics for epoch 15: 1023 clusters
Epoch: [15][20/200]	Time 0.811 (0.759)	Data 0.001 (0.052)	Loss 0.160 (0.238)
Epoch: [15][40/200]	Time 0.694 (0.773)	Data 0.001 (0.070)	Loss 1.227 (0.424)
Epoch: [15][60/200]	Time 0.691 (0.750)	Data 0.000 (0.047)	Loss 1.637 (0.679)
Epoch: [15][80/200]	Time 0.693 (0.760)	Data 0.001 (0.057)	Loss 1.194 (0.784)
Epoch: [15][100/200]	Time 0.689 (0.766)	Data 0.001 (0.063)	Loss 1.127 (0.861)
Epoch: [15][120/200]	Time 0.688 (0.754)	Data 0.000 (0.053)	Loss 1.131 (0.913)
Epoch: [15][140/200]	Time 0.691 (0.760)	Data 0.001 (0.058)	Loss 0.865 (0.936)
Epoch: [15][160/200]	Time 0.692 (0.763)	Data 0.001 (0.062)	Loss 1.178 (0.960)
Epoch: [15][180/200]	Time 0.691 (0.755)	Data 0.000 (0.055)	Loss 0.987 (0.982)
Epoch: [15][200/200]	Time 0.695 (0.758)	Data 0.001 (0.058)	Loss 0.952 (1.007)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.231 (0.263)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.232 (0.251)	Data 0.001 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.26889085769653
==> Statistics for epoch 16: 1072 clusters
Epoch: [16][20/200]	Time 0.689 (0.743)	Data 0.001 (0.053)	Loss 0.164 (0.225)
Epoch: [16][40/200]	Time 0.693 (0.763)	Data 0.001 (0.072)	Loss 1.240 (0.382)
Epoch: [16][60/200]	Time 0.695 (0.742)	Data 0.000 (0.048)	Loss 1.085 (0.658)
Epoch: [16][80/200]	Time 0.692 (0.753)	Data 0.001 (0.059)	Loss 1.287 (0.794)
Epoch: [16][100/200]	Time 2.496 (0.759)	Data 1.772 (0.065)	Loss 1.721 (0.887)
Epoch: [16][120/200]	Time 0.696 (0.751)	Data 0.001 (0.054)	Loss 1.084 (0.938)
Epoch: [16][140/200]	Time 0.689 (0.758)	Data 0.001 (0.060)	Loss 1.606 (0.986)
Epoch: [16][160/200]	Time 0.692 (0.751)	Data 0.000 (0.052)	Loss 1.189 (1.008)
Epoch: [16][180/200]	Time 0.697 (0.755)	Data 0.001 (0.056)	Loss 0.969 (1.024)
Epoch: [16][200/200]	Time 0.693 (0.758)	Data 0.001 (0.059)	Loss 0.945 (1.041)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.268)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.398 (0.253)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.816431283950806
==> Statistics for epoch 17: 1061 clusters
Epoch: [17][20/200]	Time 0.699 (0.757)	Data 0.001 (0.051)	Loss 0.130 (0.212)
Epoch: [17][40/200]	Time 0.696 (0.774)	Data 0.001 (0.067)	Loss 1.168 (0.376)
Epoch: [17][60/200]	Time 0.703 (0.753)	Data 0.000 (0.045)	Loss 1.366 (0.647)
Epoch: [17][80/200]	Time 0.723 (0.763)	Data 0.001 (0.055)	Loss 0.900 (0.790)
Epoch: [17][100/200]	Time 2.606 (0.770)	Data 1.745 (0.061)	Loss 1.359 (0.884)
Epoch: [17][120/200]	Time 0.702 (0.759)	Data 0.001 (0.051)	Loss 1.315 (0.937)
Epoch: [17][140/200]	Time 0.698 (0.764)	Data 0.001 (0.055)	Loss 1.205 (0.968)
Epoch: [17][160/200]	Time 0.694 (0.757)	Data 0.000 (0.049)	Loss 1.161 (0.992)
Epoch: [17][180/200]	Time 0.697 (0.761)	Data 0.001 (0.053)	Loss 1.594 (1.023)
Epoch: [17][200/200]	Time 0.707 (0.765)	Data 0.001 (0.057)	Loss 1.116 (1.036)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.227 (0.263)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.230 (0.250)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.75622010231018
==> Statistics for epoch 18: 1060 clusters
Epoch: [18][20/200]	Time 0.690 (0.754)	Data 0.001 (0.061)	Loss 0.214 (0.208)
Epoch: [18][40/200]	Time 0.694 (0.770)	Data 0.001 (0.074)	Loss 1.452 (0.371)
Epoch: [18][60/200]	Time 0.691 (0.749)	Data 0.000 (0.050)	Loss 1.158 (0.634)
Epoch: [18][80/200]	Time 0.708 (0.761)	Data 0.001 (0.058)	Loss 1.024 (0.757)
Epoch: [18][100/200]	Time 2.641 (0.769)	Data 1.802 (0.064)	Loss 0.999 (0.847)
Epoch: [18][120/200]	Time 0.696 (0.757)	Data 0.001 (0.054)	Loss 1.115 (0.889)
Epoch: [18][140/200]	Time 0.695 (0.763)	Data 0.001 (0.058)	Loss 1.332 (0.924)
Epoch: [18][160/200]	Time 0.695 (0.755)	Data 0.000 (0.051)	Loss 1.155 (0.951)
Epoch: [18][180/200]	Time 0.808 (0.759)	Data 0.001 (0.055)	Loss 1.224 (0.972)
Epoch: [18][200/200]	Time 0.704 (0.762)	Data 0.001 (0.057)	Loss 1.238 (0.992)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.237 (0.263)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.240 (0.250)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.083805084228516
==> Statistics for epoch 19: 1072 clusters
Epoch: [19][20/200]	Time 0.784 (0.757)	Data 0.001 (0.056)	Loss 0.282 (0.224)
Epoch: [19][40/200]	Time 0.696 (0.769)	Data 0.001 (0.071)	Loss 1.148 (0.373)
Epoch: [19][60/200]	Time 0.695 (0.747)	Data 0.000 (0.047)	Loss 1.603 (0.652)
Epoch: [19][80/200]	Time 0.695 (0.758)	Data 0.001 (0.056)	Loss 1.832 (0.779)
Epoch: [19][100/200]	Time 2.449 (0.764)	Data 1.710 (0.062)	Loss 1.187 (0.855)
Epoch: [19][120/200]	Time 0.700 (0.755)	Data 0.001 (0.052)	Loss 1.273 (0.916)
Epoch: [19][140/200]	Time 0.695 (0.760)	Data 0.001 (0.057)	Loss 0.942 (0.951)
Epoch: [19][160/200]	Time 0.691 (0.753)	Data 0.000 (0.050)	Loss 1.151 (0.973)
Epoch: [19][180/200]	Time 0.695 (0.756)	Data 0.001 (0.053)	Loss 0.991 (0.991)
