==========
Args:Namespace(dataset='msmt17', batch_size=256, workers=4, height=256, width=128, num_instances=8, eps=0.7, eps_gap=0.02, k1=30, k2=6, arch='resnet_ibn152a', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet152_IBN_Market1501/64bs_lr0.0004_ep120_warm20_seed0/resnet152_ibn_120.pth', features=0, dropout=0, momentum=0.2, drop_path_rate=0.3, hw_ratio=2, self_norm=True, conv_stem=False, lr=0.00035, weight_decay=0.0005, optimizer='SGD', epochs=50, iters=200, step_size=20, seed=1, print_freq=20, eval_step=10, temp=0.05, data_dir='/home/ma-user/work/Projects/ReIDNet_Finetune/CContrast/data', logs_dir='log/market2msmt/resnet152_ibn_cion', pooling_type='gem', use_hard=True)
==========
==> Load unlabeled dataset
=> MSMT17 loaded
Dataset statistics:
  ----------------------------------------
  subset   | # ids | # images | # cameras
  ----------------------------------------
  train    |  1041 |    32621 |        15
  query    |  3060 |    11659 |        15
  gallery  |  3060 |    82161 |        15
  ----------------------------------------
pooling_type: gem
checkpoint loaded!
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.253 (0.475)	Data 0.000 (0.022)	
Extract Features: [100/128]	Time 0.242 (0.368)	Data 0.000 (0.012)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.477463245391846
Clustering criterion: eps: 0.700
==> Statistics for epoch 0: 783 clusters
Epoch: [0][20/200]	Time 0.733 (1.249)	Data 0.000 (0.051)	Loss 3.098 (3.068)
Epoch: [0][40/200]	Time 0.745 (1.032)	Data 0.001 (0.064)	Loss 2.282 (2.906)
Epoch: [0][60/200]	Time 0.742 (0.961)	Data 0.001 (0.067)	Loss 1.951 (2.686)
Epoch: [0][80/200]	Time 0.741 (0.924)	Data 0.001 (0.069)	Loss 2.109 (2.509)
Epoch: [0][100/200]	Time 0.732 (0.901)	Data 0.001 (0.069)	Loss 1.709 (2.383)
Epoch: [0][120/200]	Time 0.730 (0.877)	Data 0.000 (0.058)	Loss 2.067 (2.295)
Epoch: [0][140/200]	Time 0.728 (0.868)	Data 0.000 (0.060)	Loss 1.653 (2.218)
Epoch: [0][160/200]	Time 0.740 (0.863)	Data 0.001 (0.062)	Loss 1.470 (2.157)
Epoch: [0][180/200]	Time 0.747 (0.859)	Data 0.001 (0.063)	Loss 1.179 (2.105)
Epoch: [0][200/200]	Time 0.734 (0.855)	Data 0.001 (0.065)	Loss 1.637 (2.052)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.281)	Data 0.000 (0.024)	
Extract Features: [100/128]	Time 0.254 (0.268)	Data 0.000 (0.012)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.54187607765198
==> Statistics for epoch 1: 954 clusters
Epoch: [1][20/200]	Time 0.729 (0.803)	Data 0.001 (0.058)	Loss 0.441 (0.434)
Epoch: [1][40/200]	Time 0.735 (0.810)	Data 0.001 (0.067)	Loss 1.913 (0.812)
Epoch: [1][60/200]	Time 0.750 (0.815)	Data 0.001 (0.071)	Loss 1.749 (1.146)
Epoch: [1][80/200]	Time 0.739 (0.799)	Data 0.001 (0.054)	Loss 1.587 (1.292)
Epoch: [1][100/200]	Time 0.744 (0.804)	Data 0.001 (0.058)	Loss 1.500 (1.397)
Epoch: [1][120/200]	Time 0.743 (0.810)	Data 0.001 (0.062)	Loss 1.601 (1.443)
Epoch: [1][140/200]	Time 0.741 (0.801)	Data 0.000 (0.054)	Loss 1.193 (1.475)
Epoch: [1][160/200]	Time 0.751 (0.803)	Data 0.002 (0.056)	Loss 1.870 (1.499)
Epoch: [1][180/200]	Time 0.747 (0.806)	Data 0.001 (0.059)	Loss 1.787 (1.524)
Epoch: [1][200/200]	Time 0.738 (0.800)	Data 0.000 (0.053)	Loss 1.555 (1.536)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.255 (0.282)	Data 0.000 (0.025)	
Extract Features: [100/128]	Time 0.249 (0.269)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.690908432006836
==> Statistics for epoch 2: 990 clusters
Epoch: [2][20/200]	Time 0.734 (0.796)	Data 0.001 (0.053)	Loss 0.331 (0.369)
Epoch: [2][40/200]	Time 0.736 (0.806)	Data 0.001 (0.064)	Loss 1.499 (0.668)
Epoch: [2][60/200]	Time 0.735 (0.783)	Data 0.000 (0.043)	Loss 1.518 (1.000)
Epoch: [2][80/200]	Time 0.739 (0.795)	Data 0.001 (0.053)	Loss 1.395 (1.181)
Epoch: [2][100/200]	Time 0.743 (0.799)	Data 0.001 (0.057)	Loss 1.582 (1.263)
Epoch: [2][120/200]	Time 0.737 (0.791)	Data 0.000 (0.048)	Loss 1.679 (1.325)
Epoch: [2][140/200]	Time 0.733 (0.795)	Data 0.001 (0.053)	Loss 1.576 (1.370)
Epoch: [2][160/200]	Time 0.736 (0.798)	Data 0.001 (0.056)	Loss 1.471 (1.392)
Epoch: [2][180/200]	Time 0.734 (0.792)	Data 0.000 (0.050)	Loss 1.407 (1.413)
Epoch: [2][200/200]	Time 0.738 (0.795)	Data 0.001 (0.052)	Loss 1.657 (1.438)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.286)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.248 (0.270)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.489678382873535
==> Statistics for epoch 3: 995 clusters
Epoch: [3][20/200]	Time 0.736 (0.799)	Data 0.001 (0.056)	Loss 0.370 (0.356)
Epoch: [3][40/200]	Time 0.739 (0.814)	Data 0.001 (0.068)	Loss 1.944 (0.619)
Epoch: [3][60/200]	Time 0.735 (0.791)	Data 0.000 (0.045)	Loss 1.292 (0.979)
Epoch: [3][80/200]	Time 0.738 (0.801)	Data 0.001 (0.056)	Loss 1.870 (1.153)
Epoch: [3][100/200]	Time 0.734 (0.808)	Data 0.001 (0.062)	Loss 1.403 (1.243)
Epoch: [3][120/200]	Time 0.732 (0.797)	Data 0.000 (0.052)	Loss 1.697 (1.311)
Epoch: [3][140/200]	Time 0.734 (0.801)	Data 0.001 (0.056)	Loss 1.282 (1.344)
Epoch: [3][160/200]	Time 0.737 (0.803)	Data 0.001 (0.059)	Loss 1.619 (1.383)
Epoch: [3][180/200]	Time 0.737 (0.797)	Data 0.000 (0.053)	Loss 1.220 (1.400)
Epoch: [3][200/200]	Time 0.738 (0.799)	Data 0.001 (0.056)	Loss 1.726 (1.422)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.242 (0.278)	Data 0.000 (0.025)	
Extract Features: [100/128]	Time 0.255 (0.266)	Data 0.001 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.64858365058899
==> Statistics for epoch 4: 1012 clusters
Epoch: [4][20/200]	Time 0.732 (0.790)	Data 0.001 (0.052)	Loss 0.336 (0.309)
Epoch: [4][40/200]	Time 0.732 (0.806)	Data 0.001 (0.065)	Loss 1.300 (0.567)
Epoch: [4][60/200]	Time 0.750 (0.787)	Data 0.000 (0.044)	Loss 1.234 (0.887)
Epoch: [4][80/200]	Time 0.739 (0.797)	Data 0.001 (0.052)	Loss 1.584 (1.029)
Epoch: [4][100/200]	Time 0.740 (0.804)	Data 0.001 (0.058)	Loss 1.690 (1.117)
Epoch: [4][120/200]	Time 0.744 (0.795)	Data 0.000 (0.049)	Loss 1.212 (1.191)
Epoch: [4][140/200]	Time 0.740 (0.801)	Data 0.001 (0.054)	Loss 1.205 (1.238)
Epoch: [4][160/200]	Time 0.742 (0.805)	Data 0.001 (0.058)	Loss 1.823 (1.278)
Epoch: [4][180/200]	Time 0.733 (0.799)	Data 0.000 (0.051)	Loss 1.842 (1.300)
Epoch: [4][200/200]	Time 0.737 (0.802)	Data 0.001 (0.054)	Loss 1.407 (1.307)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.243 (0.283)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.247 (0.268)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.374061584472656
==> Statistics for epoch 5: 1014 clusters
Epoch: [5][20/200]	Time 0.739 (0.796)	Data 0.001 (0.050)	Loss 0.269 (0.287)
Epoch: [5][40/200]	Time 0.737 (0.805)	Data 0.001 (0.063)	Loss 2.098 (0.536)
Epoch: [5][60/200]	Time 0.740 (0.785)	Data 0.000 (0.042)	Loss 1.531 (0.818)
Epoch: [5][80/200]	Time 0.741 (0.795)	Data 0.001 (0.050)	Loss 1.195 (0.976)
Epoch: [5][100/200]	Time 0.742 (0.802)	Data 0.001 (0.058)	Loss 1.273 (1.065)
Epoch: [5][120/200]	Time 0.733 (0.792)	Data 0.000 (0.049)	Loss 1.295 (1.136)
Epoch: [5][140/200]	Time 0.736 (0.796)	Data 0.001 (0.052)	Loss 1.298 (1.181)
Epoch: [5][160/200]	Time 0.738 (0.799)	Data 0.001 (0.056)	Loss 1.359 (1.213)
Epoch: [5][180/200]	Time 0.733 (0.794)	Data 0.000 (0.050)	Loss 1.115 (1.226)
Epoch: [5][200/200]	Time 0.752 (0.796)	Data 0.001 (0.052)	Loss 1.716 (1.243)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.285)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.247 (0.268)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.20585584640503
==> Statistics for epoch 6: 1012 clusters
Epoch: [6][20/200]	Time 0.744 (0.788)	Data 0.001 (0.053)	Loss 0.188 (0.258)
Epoch: [6][40/200]	Time 0.734 (0.808)	Data 0.001 (0.070)	Loss 1.603 (0.505)
Epoch: [6][60/200]	Time 0.737 (0.785)	Data 0.000 (0.047)	Loss 1.190 (0.795)
Epoch: [6][80/200]	Time 0.736 (0.798)	Data 0.001 (0.056)	Loss 1.584 (0.934)
Epoch: [6][100/200]	Time 0.735 (0.805)	Data 0.001 (0.063)	Loss 1.457 (1.027)
Epoch: [6][120/200]	Time 0.737 (0.796)	Data 0.000 (0.052)	Loss 1.708 (1.098)
Epoch: [6][140/200]	Time 0.737 (0.802)	Data 0.001 (0.057)	Loss 1.443 (1.143)
Epoch: [6][160/200]	Time 0.738 (0.806)	Data 0.001 (0.060)	Loss 1.182 (1.160)
Epoch: [6][180/200]	Time 0.735 (0.799)	Data 0.000 (0.054)	Loss 1.511 (1.185)
Epoch: [6][200/200]	Time 0.737 (0.802)	Data 0.001 (0.057)	Loss 1.224 (1.205)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.249 (0.284)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.246 (0.270)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.67734980583191
==> Statistics for epoch 7: 1017 clusters
Epoch: [7][20/200]	Time 0.740 (0.796)	Data 0.001 (0.051)	Loss 0.236 (0.253)
Epoch: [7][40/200]	Time 0.741 (0.809)	Data 0.001 (0.066)	Loss 1.426 (0.500)
Epoch: [7][60/200]	Time 0.876 (0.790)	Data 0.000 (0.045)	Loss 1.318 (0.784)
Epoch: [7][80/200]	Time 0.740 (0.801)	Data 0.001 (0.055)	Loss 0.867 (0.921)
Epoch: [7][100/200]	Time 0.741 (0.809)	Data 0.002 (0.062)	Loss 1.293 (0.996)
Epoch: [7][120/200]	Time 0.737 (0.799)	Data 0.000 (0.052)	Loss 1.106 (1.047)
Epoch: [7][140/200]	Time 0.745 (0.804)	Data 0.002 (0.056)	Loss 0.922 (1.074)
Epoch: [7][160/200]	Time 0.879 (0.807)	Data 0.001 (0.059)	Loss 1.395 (1.104)
Epoch: [7][180/200]	Time 0.741 (0.801)	Data 0.000 (0.052)	Loss 1.655 (1.130)
Epoch: [7][200/200]	Time 0.743 (0.804)	Data 0.001 (0.055)	Loss 1.010 (1.145)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.284)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.253 (0.270)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.9686324596405
==> Statistics for epoch 8: 1032 clusters
Epoch: [8][20/200]	Time 0.730 (0.797)	Data 0.001 (0.055)	Loss 0.274 (0.223)
Epoch: [8][40/200]	Time 0.730 (0.808)	Data 0.001 (0.069)	Loss 1.107 (0.416)
Epoch: [8][60/200]	Time 0.732 (0.783)	Data 0.000 (0.047)	Loss 1.106 (0.704)
Epoch: [8][80/200]	Time 0.734 (0.794)	Data 0.001 (0.055)	Loss 1.112 (0.858)
Epoch: [8][100/200]	Time 0.741 (0.800)	Data 0.001 (0.059)	Loss 1.553 (0.952)
Epoch: [8][120/200]	Time 0.742 (0.792)	Data 0.001 (0.050)	Loss 1.383 (1.009)