Epoch: [19][200/200]	Time 0.700 (0.760)	Data 0.001 (0.057)	Loss 1.246 (1.010)
Extract Features: [50/367]	Time 0.232 (0.265)	Data 0.000 (0.028)	
Extract Features: [100/367]	Time 0.232 (0.251)	Data 0.000 (0.014)	
Extract Features: [150/367]	Time 0.231 (0.247)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.233 (0.245)	Data 0.000 (0.007)	
Extract Features: [250/367]	Time 0.236 (0.243)	Data 0.001 (0.006)	
Extract Features: [300/367]	Time 0.233 (0.243)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.238 (0.243)	Data 0.000 (0.004)	
Mean AP: 63.0%

 * Finished epoch  19  model mAP: 63.0%  best: 63.0% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.258)	Data 0.001 (0.026)	
Extract Features: [100/128]	Time 0.227 (0.249)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.89892339706421
==> Statistics for epoch 20: 1022 clusters
Epoch: [20][20/200]	Time 0.691 (0.744)	Data 0.001 (0.052)	Loss 0.169 (0.211)
Epoch: [20][40/200]	Time 0.693 (0.763)	Data 0.001 (0.068)	Loss 0.909 (0.389)
Epoch: [20][60/200]	Time 0.692 (0.743)	Data 0.000 (0.046)	Loss 0.850 (0.626)
Epoch: [20][80/200]	Time 0.694 (0.755)	Data 0.001 (0.059)	Loss 1.338 (0.739)
Epoch: [20][100/200]	Time 0.699 (0.762)	Data 0.001 (0.065)	Loss 1.339 (0.822)
Epoch: [20][120/200]	Time 0.696 (0.753)	Data 0.000 (0.055)	Loss 0.971 (0.857)
Epoch: [20][140/200]	Time 0.695 (0.760)	Data 0.001 (0.060)	Loss 0.968 (0.881)
Epoch: [20][160/200]	Time 0.692 (0.764)	Data 0.001 (0.064)	Loss 0.989 (0.907)
Epoch: [20][180/200]	Time 0.695 (0.756)	Data 0.000 (0.057)	Loss 0.998 (0.925)
Epoch: [20][200/200]	Time 0.693 (0.759)	Data 0.001 (0.060)	Loss 0.931 (0.940)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.267)	Data 0.000 (0.032)	
Extract Features: [100/128]	Time 0.236 (0.253)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.023961305618286
==> Statistics for epoch 21: 1080 clusters
Epoch: [21][20/200]	Time 0.693 (0.755)	Data 0.001 (0.057)	Loss 0.182 (0.190)
Epoch: [21][40/200]	Time 0.690 (0.771)	Data 0.001 (0.071)	Loss 1.431 (0.340)
Epoch: [21][60/200]	Time 0.693 (0.747)	Data 0.000 (0.048)	Loss 1.113 (0.575)
Epoch: [21][80/200]	Time 0.700 (0.759)	Data 0.001 (0.058)	Loss 0.999 (0.701)
Epoch: [21][100/200]	Time 2.476 (0.764)	Data 1.758 (0.064)	Loss 0.905 (0.777)
Epoch: [21][120/200]	Time 0.692 (0.752)	Data 0.001 (0.054)	Loss 1.175 (0.831)
Epoch: [21][140/200]	Time 0.696 (0.756)	Data 0.001 (0.058)	Loss 1.223 (0.887)
Epoch: [21][160/200]	Time 0.699 (0.749)	Data 0.000 (0.051)	Loss 1.142 (0.916)
Epoch: [21][180/200]	Time 0.699 (0.754)	Data 0.001 (0.055)	Loss 1.509 (0.940)
Epoch: [21][200/200]	Time 0.703 (0.759)	Data 0.001 (0.058)	Loss 1.183 (0.967)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.237 (0.262)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.243 (0.250)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.01407861709595
==> Statistics for epoch 22: 1083 clusters
Epoch: [22][20/200]	Time 0.689 (0.744)	Data 0.001 (0.054)	Loss 0.136 (0.197)
Epoch: [22][40/200]	Time 0.692 (0.764)	Data 0.001 (0.073)	Loss 1.006 (0.323)
Epoch: [22][60/200]	Time 0.694 (0.742)	Data 0.000 (0.049)	Loss 1.233 (0.582)
Epoch: [22][80/200]	Time 0.691 (0.754)	Data 0.001 (0.059)	Loss 0.797 (0.699)
Epoch: [22][100/200]	Time 2.349 (0.761)	Data 1.621 (0.064)	Loss 0.748 (0.789)
Epoch: [22][120/200]	Time 0.694 (0.752)	Data 0.001 (0.053)	Loss 1.163 (0.847)
Epoch: [22][140/200]	Time 0.693 (0.758)	Data 0.001 (0.059)	Loss 1.278 (0.893)
Epoch: [22][160/200]	Time 0.827 (0.752)	Data 0.000 (0.051)	Loss 0.772 (0.916)
Epoch: [22][180/200]	Time 0.694 (0.757)	Data 0.001 (0.057)	Loss 0.816 (0.935)
Epoch: [22][200/200]	Time 0.712 (0.761)	Data 0.001 (0.060)	Loss 0.747 (0.957)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.263)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.241 (0.250)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.11863422393799
==> Statistics for epoch 23: 1070 clusters
Epoch: [23][20/200]	Time 0.691 (0.746)	Data 0.001 (0.055)	Loss 0.137 (0.187)
Epoch: [23][40/200]	Time 0.699 (0.769)	Data 0.001 (0.073)	Loss 1.049 (0.331)
Epoch: [23][60/200]	Time 0.696 (0.748)	Data 0.000 (0.049)	Loss 1.096 (0.587)
Epoch: [23][80/200]	Time 0.695 (0.761)	Data 0.001 (0.061)	Loss 1.097 (0.693)
Epoch: [23][100/200]	Time 2.478 (0.768)	Data 1.762 (0.067)	Loss 0.720 (0.771)
Epoch: [23][120/200]	Time 0.696 (0.757)	Data 0.001 (0.056)	Loss 1.018 (0.824)
Epoch: [23][140/200]	Time 0.699 (0.760)	Data 0.001 (0.060)	Loss 0.967 (0.867)
Epoch: [23][160/200]	Time 0.698 (0.753)	Data 0.000 (0.052)	Loss 1.591 (0.896)
Epoch: [23][180/200]	Time 0.692 (0.758)	Data 0.001 (0.056)	Loss 1.074 (0.911)
Epoch: [23][200/200]	Time 0.696 (0.761)	Data 0.001 (0.059)	Loss 1.122 (0.926)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.262)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.229 (0.249)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.77452254295349