Epoch: [8][140/200]	Time 0.741 (0.796)	Data 0.001 (0.053)	Loss 1.089 (1.041)
Epoch: [8][160/200]	Time 0.730 (0.789)	Data 0.000 (0.046)	Loss 1.072 (1.072)
Epoch: [8][180/200]	Time 0.752 (0.794)	Data 0.001 (0.050)	Loss 1.571 (1.095)
Epoch: [8][200/200]	Time 0.767 (0.798)	Data 0.001 (0.053)	Loss 1.238 (1.108)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.284)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.249 (0.270)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.02211618423462
==> Statistics for epoch 9: 1055 clusters
Epoch: [9][20/200]	Time 0.737 (0.802)	Data 0.001 (0.054)	Loss 0.202 (0.229)
Epoch: [9][40/200]	Time 0.730 (0.812)	Data 0.001 (0.065)	Loss 1.119 (0.423)
Epoch: [9][60/200]	Time 0.730 (0.789)	Data 0.000 (0.044)	Loss 1.027 (0.699)
Epoch: [9][80/200]	Time 0.736 (0.798)	Data 0.001 (0.052)	Loss 1.425 (0.839)
Epoch: [9][100/200]	Time 0.746 (0.803)	Data 0.001 (0.058)	Loss 1.132 (0.922)
Epoch: [9][120/200]	Time 0.750 (0.794)	Data 0.001 (0.048)	Loss 1.469 (0.988)
Epoch: [9][140/200]	Time 0.739 (0.799)	Data 0.001 (0.054)	Loss 1.323 (1.028)
Epoch: [9][160/200]	Time 0.735 (0.793)	Data 0.000 (0.047)	Loss 1.376 (1.060)
Epoch: [9][180/200]	Time 0.734 (0.797)	Data 0.001 (0.051)	Loss 1.172 (1.078)
Epoch: [9][200/200]	Time 0.740 (0.800)	Data 0.001 (0.054)	Loss 1.228 (1.097)
Extract Features: [50/367]	Time 0.245 (0.282)	Data 0.000 (0.028)	
Extract Features: [100/367]	Time 0.254 (0.268)	Data 0.000 (0.014)	
Extract Features: [150/367]	Time 0.308 (0.267)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.251 (0.269)	Data 0.000 (0.007)	
Extract Features: [250/367]	Time 0.248 (0.270)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.244 (0.270)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.245 (0.271)	Data 0.000 (0.005)	
Mean AP: 65.6%

 * Finished epoch   9  model mAP: 65.6%  best: 65.6% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.243 (0.275)	Data 0.000 (0.025)	
Extract Features: [100/128]	Time 0.244 (0.263)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.54852867126465
==> Statistics for epoch 10: 1006 clusters
Epoch: [10][20/200]	Time 0.737 (0.830)	Data 0.001 (0.056)	Loss 0.145 (0.221)
Epoch: [10][40/200]	Time 0.733 (0.825)	Data 0.001 (0.070)	Loss 1.412 (0.428)
Epoch: [10][60/200]	Time 0.732 (0.799)	Data 0.000 (0.047)	Loss 1.224 (0.685)
Epoch: [10][80/200]	Time 0.733 (0.806)	Data 0.001 (0.057)	Loss 1.246 (0.814)
Epoch: [10][100/200]	Time 0.835 (0.810)	Data 0.001 (0.062)	Loss 1.403 (0.878)
Epoch: [10][120/200]	Time 0.737 (0.798)	Data 0.000 (0.052)	Loss 1.305 (0.942)
Epoch: [10][140/200]	Time 0.734 (0.802)	Data 0.001 (0.056)	Loss 0.849 (0.976)
Epoch: [10][160/200]	Time 0.737 (0.806)	Data 0.001 (0.060)	Loss 1.185 (1.010)
Epoch: [10][180/200]	Time 0.732 (0.798)	Data 0.000 (0.053)	Loss 1.019 (1.030)
Epoch: [10][200/200]	Time 0.739 (0.801)	Data 0.001 (0.056)	Loss 1.356 (1.046)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.287)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.249 (0.273)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.862776041030884
==> Statistics for epoch 11: 1016 clusters
Epoch: [11][20/200]	Time 0.731 (0.791)	Data 0.001 (0.056)	Loss 0.167 (0.222)
Epoch: [11][40/200]	Time 0.739 (0.809)	Data 0.001 (0.071)	Loss 1.192 (0.441)
Epoch: [11][60/200]	Time 0.731 (0.789)	Data 0.000 (0.048)	Loss 1.453 (0.685)
Epoch: [11][80/200]	Time 0.748 (0.797)	Data 0.001 (0.056)	Loss 0.974 (0.822)
Epoch: [11][100/200]	Time 0.745 (0.804)	Data 0.001 (0.061)	Loss 1.303 (0.888)
Epoch: [11][120/200]	Time 0.736 (0.793)	Data 0.000 (0.051)	Loss 1.817 (0.951)
Epoch: [11][140/200]	Time 0.737 (0.800)	Data 0.001 (0.057)	Loss 1.071 (0.983)
Epoch: [11][160/200]	Time 0.741 (0.803)	Data 0.001 (0.059)	Loss 1.025 (1.004)
Epoch: [11][180/200]	Time 0.731 (0.797)	Data 0.000 (0.053)	Loss 1.256 (1.028)
Epoch: [11][200/200]	Time 0.743 (0.801)	Data 0.001 (0.056)	Loss 0.948 (1.049)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.290)	Data 0.000 (0.032)	
Extract Features: [100/128]	Time 0.245 (0.274)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.61936068534851
==> Statistics for epoch 12: 1026 clusters
Epoch: [12][20/200]	Time 0.734 (0.800)	Data 0.002 (0.056)	Loss 0.153 (0.212)
Epoch: [12][40/200]	Time 0.738 (0.819)	Data 0.001 (0.074)	Loss 0.828 (0.394)
Epoch: [12][60/200]	Time 0.736 (0.794)	Data 0.000 (0.050)	Loss 1.232 (0.636)
Epoch: [12][80/200]	Time 0.735 (0.807)	Data 0.001 (0.061)	Loss 1.461 (0.779)
Epoch: [12][100/200]	Time 0.749 (0.814)	Data 0.001 (0.068)	Loss 0.910 (0.854)
Epoch: [12][120/200]	Time 0.738 (0.804)	Data 0.001 (0.057)	Loss 1.276 (0.901)
Epoch: [12][140/200]	Time 0.739 (0.809)	Data 0.001 (0.062)	Loss 0.897 (0.944)
Epoch: [12][160/200]	Time 0.739 (0.800)	Data 0.000 (0.054)	Loss 0.939 (0.971)
Epoch: [12][180/200]	Time 0.735 (0.803)	Data 0.001 (0.058)	Loss 1.097 (0.990)
Epoch: [12][200/200]	Time 0.753 (0.807)	Data 0.001 (0.061)	Loss 1.024 (1.005)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.244 (0.286)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.247 (0.271)	Data 0.001 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.442060708999634
==> Statistics for epoch 13: 1039 clusters
Epoch: [13][20/200]	Time 0.734 (0.792)	Data 0.001 (0.057)	Loss 0.250 (0.207)
Epoch: [13][40/200]	Time 0.735 (0.811)	Data 0.001 (0.073)	Loss 0.746 (0.372)
Epoch: [13][60/200]	Time 0.736 (0.786)	Data 0.000 (0.049)	Loss 1.016 (0.631)
Epoch: [13][80/200]	Time 0.736 (0.801)	Data 0.001 (0.061)	Loss 1.262 (0.756)
Epoch: [13][100/200]	Time 0.734 (0.810)	Data 0.001 (0.067)	Loss 1.254 (0.837)
Epoch: [13][120/200]	Time 0.737 (0.799)	Data 0.001 (0.056)	Loss 1.148 (0.907)
Epoch: [13][140/200]	Time 0.745 (0.804)	Data 0.001 (0.061)	Loss 1.340 (0.941)
Epoch: [13][160/200]	Time 0.738 (0.798)	Data 0.000 (0.054)	Loss 1.368 (0.968)
Epoch: [13][180/200]	Time 0.749 (0.804)	Data 0.001 (0.058)	Loss 1.231 (0.994)
Epoch: [13][200/200]	Time 0.736 (0.808)	Data 0.001 (0.062)	Loss 1.042 (1.006)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.248 (0.285)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.363 (0.272)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.65085053443909
==> Statistics for epoch 14: 1016 clusters
Epoch: [14][20/200]	Time 0.736 (0.798)	Data 0.001 (0.059)	Loss 0.171 (0.212)
Epoch: [14][40/200]	Time 0.731 (0.816)	Data 0.001 (0.071)	Loss 1.202 (0.393)
Epoch: [14][60/200]	Time 0.734 (0.789)	Data 0.000 (0.047)	Loss 0.998 (0.640)
Epoch: [14][80/200]	Time 0.739 (0.799)	Data 0.001 (0.056)	Loss 0.935 (0.742)
Epoch: [14][100/200]	Time 0.741 (0.805)	Data 0.001 (0.061)	Loss 1.329 (0.823)
Epoch: [14][120/200]	Time 0.742 (0.796)	Data 0.000 (0.051)	Loss 1.244 (0.879)
Epoch: [14][140/200]	Time 0.753 (0.800)	Data 0.001 (0.056)	Loss 1.003 (0.913)
Epoch: [14][160/200]	Time 0.739 (0.805)	Data 0.001 (0.061)	Loss 1.287 (0.936)
Epoch: [14][180/200]	Time 0.742 (0.798)	Data 0.000 (0.054)	Loss 1.311 (0.954)
Epoch: [14][200/200]	Time 0.733 (0.800)	Data 0.002 (0.057)	Loss 0.898 (0.967)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.283)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.249 (0.269)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.159518003463745
==> Statistics for epoch 15: 1033 clusters
Epoch: [15][20/200]	Time 0.824 (0.789)	Data 0.001 (0.052)	Loss 0.251 (0.208)
Epoch: [15][40/200]	Time 0.732 (0.809)	Data 0.001 (0.069)	Loss 0.955 (0.389)
Epoch: [15][60/200]	Time 0.728 (0.784)	Data 0.000 (0.046)	Loss 1.348 (0.624)
Epoch: [15][80/200]	Time 0.738 (0.799)	Data 0.001 (0.057)	Loss 1.097 (0.738)
Epoch: [15][100/200]	Time 0.739 (0.806)	Data 0.001 (0.064)	Loss 1.141 (0.818)
Epoch: [15][120/200]	Time 0.850 (0.796)	Data 0.001 (0.053)	Loss 0.980 (0.866)
Epoch: [15][140/200]	Time 0.735 (0.800)	Data 0.001 (0.057)	Loss 1.240 (0.904)
Epoch: [15][160/200]	Time 0.732 (0.792)	Data 0.000 (0.050)	Loss 1.309 (0.934)
Epoch: [15][180/200]	Time 0.731 (0.796)	Data 0.001 (0.054)	Loss 1.823 (0.950)
Epoch: [15][200/200]	Time 0.736 (0.799)	Data 0.001 (0.057)	Loss 0.822 (0.970)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.283)	Data 0.001 (0.027)	
Extract Features: [100/128]	Time 0.254 (0.270)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.91519260406494
==> Statistics for epoch 16: 1030 clusters
Epoch: [16][20/200]	Time 0.728 (0.797)	Data 0.001 (0.057)	Loss 0.189 (0.203)
Epoch: [16][40/200]	Time 0.737 (0.808)	Data 0.001 (0.072)	Loss 1.097 (0.366)
Epoch: [16][60/200]	Time 0.738 (0.787)	Data 0.000 (0.049)	Loss 0.757 (0.574)
Epoch: [16][80/200]	Time 0.735 (0.799)	Data 0.001 (0.059)	Loss 1.056 (0.696)
Epoch: [16][100/200]	Time 0.738 (0.806)	Data 0.001 (0.064)	Loss 1.093 (0.765)
Epoch: [16][120/200]	Time 0.870 (0.797)	Data 0.001 (0.053)	Loss 0.837 (0.824)
Epoch: [16][140/200]	Time 0.742 (0.802)	Data 0.001 (0.058)	Loss 1.050 (0.869)
Epoch: [16][160/200]	Time 0.743 (0.795)	Data 0.000 (0.051)	Loss 1.277 (0.895)
Epoch: [16][180/200]	Time 0.737 (0.801)	Data 0.001 (0.055)	Loss 1.072 (0.912)
Epoch: [16][200/200]	Time 0.739 (0.804)	Data 0.001 (0.058)	Loss 1.244 (0.926)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.281)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.246 (0.268)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.7962429523468
==> Statistics for epoch 17: 1041 clusters
Epoch: [17][20/200]	Time 0.735 (0.798)	Data 0.001 (0.057)	Loss 0.164 (0.190)
Epoch: [17][40/200]	Time 0.739 (0.821)	Data 0.001 (0.077)	Loss 1.076 (0.351)
Epoch: [17][60/200]	Time 0.738 (0.796)	Data 0.000 (0.051)	Loss 0.877 (0.589)
Epoch: [17][80/200]	Time 0.739 (0.805)	Data 0.001 (0.059)	Loss 1.017 (0.711)
Epoch: [17][100/200]	Time 0.737 (0.814)	Data 0.001 (0.067)	Loss 1.198 (0.773)
Epoch: [17][120/200]	Time 0.740 (0.802)	Data 0.001 (0.056)	Loss 1.092 (0.825)
Epoch: [17][140/200]	Time 0.745 (0.807)	Data 0.001 (0.060)	Loss 1.261 (0.863)
Epoch: [17][160/200]	Time 0.735 (0.800)	Data 0.000 (0.053)	Loss 1.008 (0.882)
Epoch: [17][180/200]	Time 0.739 (0.804)	Data 0.001 (0.057)	Loss 1.210 (0.896)