==> Statistics for epoch 24: 1074 clusters
Epoch: [24][20/200]	Time 0.692 (0.755)	Data 0.001 (0.056)	Loss 0.183 (0.173)
Epoch: [24][40/200]	Time 0.694 (0.771)	Data 0.001 (0.072)	Loss 1.351 (0.340)
Epoch: [24][60/200]	Time 0.691 (0.748)	Data 0.000 (0.048)	Loss 0.907 (0.574)
Epoch: [24][80/200]	Time 0.697 (0.758)	Data 0.001 (0.059)	Loss 1.046 (0.703)
Epoch: [24][100/200]	Time 2.476 (0.765)	Data 1.747 (0.065)	Loss 1.468 (0.794)
Epoch: [24][120/200]	Time 0.693 (0.753)	Data 0.001 (0.054)	Loss 1.316 (0.860)
Epoch: [24][140/200]	Time 0.694 (0.757)	Data 0.001 (0.059)	Loss 0.789 (0.896)
Epoch: [24][160/200]	Time 0.693 (0.749)	Data 0.000 (0.052)	Loss 1.005 (0.917)
Epoch: [24][180/200]	Time 0.691 (0.753)	Data 0.001 (0.055)	Loss 1.032 (0.938)
Epoch: [24][200/200]	Time 0.694 (0.756)	Data 0.001 (0.058)	Loss 1.269 (0.952)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.261)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.227 (0.249)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.8487012386322
==> Statistics for epoch 25: 1065 clusters
Epoch: [25][20/200]	Time 0.689 (0.740)	Data 0.001 (0.048)	Loss 0.164 (0.186)
Epoch: [25][40/200]	Time 0.699 (0.761)	Data 0.001 (0.065)	Loss 1.343 (0.331)
Epoch: [25][60/200]	Time 0.699 (0.740)	Data 0.000 (0.043)	Loss 1.166 (0.560)
Epoch: [25][80/200]	Time 0.710 (0.754)	Data 0.001 (0.057)	Loss 1.184 (0.687)
Epoch: [25][100/200]	Time 2.352 (0.759)	Data 1.627 (0.062)	Loss 1.552 (0.775)
Epoch: [25][120/200]	Time 0.701 (0.749)	Data 0.001 (0.052)	Loss 0.634 (0.826)
Epoch: [25][140/200]	Time 0.693 (0.755)	Data 0.001 (0.056)	Loss 1.373 (0.875)
Epoch: [25][160/200]	Time 0.695 (0.748)	Data 0.000 (0.049)	Loss 1.003 (0.889)
Epoch: [25][180/200]	Time 0.697 (0.753)	Data 0.001 (0.054)	Loss 1.022 (0.903)
Epoch: [25][200/200]	Time 0.702 (0.758)	Data 0.001 (0.058)	Loss 1.210 (0.920)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.265)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.234 (0.252)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.34044575691223
==> Statistics for epoch 26: 1079 clusters
Epoch: [26][20/200]	Time 0.691 (0.750)	Data 0.001 (0.058)	Loss 0.145 (0.187)
Epoch: [26][40/200]	Time 0.694 (0.767)	Data 0.001 (0.071)	Loss 0.608 (0.324)
Epoch: [26][60/200]	Time 0.690 (0.743)	Data 0.000 (0.048)	Loss 1.128 (0.595)
Epoch: [26][80/200]	Time 0.699 (0.754)	Data 0.001 (0.058)	Loss 1.055 (0.722)
Epoch: [26][100/200]	Time 2.451 (0.761)	Data 1.661 (0.063)	Loss 0.778 (0.788)
Epoch: [26][120/200]	Time 0.694 (0.751)	Data 0.001 (0.053)	Loss 0.935 (0.836)
Epoch: [26][140/200]	Time 0.703 (0.756)	Data 0.001 (0.057)	Loss 1.261 (0.877)
Epoch: [26][160/200]	Time 0.692 (0.748)	Data 0.000 (0.050)	Loss 1.405 (0.910)
Epoch: [26][180/200]	Time 0.705 (0.753)	Data 0.001 (0.054)	Loss 1.117 (0.930)
Epoch: [26][200/200]	Time 0.695 (0.757)	Data 0.001 (0.057)	Loss 1.489 (0.953)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.233 (0.261)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.230 (0.248)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.59124183654785
==> Statistics for epoch 27: 1089 clusters
Epoch: [27][20/200]	Time 0.696 (0.747)	Data 0.001 (0.055)	Loss 0.153 (0.208)
Epoch: [27][40/200]	Time 0.698 (0.768)	Data 0.001 (0.070)	Loss 1.461 (0.319)
Epoch: [27][60/200]	Time 0.693 (0.744)	Data 0.001 (0.047)	Loss 1.348 (0.580)
Epoch: [27][80/200]	Time 0.697 (0.757)	Data 0.001 (0.059)	Loss 0.929 (0.709)
Epoch: [27][100/200]	Time 0.695 (0.747)	Data 0.000 (0.048)	Loss 1.028 (0.786)
Epoch: [27][120/200]	Time 0.696 (0.755)	Data 0.001 (0.055)	Loss 1.023 (0.838)
Epoch: [27][140/200]	Time 0.698 (0.762)	Data 0.002 (0.061)	Loss 1.171 (0.877)
Epoch: [27][160/200]	Time 0.697 (0.755)	Data 0.001 (0.053)	Loss 1.148 (0.901)
Epoch: [27][180/200]	Time 0.698 (0.760)	Data 0.001 (0.058)	Loss 1.021 (0.917)
Epoch: [27][200/200]	Time 0.695 (0.754)	Data 0.000 (0.052)	Loss 1.288 (0.936)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.264)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.227 (0.250)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.41336154937744
==> Statistics for epoch 28: 1079 clusters
Epoch: [28][20/200]	Time 0.690 (0.751)	Data 0.001 (0.050)	Loss 0.305 (0.187)
Epoch: [28][40/200]	Time 0.693 (0.776)	Data 0.001 (0.073)	Loss 0.951 (0.326)
Epoch: [28][60/200]	Time 0.692 (0.751)	Data 0.000 (0.049)	Loss 0.999 (0.577)
Epoch: [28][80/200]	Time 0.839 (0.764)	Data 0.001 (0.059)	Loss 1.055 (0.681)
Epoch: [28][100/200]	Time 2.706 (0.771)	Data 1.835 (0.066)	Loss 0.703 (0.757)
Epoch: [28][120/200]	Time 0.701 (0.758)	Data 0.001 (0.055)	Loss 0.713 (0.817)
Epoch: [28][140/200]	Time 0.696 (0.764)	Data 0.001 (0.061)	Loss 1.365 (0.865)
Epoch: [28][160/200]	Time 0.695 (0.756)	Data 0.000 (0.053)	Loss 0.852 (0.893)
Epoch: [28][180/200]	Time 0.697 (0.760)	Data 0.001 (0.058)	Loss 1.025 (0.917)
Epoch: [28][200/200]	Time 0.699 (0.764)	Data 0.001 (0.061)	Loss 0.867 (0.927)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.265)	Data 0.001 (0.031)	
Extract Features: [100/128]	Time 0.230 (0.249)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.99564266204834
==> Statistics for epoch 29: 1084 clusters
Epoch: [29][20/200]	Time 0.689 (0.752)	Data 0.001 (0.053)	Loss 0.113 (0.191)
Epoch: [29][40/200]	Time 0.692 (0.767)	Data 0.001 (0.071)	Loss 0.760 (0.318)
Epoch: [29][60/200]	Time 0.696 (0.746)	Data 0.000 (0.048)	Loss 1.588 (0.552)
Epoch: [29][80/200]	Time 0.699 (0.757)	Data 0.001 (0.059)	Loss 0.833 (0.692)
Epoch: [29][100/200]	Time 2.472 (0.762)	Data 1.737 (0.065)	Loss 0.923 (0.745)
Epoch: [29][120/200]	Time 0.699 (0.751)	Data 0.001 (0.054)	Loss 0.630 (0.797)
Epoch: [29][140/200]	Time 0.700 (0.758)	Data 0.001 (0.059)	Loss 0.915 (0.837)
Epoch: [29][160/200]	Time 0.695 (0.752)	Data 0.000 (0.052)	Loss 0.807 (0.860)
Epoch: [29][180/200]	Time 0.710 (0.758)	Data 0.001 (0.056)	Loss 0.918 (0.893)
Epoch: [29][200/200]	Time 0.700 (0.761)	Data 0.001 (0.059)	Loss 1.192 (0.909)
Extract Features: [50/367]	Time 0.251 (0.267)	Data 0.000 (0.029)	
Extract Features: [100/367]	Time 0.230 (0.251)	Data 0.000 (0.015)	
Extract Features: [150/367]	Time 0.231 (0.247)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.227 (0.244)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.234 (0.243)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.233 (0.242)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.236 (0.241)	Data 0.000 (0.004)	
Mean AP: 70.0%

 * Finished epoch  29  model mAP: 70.0%  best: 70.0% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.226 (0.272)	Data 0.000 (0.039)	
Extract Features: [100/128]	Time 0.350 (0.253)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.454177141189575
==> Statistics for epoch 30: 1080 clusters
Epoch: [30][20/200]	Time 0.695 (0.756)	Data 0.001 (0.055)	Loss 0.143 (0.179)
Epoch: [30][40/200]	Time 0.692 (0.774)	Data 0.001 (0.074)	Loss 1.166 (0.308)
Epoch: [30][60/200]	Time 0.692 (0.749)	Data 0.000 (0.050)	Loss 1.103 (0.554)
Epoch: [30][80/200]	Time 0.696 (0.759)	Data 0.001 (0.059)	Loss 1.013 (0.657)
Epoch: [30][100/200]	Time 2.623 (0.767)	Data 1.883 (0.066)	Loss 1.087 (0.746)
Epoch: [30][120/200]	Time 0.696 (0.756)	Data 0.001 (0.055)	Loss 1.028 (0.795)
Epoch: [30][140/200]	Time 0.701 (0.760)	Data 0.001 (0.060)	Loss 1.537 (0.842)
Epoch: [30][160/200]	Time 0.701 (0.752)	Data 0.000 (0.052)	Loss 1.080 (0.867)
Epoch: [30][180/200]	Time 0.704 (0.757)	Data 0.001 (0.057)	Loss 1.223 (0.891)
Epoch: [30][200/200]	Time 0.697 (0.760)	Data 0.001 (0.060)	Loss 1.068 (0.912)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.230 (0.268)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.234 (0.252)	Data 0.001 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.48529648780823
==> Statistics for epoch 31: 1075 clusters
Epoch: [31][20/200]	Time 0.699 (0.743)	Data 0.001 (0.051)	Loss 0.236 (0.183)
Epoch: [31][40/200]	Time 0.691 (0.760)	Data 0.001 (0.069)	Loss 1.019 (0.296)
Epoch: [31][60/200]	Time 0.690 (0.737)	Data 0.000 (0.046)	Loss 1.170 (0.570)
Epoch: [31][80/200]	Time 0.694 (0.748)	Data 0.001 (0.056)	Loss 1.138 (0.685)
Epoch: [31][100/200]	Time 2.603 (0.757)	Data 1.879 (0.064)	Loss 1.223 (0.766)
Epoch: [31][120/200]	Time 0.698 (0.749)	Data 0.001 (0.053)	Loss 1.128 (0.799)
Epoch: [31][140/200]	Time 0.705 (0.755)	Data 0.001 (0.059)	Loss 0.858 (0.835)
Epoch: [31][160/200]	Time 0.702 (0.750)	Data 0.000 (0.052)	Loss 1.002 (0.868)
Epoch: [31][180/200]	Time 0.696 (0.754)	Data 0.001 (0.056)	Loss 0.867 (0.895)
Epoch: [31][200/200]	Time 0.701 (0.758)	Data 0.002 (0.059)	Loss 0.890 (0.913)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.231 (0.261)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.227 (0.248)	Data 0.001 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.883899211883545
==> Statistics for epoch 32: 1075 clusters
Epoch: [32][20/200]	Time 0.689 (0.754)	Data 0.001 (0.055)	Loss 0.158 (0.166)
Epoch: [32][40/200]	Time 0.691 (0.772)	Data 0.001 (0.076)	Loss 1.193 (0.318)
Epoch: [32][60/200]	Time 0.689 (0.748)	Data 0.000 (0.051)	Loss 1.250 (0.545)
Epoch: [32][80/200]	Time 0.694 (0.758)	Data 0.001 (0.062)	Loss 1.113 (0.670)
Epoch: [32][100/200]	Time 2.568 (0.766)	Data 1.856 (0.068)	Loss 1.028 (0.744)
Epoch: [32][120/200]	Time 0.691 (0.754)	Data 0.001 (0.057)	Loss 1.295 (0.806)
Epoch: [32][140/200]	Time 0.693 (0.757)	Data 0.001 (0.061)	Loss 1.215 (0.842)
Epoch: [32][160/200]	Time 0.692 (0.749)	Data 0.000 (0.053)	Loss 1.490 (0.870)
Epoch: [32][180/200]	Time 0.816 (0.754)	Data 0.002 (0.057)	Loss 1.277 (0.897)