Epoch: [17][200/200]	Time 0.735 (0.807)	Data 0.001 (0.061)	Loss 1.036 (0.913)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.244 (0.287)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.247 (0.271)	Data 0.001 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.07065463066101
==> Statistics for epoch 18: 1044 clusters
Epoch: [18][20/200]	Time 0.749 (0.793)	Data 0.002 (0.055)	Loss 0.075 (0.183)
Epoch: [18][40/200]	Time 0.735 (0.814)	Data 0.001 (0.073)	Loss 1.097 (0.348)
Epoch: [18][60/200]	Time 0.735 (0.791)	Data 0.000 (0.049)	Loss 0.929 (0.590)
Epoch: [18][80/200]	Time 0.737 (0.802)	Data 0.001 (0.060)	Loss 0.917 (0.685)
Epoch: [18][100/200]	Time 0.737 (0.810)	Data 0.002 (0.067)	Loss 0.698 (0.747)
Epoch: [18][120/200]	Time 0.740 (0.800)	Data 0.001 (0.056)	Loss 0.800 (0.788)
Epoch: [18][140/200]	Time 0.741 (0.805)	Data 0.001 (0.061)	Loss 0.771 (0.816)
Epoch: [18][160/200]	Time 0.744 (0.798)	Data 0.000 (0.053)	Loss 0.966 (0.850)
Epoch: [18][180/200]	Time 0.832 (0.802)	Data 0.001 (0.057)	Loss 1.053 (0.870)
Epoch: [18][200/200]	Time 0.744 (0.804)	Data 0.001 (0.060)	Loss 1.145 (0.884)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.353 (0.285)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.244 (0.270)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.78887701034546
==> Statistics for epoch 19: 1018 clusters
Epoch: [19][20/200]	Time 0.730 (0.795)	Data 0.001 (0.051)	Loss 0.180 (0.153)
Epoch: [19][40/200]	Time 0.737 (0.812)	Data 0.001 (0.068)	Loss 0.904 (0.334)
Epoch: [19][60/200]	Time 0.732 (0.790)	Data 0.000 (0.046)	Loss 0.934 (0.565)
Epoch: [19][80/200]	Time 0.740 (0.798)	Data 0.001 (0.055)	Loss 1.049 (0.663)
Epoch: [19][100/200]	Time 0.740 (0.805)	Data 0.001 (0.062)	Loss 0.871 (0.729)
Epoch: [19][120/200]	Time 0.734 (0.793)	Data 0.000 (0.051)	Loss 1.221 (0.789)
Epoch: [19][140/200]	Time 0.737 (0.799)	Data 0.001 (0.056)	Loss 1.172 (0.818)
Epoch: [19][160/200]	Time 0.737 (0.803)	Data 0.001 (0.060)	Loss 1.135 (0.839)
Epoch: [19][180/200]	Time 0.734 (0.796)	Data 0.000 (0.053)	Loss 0.782 (0.858)
Epoch: [19][200/200]	Time 0.732 (0.798)	Data 0.001 (0.056)	Loss 0.952 (0.872)
Extract Features: [50/367]	Time 0.251 (0.285)	Data 0.000 (0.031)	
Extract Features: [100/367]	Time 0.245 (0.271)	Data 0.000 (0.016)	
Extract Features: [150/367]	Time 0.248 (0.266)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.244 (0.263)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.247 (0.261)	Data 0.000 (0.007)	
Extract Features: [300/367]	Time 0.246 (0.260)	Data 0.000 (0.006)	
Extract Features: [350/367]	Time 0.246 (0.259)	Data 0.000 (0.005)	
Mean AP: 71.4%

 * Finished epoch  19  model mAP: 71.4%  best: 71.4% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.249 (0.283)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.252 (0.268)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.479060888290405
==> Statistics for epoch 20: 1034 clusters
Epoch: [20][20/200]	Time 0.733 (0.800)	Data 0.001 (0.058)	Loss 0.142 (0.175)
Epoch: [20][40/200]	Time 0.734 (0.814)	Data 0.001 (0.070)	Loss 0.802 (0.313)
Epoch: [20][60/200]	Time 0.727 (0.789)	Data 0.000 (0.047)	Loss 0.913 (0.544)
Epoch: [20][80/200]	Time 0.734 (0.800)	Data 0.001 (0.055)	Loss 1.007 (0.646)
Epoch: [20][100/200]	Time 0.733 (0.806)	Data 0.002 (0.061)	Loss 1.126 (0.711)
Epoch: [20][120/200]	Time 0.732 (0.794)	Data 0.001 (0.051)	Loss 1.129 (0.752)
Epoch: [20][140/200]	Time 0.736 (0.798)	Data 0.001 (0.055)	Loss 0.835 (0.783)
Epoch: [20][160/200]	Time 0.732 (0.791)	Data 0.000 (0.049)	Loss 0.828 (0.804)
Epoch: [20][180/200]	Time 0.736 (0.794)	Data 0.001 (0.052)	Loss 0.788 (0.831)
Epoch: [20][200/200]	Time 0.729 (0.797)	Data 0.001 (0.055)	Loss 1.091 (0.843)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.281)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.252 (0.266)	Data 0.001 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.83699369430542
==> Statistics for epoch 21: 1057 clusters
Epoch: [21][20/200]	Time 0.733 (0.787)	Data 0.001 (0.052)	Loss 0.166 (0.155)
Epoch: [21][40/200]	Time 0.735 (0.814)	Data 0.001 (0.072)	Loss 1.032 (0.293)
Epoch: [21][60/200]	Time 0.850 (0.791)	Data 0.000 (0.048)	Loss 0.925 (0.515)
Epoch: [21][80/200]	Time 0.737 (0.802)	Data 0.001 (0.058)	Loss 0.851 (0.624)
Epoch: [21][100/200]	Time 2.626 (0.810)	Data 1.864 (0.066)	Loss 0.942 (0.691)
Epoch: [21][120/200]	Time 0.740 (0.798)	Data 0.001 (0.055)	Loss 0.994 (0.743)
Epoch: [21][140/200]	Time 0.741 (0.802)	Data 0.001 (0.059)	Loss 1.145 (0.786)
Epoch: [21][160/200]	Time 0.739 (0.795)	Data 0.000 (0.051)	Loss 0.920 (0.808)
Epoch: [21][180/200]	Time 0.751 (0.800)	Data 0.001 (0.056)	Loss 0.931 (0.828)
Epoch: [21][200/200]	Time 0.744 (0.803)	Data 0.001 (0.059)	Loss 0.819 (0.833)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.280)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.248 (0.267)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.8584885597229
==> Statistics for epoch 22: 1054 clusters