Epoch: [32][200/200]	Time 0.695 (0.757)	Data 0.001 (0.060)	Loss 1.209 (0.912)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.225 (0.264)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.228 (0.248)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.37938666343689
==> Statistics for epoch 33: 1086 clusters
Epoch: [33][20/200]	Time 0.688 (0.746)	Data 0.001 (0.059)	Loss 0.155 (0.194)
Epoch: [33][40/200]	Time 0.690 (0.767)	Data 0.001 (0.076)	Loss 1.093 (0.323)
Epoch: [33][60/200]	Time 0.690 (0.745)	Data 0.000 (0.051)	Loss 0.876 (0.557)
Epoch: [33][80/200]	Time 0.693 (0.755)	Data 0.001 (0.059)	Loss 1.402 (0.691)
Epoch: [33][100/200]	Time 2.399 (0.761)	Data 1.662 (0.064)	Loss 1.337 (0.760)
Epoch: [33][120/200]	Time 0.694 (0.751)	Data 0.001 (0.054)	Loss 1.100 (0.799)
Epoch: [33][140/200]	Time 0.691 (0.756)	Data 0.001 (0.058)	Loss 0.738 (0.843)
Epoch: [33][160/200]	Time 0.694 (0.748)	Data 0.000 (0.051)	Loss 1.029 (0.873)
Epoch: [33][180/200]	Time 0.690 (0.752)	Data 0.001 (0.054)	Loss 1.126 (0.887)
Epoch: [33][200/200]	Time 0.694 (0.755)	Data 0.001 (0.057)	Loss 0.942 (0.906)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.228 (0.264)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.230 (0.249)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.94814729690552
==> Statistics for epoch 34: 1084 clusters
Epoch: [34][20/200]	Time 0.689 (0.753)	Data 0.001 (0.049)	Loss 0.098 (0.174)
Epoch: [34][40/200]	Time 0.690 (0.770)	Data 0.001 (0.069)	Loss 1.039 (0.311)
Epoch: [34][60/200]	Time 0.695 (0.747)	Data 0.000 (0.046)	Loss 1.127 (0.544)
Epoch: [34][80/200]	Time 0.694 (0.757)	Data 0.001 (0.057)	Loss 0.815 (0.663)
Epoch: [34][100/200]	Time 2.528 (0.764)	Data 1.795 (0.063)	Loss 1.550 (0.756)
Epoch: [34][120/200]	Time 0.695 (0.753)	Data 0.001 (0.053)	Loss 1.147 (0.807)
Epoch: [34][140/200]	Time 0.697 (0.759)	Data 0.001 (0.059)	Loss 1.145 (0.833)
Epoch: [34][160/200]	Time 0.700 (0.752)	Data 0.000 (0.051)	Loss 0.856 (0.852)
Epoch: [34][180/200]	Time 0.697 (0.756)	Data 0.001 (0.056)	Loss 1.090 (0.872)
Epoch: [34][200/200]	Time 0.698 (0.759)	Data 0.002 (0.059)	Loss 0.729 (0.891)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.263)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.237 (0.249)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.970627546310425
==> Statistics for epoch 35: 1081 clusters
Epoch: [35][20/200]	Time 0.689 (0.747)	Data 0.001 (0.054)	Loss 0.134 (0.173)
Epoch: [35][40/200]	Time 0.689 (0.771)	Data 0.001 (0.074)	Loss 1.202 (0.310)
Epoch: [35][60/200]	Time 0.696 (0.748)	Data 0.000 (0.050)	Loss 0.982 (0.557)
Epoch: [35][80/200]	Time 0.696 (0.760)	Data 0.001 (0.060)	Loss 1.212 (0.689)
Epoch: [35][100/200]	Time 2.497 (0.768)	Data 1.771 (0.066)	Loss 0.911 (0.759)
Epoch: [35][120/200]	Time 0.723 (0.758)	Data 0.001 (0.055)	Loss 1.027 (0.819)
Epoch: [35][140/200]	Time 0.696 (0.763)	Data 0.001 (0.060)	Loss 1.383 (0.856)
Epoch: [35][160/200]	Time 0.696 (0.756)	Data 0.000 (0.053)	Loss 0.973 (0.894)
Epoch: [35][180/200]	Time 0.705 (0.760)	Data 0.001 (0.057)	Loss 1.126 (0.914)
Epoch: [35][200/200]	Time 0.692 (0.762)	Data 0.001 (0.060)	Loss 0.735 (0.928)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.225 (0.264)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.229 (0.249)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.724019050598145
==> Statistics for epoch 36: 1082 clusters
Epoch: [36][20/200]	Time 0.687 (0.744)	Data 0.001 (0.047)	Loss 0.314 (0.176)
Epoch: [36][40/200]	Time 0.686 (0.763)	Data 0.001 (0.069)	Loss 1.114 (0.308)
Epoch: [36][60/200]	Time 0.690 (0.740)	Data 0.000 (0.046)	Loss 0.948 (0.536)
Epoch: [36][80/200]	Time 0.692 (0.749)	Data 0.001 (0.056)	Loss 1.307 (0.665)
Epoch: [36][100/200]	Time 2.524 (0.757)	Data 1.782 (0.062)	Loss 0.838 (0.731)
Epoch: [36][120/200]	Time 0.698 (0.747)	Data 0.001 (0.052)	Loss 1.186 (0.790)
Epoch: [36][140/200]	Time 0.701 (0.753)	Data 0.001 (0.057)	Loss 1.004 (0.831)
Epoch: [36][160/200]	Time 0.690 (0.745)	Data 0.000 (0.050)	Loss 1.064 (0.860)
Epoch: [36][180/200]	Time 0.692 (0.749)	Data 0.001 (0.055)	Loss 1.201 (0.878)
Epoch: [36][200/200]	Time 0.755 (0.754)	Data 0.001 (0.059)	Loss 0.900 (0.897)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.233 (0.262)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.230 (0.249)	Data 0.001 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.44606828689575
==> Statistics for epoch 37: 1084 clusters
Epoch: [37][20/200]	Time 0.701 (0.746)	Data 0.001 (0.055)	Loss 0.250 (0.179)
Epoch: [37][40/200]	Time 0.687 (0.761)	Data 0.001 (0.071)	Loss 1.661 (0.327)
Epoch: [37][60/200]	Time 0.690 (0.737)	Data 0.000 (0.048)	Loss 1.014 (0.558)
Epoch: [37][80/200]	Time 0.692 (0.748)	Data 0.001 (0.056)	Loss 1.463 (0.678)
Epoch: [37][100/200]	Time 2.524 (0.756)	Data 1.795 (0.063)	Loss 1.320 (0.745)
Epoch: [37][120/200]	Time 0.694 (0.747)	Data 0.001 (0.053)	Loss 1.231 (0.799)
Epoch: [37][140/200]	Time 0.694 (0.753)	Data 0.001 (0.058)	Loss 1.128 (0.838)
Epoch: [37][160/200]	Time 0.700 (0.747)	Data 0.000 (0.051)	Loss 1.563 (0.859)
Epoch: [37][180/200]	Time 0.704 (0.753)	Data 0.001 (0.056)	Loss 1.226 (0.881)
Epoch: [37][200/200]	Time 0.700 (0.758)	Data 0.001 (0.059)	Loss 1.008 (0.900)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.228 (0.265)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.231 (0.253)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.38360667228699
==> Statistics for epoch 38: 1085 clusters
Epoch: [38][20/200]	Time 0.689 (0.763)	Data 0.001 (0.058)	Loss 0.174 (0.163)
Epoch: [38][40/200]	Time 0.690 (0.777)	Data 0.001 (0.073)	Loss 0.895 (0.320)
Epoch: [38][60/200]	Time 0.693 (0.751)	Data 0.000 (0.049)	Loss 1.124 (0.535)
Epoch: [38][80/200]	Time 0.694 (0.762)	Data 0.001 (0.059)	Loss 1.154 (0.661)
Epoch: [38][100/200]	Time 2.351 (0.766)	Data 1.608 (0.063)	Loss 0.912 (0.708)
Epoch: [38][120/200]	Time 0.698 (0.756)	Data 0.001 (0.053)	Loss 1.352 (0.761)
Epoch: [38][140/200]	Time 0.696 (0.760)	Data 0.001 (0.057)	Loss 1.085 (0.813)
Epoch: [38][160/200]	Time 0.700 (0.753)	Data 0.000 (0.050)	Loss 0.883 (0.840)
Epoch: [38][180/200]	Time 0.693 (0.757)	Data 0.001 (0.054)	Loss 1.126 (0.860)
Epoch: [38][200/200]	Time 0.701 (0.760)	Data 0.001 (0.057)	Loss 1.182 (0.880)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.231 (0.262)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.344 (0.250)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.93404936790466
==> Statistics for epoch 39: 1086 clusters
Epoch: [39][20/200]	Time 0.787 (0.760)	Data 0.001 (0.059)	Loss 0.101 (0.175)
Epoch: [39][40/200]	Time 0.701 (0.776)	Data 0.001 (0.076)	Loss 0.875 (0.311)
Epoch: [39][60/200]	Time 0.692 (0.751)	Data 0.000 (0.051)	Loss 0.842 (0.533)
Epoch: [39][80/200]	Time 0.696 (0.762)	Data 0.001 (0.061)	Loss 0.639 (0.653)
Epoch: [39][100/200]	Time 2.584 (0.768)	Data 1.856 (0.067)	Loss 0.835 (0.729)
Epoch: [39][120/200]	Time 0.701 (0.757)	Data 0.001 (0.056)	Loss 0.907 (0.777)
Epoch: [39][140/200]	Time 0.698 (0.761)	Data 0.001 (0.061)	Loss 0.656 (0.824)
Epoch: [39][160/200]	Time 0.694 (0.754)	Data 0.000 (0.053)	Loss 1.024 (0.854)
Epoch: [39][180/200]	Time 0.700 (0.758)	Data 0.001 (0.058)	Loss 1.288 (0.870)
Epoch: [39][200/200]	Time 0.699 (0.762)	Data 0.001 (0.061)	Loss 1.060 (0.889)
Extract Features: [50/367]	Time 0.230 (0.268)	Data 0.000 (0.031)	
Extract Features: [100/367]	Time 0.232 (0.252)	Data 0.000 (0.016)	
Extract Features: [150/367]	Time 0.231 (0.247)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.378 (0.245)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.231 (0.243)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.231 (0.242)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.239 (0.242)	Data 0.001 (0.005)	
Mean AP: 70.0%

 * Finished epoch  39  model mAP: 70.0%  best: 70.0% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.228 (0.263)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.360 (0.249)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.549920082092285
==> Statistics for epoch 40: 1080 clusters
Epoch: [40][20/200]	Time 0.689 (0.741)	Data 0.001 (0.049)	Loss 0.107 (0.170)
Epoch: [40][40/200]	Time 0.700 (0.759)	Data 0.001 (0.065)	Loss 1.183 (0.300)
Epoch: [40][60/200]	Time 0.693 (0.742)	Data 0.000 (0.043)	Loss 0.991 (0.537)
Epoch: [40][80/200]	Time 0.693 (0.755)	Data 0.001 (0.054)	Loss 1.111 (0.669)
Epoch: [40][100/200]	Time 2.523 (0.763)	Data 1.810 (0.061)	Loss 0.710 (0.741)
Epoch: [40][120/200]	Time 0.709 (0.753)	Data 0.001 (0.051)	Loss 0.822 (0.798)
Epoch: [40][140/200]	Time 0.695 (0.760)	Data 0.001 (0.058)	Loss 1.123 (0.826)
Epoch: [40][160/200]	Time 0.698 (0.753)	Data 0.000 (0.051)	Loss 1.084 (0.854)
Epoch: [40][180/200]	Time 0.691 (0.757)	Data 0.001 (0.055)	Loss 1.415 (0.875)
Epoch: [40][200/200]	Time 0.703 (0.759)	Data 0.001 (0.058)	Loss 1.129 (0.888)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.227 (0.262)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.236 (0.251)	Data 0.001 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.75130558013916
==> Statistics for epoch 41: 1092 clusters
Epoch: [41][20/200]	Time 0.690 (0.747)	Data 0.001 (0.050)	Loss 0.167 (0.164)
Epoch: [41][40/200]	Time 0.695 (0.763)	Data 0.001 (0.066)	Loss 0.808 (0.314)
Epoch: [41][60/200]	Time 0.697 (0.743)	Data 0.001 (0.044)	Loss 0.704 (0.532)
Epoch: [41][80/200]	Time 0.696 (0.755)	Data 0.001 (0.055)	Loss 0.840 (0.652)
Epoch: [41][100/200]	Time 0.693 (0.744)	Data 0.000 (0.045)	Loss 1.043 (0.738)
Epoch: [41][120/200]	Time 0.696 (0.751)	Data 0.001 (0.051)	Loss 0.964 (0.795)
Epoch: [41][140/200]	Time 0.691 (0.756)	Data 0.001 (0.057)	Loss 0.838 (0.832)
Epoch: [41][160/200]	Time 0.696 (0.749)	Data 0.001 (0.050)	Loss 0.902 (0.855)