Epoch: [22][20/200]	Time 0.737 (0.790)	Data 0.001 (0.054)	Loss 0.140 (0.157)
Epoch: [22][40/200]	Time 0.731 (0.812)	Data 0.001 (0.072)	Loss 0.937 (0.298)
Epoch: [22][60/200]	Time 0.731 (0.785)	Data 0.000 (0.048)	Loss 0.767 (0.500)
Epoch: [22][80/200]	Time 0.735 (0.798)	Data 0.001 (0.057)	Loss 0.896 (0.624)
Epoch: [22][100/200]	Time 0.733 (0.806)	Data 0.001 (0.063)	Loss 1.332 (0.699)
Epoch: [22][120/200]	Time 0.742 (0.796)	Data 0.001 (0.053)	Loss 0.926 (0.748)
Epoch: [22][140/200]	Time 0.734 (0.800)	Data 0.001 (0.057)	Loss 0.757 (0.772)
Epoch: [22][160/200]	Time 0.734 (0.793)	Data 0.000 (0.050)	Loss 1.194 (0.794)
Epoch: [22][180/200]	Time 0.739 (0.796)	Data 0.001 (0.054)	Loss 0.610 (0.811)
Epoch: [22][200/200]	Time 0.735 (0.800)	Data 0.001 (0.057)	Loss 0.775 (0.821)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.280)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.253 (0.268)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.401376247406006
==> Statistics for epoch 23: 1039 clusters
Epoch: [23][20/200]	Time 0.739 (0.797)	Data 0.001 (0.057)	Loss 0.183 (0.153)
Epoch: [23][40/200]	Time 0.733 (0.809)	Data 0.001 (0.069)	Loss 1.157 (0.302)
Epoch: [23][60/200]	Time 0.729 (0.786)	Data 0.000 (0.046)	Loss 0.780 (0.508)
Epoch: [23][80/200]	Time 0.742 (0.796)	Data 0.002 (0.055)	Loss 0.962 (0.625)
Epoch: [23][100/200]	Time 0.742 (0.804)	Data 0.001 (0.061)	Loss 1.095 (0.671)
Epoch: [23][120/200]	Time 0.741 (0.795)	Data 0.001 (0.051)	Loss 0.957 (0.713)
Epoch: [23][140/200]	Time 0.746 (0.800)	Data 0.001 (0.056)	Loss 0.848 (0.751)
Epoch: [23][160/200]	Time 0.733 (0.792)	Data 0.000 (0.049)	Loss 0.989 (0.776)
Epoch: [23][180/200]	Time 0.742 (0.797)	Data 0.001 (0.053)	Loss 0.920 (0.802)
Epoch: [23][200/200]	Time 0.738 (0.801)	Data 0.001 (0.057)	Loss 1.312 (0.818)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.282)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.249 (0.271)	Data 0.001 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.349873065948486
==> Statistics for epoch 24: 1055 clusters
Epoch: [24][20/200]	Time 0.733 (0.798)	Data 0.001 (0.051)	Loss 0.158 (0.162)
Epoch: [24][40/200]	Time 0.750 (0.815)	Data 0.001 (0.070)	Loss 1.308 (0.312)
Epoch: [24][60/200]	Time 0.738 (0.789)	Data 0.000 (0.047)	Loss 0.829 (0.514)
Epoch: [24][80/200]	Time 0.740 (0.802)	Data 0.001 (0.058)	Loss 0.924 (0.606)
Epoch: [24][100/200]	Time 0.740 (0.810)	Data 0.002 (0.066)	Loss 1.048 (0.675)
Epoch: [24][120/200]	Time 0.734 (0.800)	Data 0.001 (0.055)	Loss 1.185 (0.724)
Epoch: [24][140/200]	Time 0.737 (0.806)	Data 0.001 (0.060)	Loss 1.012 (0.756)
Epoch: [24][160/200]	Time 0.875 (0.798)	Data 0.000 (0.053)	Loss 0.908 (0.783)
Epoch: [24][180/200]	Time 0.737 (0.802)	Data 0.001 (0.057)	Loss 0.813 (0.798)
Epoch: [24][200/200]	Time 0.738 (0.806)	Data 0.001 (0.061)	Loss 0.795 (0.810)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.249 (0.285)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.247 (0.271)	Data 0.001 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.64721465110779
==> Statistics for epoch 25: 1050 clusters
Epoch: [25][20/200]	Time 0.731 (0.805)	Data 0.001 (0.054)	Loss 0.140 (0.153)
Epoch: [25][40/200]	Time 0.736 (0.817)	Data 0.001 (0.072)	Loss 1.161 (0.301)
Epoch: [25][60/200]	Time 0.748 (0.795)	Data 0.000 (0.049)	Loss 0.721 (0.494)
Epoch: [25][80/200]	Time 0.747 (0.802)	Data 0.001 (0.057)	Loss 1.342 (0.622)
Epoch: [25][100/200]	Time 0.735 (0.808)	Data 0.001 (0.063)	Loss 1.021 (0.683)
Epoch: [25][120/200]	Time 0.738 (0.798)	Data 0.001 (0.052)	Loss 1.066 (0.736)
Epoch: [25][140/200]	Time 0.737 (0.805)	Data 0.001 (0.058)	Loss 1.095 (0.765)
Epoch: [25][160/200]	Time 0.744 (0.798)	Data 0.000 (0.051)	Loss 0.693 (0.783)
Epoch: [25][180/200]	Time 0.743 (0.803)	Data 0.001 (0.056)	Loss 0.802 (0.797)
Epoch: [25][200/200]	Time 0.745 (0.806)	Data 0.001 (0.059)	Loss 0.778 (0.810)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.287)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.260 (0.272)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.728174924850464
==> Statistics for epoch 26: 1048 clusters
Epoch: [26][20/200]	Time 0.730 (0.792)	Data 0.002 (0.054)	Loss 0.142 (0.161)
Epoch: [26][40/200]	Time 0.733 (0.812)	Data 0.001 (0.071)	Loss 0.942 (0.315)
Epoch: [26][60/200]	Time 0.739 (0.788)	Data 0.000 (0.048)	Loss 1.100 (0.503)
Epoch: [26][80/200]	Time 0.739 (0.799)	Data 0.001 (0.058)	Loss 0.910 (0.619)
Epoch: [26][100/200]	Time 0.734 (0.807)	Data 0.001 (0.066)	Loss 0.722 (0.672)
Epoch: [26][120/200]	Time 0.733 (0.796)	Data 0.001 (0.055)	Loss 0.698 (0.712)
Epoch: [26][140/200]	Time 0.733 (0.800)	Data 0.001 (0.060)	Loss 0.771 (0.749)
Epoch: [26][160/200]	Time 0.733 (0.793)	Data 0.000 (0.052)	Loss 0.846 (0.773)
Epoch: [26][180/200]	Time 0.737 (0.798)	Data 0.001 (0.057)	Loss 0.895 (0.792)
Epoch: [26][200/200]	Time 0.734 (0.801)	Data 0.001 (0.059)	Loss 0.922 (0.811)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.281)	Data 0.001 (0.026)	
Extract Features: [100/128]	Time 0.248 (0.268)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.32433199882507
==> Statistics for epoch 27: 1042 clusters
Epoch: [27][20/200]	Time 0.730 (0.784)	Data 0.001 (0.054)	Loss 0.127 (0.159)
Epoch: [27][40/200]	Time 0.731 (0.805)	Data 0.001 (0.069)	Loss 0.580 (0.293)
Epoch: [27][60/200]	Time 0.833 (0.785)	Data 0.000 (0.046)	Loss 0.787 (0.521)
Epoch: [27][80/200]	Time 0.741 (0.796)	Data 0.001 (0.057)	Loss 0.905 (0.632)
Epoch: [27][100/200]	Time 0.740 (0.804)	Data 0.003 (0.062)	Loss 0.853 (0.680)
Epoch: [27][120/200]	Time 0.733 (0.795)	Data 0.001 (0.052)	Loss 1.039 (0.731)
Epoch: [27][140/200]	Time 0.740 (0.800)	Data 0.001 (0.057)	Loss 1.199 (0.760)
Epoch: [27][160/200]	Time 0.738 (0.792)	Data 0.000 (0.050)	Loss 0.881 (0.784)
Epoch: [27][180/200]	Time 0.737 (0.798)	Data 0.001 (0.054)	Loss 0.827 (0.798)
Epoch: [27][200/200]	Time 0.733 (0.802)	Data 0.001 (0.057)	Loss 0.941 (0.815)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.283)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.247 (0.271)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.8758430480957
==> Statistics for epoch 28: 1049 clusters
Epoch: [28][20/200]	Time 0.724 (0.790)	Data 0.001 (0.055)	Loss 0.149 (0.160)
Epoch: [28][40/200]	Time 0.729 (0.805)	Data 0.001 (0.068)	Loss 0.747 (0.300)
Epoch: [28][60/200]	Time 0.735 (0.780)	Data 0.000 (0.045)	Loss 0.821 (0.491)
Epoch: [28][80/200]	Time 0.732 (0.792)	Data 0.001 (0.055)	Loss 0.962 (0.608)
Epoch: [28][100/200]	Time 0.733 (0.799)	Data 0.001 (0.061)	Loss 0.823 (0.655)
Epoch: [28][120/200]	Time 0.738 (0.791)	Data 0.001 (0.051)	Loss 0.720 (0.714)
Epoch: [28][140/200]	Time 0.747 (0.798)	Data 0.001 (0.057)	Loss 0.892 (0.745)
Epoch: [28][160/200]	Time 0.735 (0.791)	Data 0.000 (0.050)	Loss 0.786 (0.767)
Epoch: [28][180/200]	Time 0.739 (0.796)	Data 0.001 (0.054)	Loss 0.771 (0.786)
Epoch: [28][200/200]	Time 0.743 (0.800)	Data 0.001 (0.058)	Loss 0.895 (0.802)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.252 (0.279)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.261 (0.267)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.87853240966797
==> Statistics for epoch 29: 1046 clusters
Epoch: [29][20/200]	Time 0.733 (0.797)	Data 0.001 (0.057)	Loss 0.087 (0.164)
Epoch: [29][40/200]	Time 0.735 (0.816)	Data 0.001 (0.073)	Loss 1.033 (0.324)
Epoch: [29][60/200]	Time 0.737 (0.792)	Data 0.000 (0.049)	Loss 0.931 (0.486)
Epoch: [29][80/200]	Time 0.738 (0.804)	Data 0.001 (0.061)	Loss 0.791 (0.585)
Epoch: [29][100/200]	Time 0.739 (0.810)	Data 0.001 (0.066)	Loss 1.043 (0.673)
Epoch: [29][120/200]	Time 0.744 (0.800)	Data 0.001 (0.056)	Loss 0.952 (0.724)
Epoch: [29][140/200]	Time 0.738 (0.803)	Data 0.001 (0.059)	Loss 0.718 (0.746)
Epoch: [29][160/200]	Time 0.738 (0.795)	Data 0.000 (0.052)	Loss 1.014 (0.767)
Epoch: [29][180/200]	Time 0.855 (0.801)	Data 0.001 (0.057)	Loss 0.819 (0.784)
Epoch: [29][200/200]	Time 0.734 (0.805)	Data 0.001 (0.061)	Loss 0.810 (0.807)
Extract Features: [50/367]	Time 0.247 (0.291)	Data 0.000 (0.032)	
Extract Features: [100/367]	Time 0.256 (0.274)	Data 0.000 (0.016)	
Extract Features: [150/367]	Time 0.255 (0.270)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.251 (0.267)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.259 (0.265)	Data 0.001 (0.007)	
Extract Features: [300/367]	Time 0.246 (0.265)	Data 0.000 (0.006)	
Extract Features: [350/367]	Time 0.248 (0.264)	Data 0.000 (0.005)	
Mean AP: 73.4%

 * Finished epoch  29  model mAP: 73.4%  best: 73.4% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.363 (0.284)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.245 (0.268)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.11540985107422
==> Statistics for epoch 30: 1056 clusters
Epoch: [30][20/200]	Time 0.729 (0.793)	Data 0.001 (0.053)	Loss 0.189 (0.150)
Epoch: [30][40/200]	Time 0.734 (0.811)	Data 0.001 (0.071)	Loss 1.095 (0.305)
Epoch: [30][60/200]	Time 0.732 (0.790)	Data 0.000 (0.047)	Loss 1.075 (0.511)
Epoch: [30][80/200]	Time 0.741 (0.802)	Data 0.001 (0.059)	Loss 0.873 (0.621)
Epoch: [30][100/200]	Time 2.553 (0.808)	Data 1.740 (0.065)	Loss 0.957 (0.697)
Epoch: [30][120/200]	Time 0.735 (0.797)	Data 0.001 (0.054)	Loss 0.746 (0.743)
Epoch: [30][140/200]	Time 0.740 (0.804)	Data 0.001 (0.060)	Loss 0.734 (0.762)
Epoch: [30][160/200]	Time 0.737 (0.797)	Data 0.000 (0.052)	Loss 1.070 (0.783)
Epoch: [30][180/200]	Time 0.847 (0.802)	Data 0.001 (0.056)	Loss 0.864 (0.793)
Epoch: [30][200/200]	Time 0.744 (0.805)	Data 0.001 (0.060)	Loss 0.768 (0.805)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.249 (0.287)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.246 (0.271)	Data 0.001 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.866365909576416
==> Statistics for epoch 31: 1047 clusters
Epoch: [31][20/200]	Time 0.737 (0.796)	Data 0.001 (0.055)	Loss 0.123 (0.140)
Epoch: [31][40/200]	Time 0.737 (0.815)	Data 0.001 (0.073)	Loss 1.004 (0.302)
Epoch: [31][60/200]	Time 0.732 (0.791)	Data 0.000 (0.049)	Loss 0.635 (0.486)