Epoch: [41][180/200]	Time 0.697 (0.752)	Data 0.001 (0.054)	Loss 1.139 (0.875)
Epoch: [41][200/200]	Time 0.692 (0.746)	Data 0.000 (0.049)	Loss 0.824 (0.887)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.239 (0.265)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.233 (0.252)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.06705117225647
==> Statistics for epoch 42: 1082 clusters
Epoch: [42][20/200]	Time 0.689 (0.746)	Data 0.001 (0.055)	Loss 0.134 (0.165)
Epoch: [42][40/200]	Time 0.695 (0.764)	Data 0.001 (0.073)	Loss 0.638 (0.312)
Epoch: [42][60/200]	Time 0.697 (0.741)	Data 0.000 (0.049)	Loss 1.265 (0.520)
Epoch: [42][80/200]	Time 0.701 (0.753)	Data 0.001 (0.059)	Loss 0.976 (0.640)
Epoch: [42][100/200]	Time 2.403 (0.759)	Data 1.630 (0.064)	Loss 0.947 (0.722)
Epoch: [42][120/200]	Time 0.696 (0.749)	Data 0.001 (0.053)	Loss 1.099 (0.784)
Epoch: [42][140/200]	Time 0.697 (0.756)	Data 0.001 (0.058)	Loss 1.117 (0.817)
Epoch: [42][160/200]	Time 0.701 (0.750)	Data 0.000 (0.051)	Loss 0.859 (0.849)
Epoch: [42][180/200]	Time 0.696 (0.755)	Data 0.001 (0.054)	Loss 0.717 (0.868)
Epoch: [42][200/200]	Time 0.708 (0.760)	Data 0.001 (0.058)	Loss 1.382 (0.884)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.269)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.231 (0.253)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.30507564544678
==> Statistics for epoch 43: 1084 clusters
Epoch: [43][20/200]	Time 0.690 (0.752)	Data 0.002 (0.054)	Loss 0.127 (0.165)
Epoch: [43][40/200]	Time 0.700 (0.766)	Data 0.001 (0.071)	Loss 0.651 (0.305)
Epoch: [43][60/200]	Time 0.690 (0.743)	Data 0.000 (0.047)	Loss 0.942 (0.535)
Epoch: [43][80/200]	Time 0.806 (0.758)	Data 0.001 (0.058)	Loss 0.961 (0.659)
Epoch: [43][100/200]	Time 2.528 (0.764)	Data 1.812 (0.065)	Loss 0.969 (0.731)
Epoch: [43][120/200]	Time 0.700 (0.755)	Data 0.001 (0.054)	Loss 1.017 (0.787)
Epoch: [43][140/200]	Time 0.698 (0.761)	Data 0.001 (0.059)	Loss 0.782 (0.827)
Epoch: [43][160/200]	Time 0.693 (0.754)	Data 0.000 (0.052)	Loss 0.972 (0.854)
Epoch: [43][180/200]	Time 0.699 (0.758)	Data 0.001 (0.055)	Loss 0.987 (0.876)
Epoch: [43][200/200]	Time 0.710 (0.762)	Data 0.001 (0.058)	Loss 0.810 (0.893)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.227 (0.266)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.237 (0.251)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.921876192092896
==> Statistics for epoch 44: 1083 clusters
Epoch: [44][20/200]	Time 0.695 (0.768)	Data 0.001 (0.055)	Loss 0.175 (0.196)
Epoch: [44][40/200]	Time 0.690 (0.782)	Data 0.001 (0.076)	Loss 1.067 (0.325)
Epoch: [44][60/200]	Time 0.693 (0.756)	Data 0.000 (0.051)	Loss 1.191 (0.581)
Epoch: [44][80/200]	Time 0.692 (0.765)	Data 0.001 (0.060)	Loss 0.807 (0.686)
Epoch: [44][100/200]	Time 2.458 (0.769)	Data 1.724 (0.066)	Loss 1.007 (0.755)
Epoch: [44][120/200]	Time 0.696 (0.758)	Data 0.001 (0.055)	Loss 1.038 (0.799)
Epoch: [44][140/200]	Time 0.699 (0.761)	Data 0.001 (0.059)	Loss 0.876 (0.833)
Epoch: [44][160/200]	Time 0.695 (0.754)	Data 0.000 (0.051)	Loss 1.267 (0.866)
Epoch: [44][180/200]	Time 0.697 (0.758)	Data 0.001 (0.056)	Loss 1.020 (0.884)
Epoch: [44][200/200]	Time 0.696 (0.762)	Data 0.001 (0.060)	Loss 0.829 (0.896)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.226 (0.267)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.230 (0.251)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.681679248809814
==> Statistics for epoch 45: 1094 clusters
Epoch: [45][20/200]	Time 0.697 (0.752)	Data 0.001 (0.056)	Loss 0.185 (0.165)
Epoch: [45][40/200]	Time 0.690 (0.765)	Data 0.001 (0.069)	Loss 1.109 (0.300)
Epoch: [45][60/200]	Time 0.707 (0.742)	Data 0.004 (0.046)	Loss 0.943 (0.541)
Epoch: [45][80/200]	Time 0.696 (0.754)	Data 0.001 (0.056)	Loss 1.277 (0.650)
Epoch: [45][100/200]	Time 0.693 (0.743)	Data 0.000 (0.045)	Loss 1.135 (0.723)
Epoch: [45][120/200]	Time 0.696 (0.750)	Data 0.001 (0.053)	Loss 0.646 (0.772)
Epoch: [45][140/200]	Time 0.691 (0.756)	Data 0.001 (0.058)	Loss 1.194 (0.817)
Epoch: [45][160/200]	Time 0.695 (0.748)	Data 0.001 (0.051)	Loss 1.038 (0.849)
Epoch: [45][180/200]	Time 0.694 (0.754)	Data 0.001 (0.056)	Loss 0.861 (0.869)
Epoch: [45][200/200]	Time 0.692 (0.750)	Data 0.000 (0.051)	Loss 1.358 (0.886)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.234 (0.266)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.231 (0.251)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.73172402381897
==> Statistics for epoch 46: 1085 clusters
Epoch: [46][20/200]	Time 0.697 (0.740)	Data 0.001 (0.049)	Loss 0.189 (0.179)
Epoch: [46][40/200]	Time 0.699 (0.761)	Data 0.001 (0.068)	Loss 0.863 (0.312)
Epoch: [46][60/200]	Time 0.704 (0.740)	Data 0.000 (0.046)	Loss 1.104 (0.540)
Epoch: [46][80/200]	Time 0.698 (0.753)	Data 0.001 (0.056)	Loss 0.917 (0.670)
Epoch: [46][100/200]	Time 2.552 (0.761)	Data 1.810 (0.063)	Loss 1.410 (0.750)
Epoch: [46][120/200]	Time 0.697 (0.752)	Data 0.001 (0.053)	Loss 1.284 (0.796)
Epoch: [46][140/200]	Time 0.694 (0.758)	Data 0.001 (0.057)	Loss 1.135 (0.825)
Epoch: [46][160/200]	Time 0.689 (0.750)	Data 0.000 (0.050)	Loss 0.871 (0.862)
Epoch: [46][180/200]	Time 0.699 (0.756)	Data 0.001 (0.055)	Loss 1.007 (0.882)
Epoch: [46][200/200]	Time 0.706 (0.759)	Data 0.001 (0.058)	Loss 0.905 (0.892)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.229 (0.263)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.228 (0.249)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.53077793121338
==> Statistics for epoch 47: 1084 clusters
Epoch: [47][20/200]	Time 0.696 (0.748)	Data 0.002 (0.050)	Loss 0.174 (0.178)
Epoch: [47][40/200]	Time 0.691 (0.766)	Data 0.001 (0.071)	Loss 0.918 (0.327)
Epoch: [47][60/200]	Time 0.816 (0.744)	Data 0.000 (0.047)	Loss 1.164 (0.563)
Epoch: [47][80/200]	Time 0.691 (0.754)	Data 0.001 (0.058)	Loss 1.029 (0.687)
Epoch: [47][100/200]	Time 2.493 (0.760)	Data 1.783 (0.065)	Loss 1.175 (0.767)
Epoch: [47][120/200]	Time 0.700 (0.752)	Data 0.001 (0.054)	Loss 0.917 (0.808)
Epoch: [47][140/200]	Time 0.702 (0.757)	Data 0.001 (0.058)	Loss 0.876 (0.843)
Epoch: [47][160/200]	Time 0.695 (0.751)	Data 0.000 (0.051)	Loss 0.829 (0.878)
Epoch: [47][180/200]	Time 0.697 (0.755)	Data 0.001 (0.055)	Loss 0.728 (0.893)
Epoch: [47][200/200]	Time 0.701 (0.760)	Data 0.001 (0.058)	Loss 1.221 (0.906)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.232 (0.264)	Data 0.001 (0.027)	
Extract Features: [100/128]	Time 0.231 (0.250)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.17784237861633
==> Statistics for epoch 48: 1086 clusters
Epoch: [48][20/200]	Time 0.689 (0.757)	Data 0.001 (0.058)	Loss 0.213 (0.173)
Epoch: [48][40/200]	Time 0.691 (0.779)	Data 0.001 (0.078)	Loss 0.992 (0.298)
Epoch: [48][60/200]	Time 0.802 (0.754)	Data 0.000 (0.052)	Loss 1.182 (0.549)
Epoch: [48][80/200]	Time 0.694 (0.766)	Data 0.001 (0.064)	Loss 0.936 (0.649)
Epoch: [48][100/200]	Time 2.597 (0.772)	Data 1.838 (0.070)	Loss 0.819 (0.730)
Epoch: [48][120/200]	Time 0.695 (0.760)	Data 0.001 (0.058)	Loss 1.080 (0.781)
Epoch: [48][140/200]	Time 0.693 (0.764)	Data 0.001 (0.063)	Loss 0.732 (0.822)
Epoch: [48][160/200]	Time 0.694 (0.757)	Data 0.000 (0.055)	Loss 1.418 (0.844)
Epoch: [48][180/200]	Time 0.697 (0.761)	Data 0.001 (0.059)	Loss 0.940 (0.870)
Epoch: [48][200/200]	Time 0.695 (0.764)	Data 0.001 (0.062)	Loss 0.763 (0.885)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.236 (0.264)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.229 (0.250)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.518275022506714
==> Statistics for epoch 49: 1087 clusters
Epoch: [49][20/200]	Time 0.691 (0.749)	Data 0.001 (0.049)	Loss 0.116 (0.200)
Epoch: [49][40/200]	Time 0.690 (0.770)	Data 0.001 (0.070)	Loss 0.968 (0.307)
Epoch: [49][60/200]	Time 0.694 (0.746)	Data 0.000 (0.047)	Loss 0.993 (0.557)
Epoch: [49][80/200]	Time 0.693 (0.757)	Data 0.001 (0.058)	Loss 0.694 (0.673)
Epoch: [49][100/200]	Time 2.573 (0.765)	Data 1.845 (0.065)	Loss 0.797 (0.743)
Epoch: [49][120/200]	Time 0.694 (0.754)	Data 0.001 (0.055)	Loss 1.144 (0.801)
Epoch: [49][140/200]	Time 0.705 (0.760)	Data 0.001 (0.060)	Loss 1.084 (0.838)
Epoch: [49][160/200]	Time 0.698 (0.753)	Data 0.000 (0.052)	Loss 0.866 (0.860)
Epoch: [49][180/200]	Time 0.695 (0.756)	Data 0.001 (0.056)	Loss 0.781 (0.878)
Epoch: [49][200/200]	Time 0.694 (0.760)	Data 0.001 (0.059)	Loss 0.985 (0.893)
Extract Features: [50/367]	Time 0.229 (0.265)	Data 0.000 (0.029)	
Extract Features: [100/367]	Time 0.233 (0.251)	Data 0.000 (0.015)	
Extract Features: [150/367]	Time 0.230 (0.247)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.229 (0.245)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.445 (0.244)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.238 (0.242)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.239 (0.241)	Data 0.000 (0.005)	
Mean AP: 70.1%

 * Finished epoch  49  model mAP: 70.1%  best: 70.1% *

==> Test with the best model:
=> Loaded checkpoint 'log/market2msmt/resnet152_cion/model_best.pth.tar'
Extract Features: [50/367]	Time 0.225 (0.259)	Data 0.000 (0.028)	
Extract Features: [100/367]	Time 0.227 (0.246)	Data 0.000 (0.014)	
Extract Features: [150/367]	Time 0.231 (0.243)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.228 (0.240)	Data 0.000 (0.007)	
Extract Features: [250/367]	Time 0.238 (0.240)	Data 0.001 (0.006)	
Extract Features: [300/367]	Time 0.229 (0.239)	Data 0.001 (0.005)	
Extract Features: [350/367]	Time 0.228 (0.239)	Data 0.004 (0.004)	
Mean AP: 70.1%
CMC Scores:
  top-1          88.1%
  top-5          93.6%
  top-10         95.0%
Total running time:  4:02:25.456779