Epoch: [31][80/200]	Time 0.731 (0.801)	Data 0.001 (0.059)	Loss 0.715 (0.596)
Epoch: [31][100/200]	Time 0.867 (0.806)	Data 0.001 (0.065)	Loss 1.342 (0.650)
Epoch: [31][120/200]	Time 0.731 (0.795)	Data 0.001 (0.054)	Loss 0.950 (0.699)
Epoch: [31][140/200]	Time 0.734 (0.801)	Data 0.001 (0.060)	Loss 1.226 (0.727)
Epoch: [31][160/200]	Time 0.732 (0.795)	Data 0.000 (0.052)	Loss 0.886 (0.755)
Epoch: [31][180/200]	Time 0.735 (0.799)	Data 0.001 (0.056)	Loss 0.845 (0.775)
Epoch: [31][200/200]	Time 0.885 (0.802)	Data 0.001 (0.059)	Loss 0.959 (0.790)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.287)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.248 (0.273)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.278512477874756
==> Statistics for epoch 32: 1044 clusters
Epoch: [32][20/200]	Time 0.727 (0.784)	Data 0.001 (0.052)	Loss 0.148 (0.158)
Epoch: [32][40/200]	Time 0.743 (0.811)	Data 0.001 (0.070)	Loss 1.199 (0.309)
Epoch: [32][60/200]	Time 0.733 (0.786)	Data 0.000 (0.047)	Loss 0.857 (0.508)
Epoch: [32][80/200]	Time 0.739 (0.797)	Data 0.001 (0.057)	Loss 0.880 (0.597)
Epoch: [32][100/200]	Time 0.738 (0.804)	Data 0.002 (0.063)	Loss 1.046 (0.649)
Epoch: [32][120/200]	Time 0.735 (0.792)	Data 0.001 (0.053)	Loss 0.949 (0.687)
Epoch: [32][140/200]	Time 0.733 (0.799)	Data 0.001 (0.058)	Loss 0.878 (0.720)
Epoch: [32][160/200]	Time 0.732 (0.791)	Data 0.000 (0.051)	Loss 1.048 (0.747)
Epoch: [32][180/200]	Time 0.734 (0.794)	Data 0.001 (0.054)	Loss 1.023 (0.769)
Epoch: [32][200/200]	Time 0.732 (0.797)	Data 0.001 (0.057)	Loss 0.692 (0.784)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.250 (0.286)	Data 0.001 (0.028)	
Extract Features: [100/128]	Time 0.247 (0.270)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.31527400016785
==> Statistics for epoch 33: 1055 clusters
Epoch: [33][20/200]	Time 0.728 (0.792)	Data 0.001 (0.052)	Loss 0.191 (0.170)
Epoch: [33][40/200]	Time 0.733 (0.810)	Data 0.001 (0.070)	Loss 0.716 (0.293)
Epoch: [33][60/200]	Time 0.731 (0.786)	Data 0.000 (0.047)	Loss 0.954 (0.517)
Epoch: [33][80/200]	Time 0.738 (0.796)	Data 0.001 (0.056)	Loss 0.647 (0.624)
Epoch: [33][100/200]	Time 0.758 (0.805)	Data 0.001 (0.062)	Loss 0.708 (0.670)
Epoch: [33][120/200]	Time 0.743 (0.796)	Data 0.001 (0.052)	Loss 1.224 (0.716)
Epoch: [33][140/200]	Time 0.736 (0.803)	Data 0.001 (0.058)	Loss 1.016 (0.748)
Epoch: [33][160/200]	Time 0.732 (0.795)	Data 0.000 (0.051)	Loss 0.869 (0.764)
Epoch: [33][180/200]	Time 0.735 (0.800)	Data 0.001 (0.055)	Loss 0.715 (0.781)
Epoch: [33][200/200]	Time 0.735 (0.802)	Data 0.001 (0.058)	Loss 1.116 (0.792)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.250 (0.286)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.251 (0.272)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.383238315582275
==> Statistics for epoch 34: 1054 clusters
Epoch: [34][20/200]	Time 0.725 (0.795)	Data 0.001 (0.050)	Loss 0.117 (0.168)
Epoch: [34][40/200]	Time 0.749 (0.811)	Data 0.001 (0.068)	Loss 0.880 (0.308)
Epoch: [34][60/200]	Time 0.739 (0.789)	Data 0.000 (0.046)	Loss 0.813 (0.498)
Epoch: [34][80/200]	Time 0.736 (0.802)	Data 0.001 (0.058)	Loss 0.613 (0.592)
Epoch: [34][100/200]	Time 0.739 (0.808)	Data 0.001 (0.064)	Loss 1.019 (0.669)
Epoch: [34][120/200]	Time 0.735 (0.798)	Data 0.001 (0.053)	Loss 0.882 (0.707)
Epoch: [34][140/200]	Time 0.843 (0.802)	Data 0.001 (0.058)	Loss 0.496 (0.736)
Epoch: [34][160/200]	Time 0.735 (0.794)	Data 0.000 (0.051)	Loss 0.741 (0.762)
Epoch: [34][180/200]	Time 0.735 (0.797)	Data 0.001 (0.054)	Loss 0.994 (0.779)
Epoch: [34][200/200]	Time 0.732 (0.801)	Data 0.001 (0.058)	Loss 1.132 (0.789)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.245 (0.282)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.249 (0.268)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 58.87746286392212
==> Statistics for epoch 35: 1056 clusters
Epoch: [35][20/200]	Time 0.870 (0.799)	Data 0.001 (0.056)	Loss 0.075 (0.151)
Epoch: [35][40/200]	Time 0.736 (0.813)	Data 0.001 (0.072)	Loss 1.190 (0.282)
Epoch: [35][60/200]	Time 0.738 (0.788)	Data 0.000 (0.048)	Loss 0.995 (0.478)
Epoch: [35][80/200]	Time 0.736 (0.799)	Data 0.001 (0.059)	Loss 0.889 (0.589)
Epoch: [35][100/200]	Time 2.510 (0.806)	Data 1.724 (0.064)	Loss 0.565 (0.650)
Epoch: [35][120/200]	Time 0.745 (0.794)	Data 0.001 (0.054)	Loss 0.912 (0.695)
Epoch: [35][140/200]	Time 0.729 (0.799)	Data 0.001 (0.059)	Loss 0.821 (0.735)
Epoch: [35][160/200]	Time 0.727 (0.791)	Data 0.000 (0.052)	Loss 1.212 (0.753)
Epoch: [35][180/200]	Time 0.733 (0.795)	Data 0.001 (0.055)	Loss 0.778 (0.771)
Epoch: [35][200/200]	Time 0.741 (0.799)	Data 0.001 (0.058)	Loss 0.678 (0.783)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.242 (0.285)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.248 (0.272)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.48000693321228
==> Statistics for epoch 36: 1063 clusters
Epoch: [36][20/200]	Time 0.734 (0.806)	Data 0.001 (0.056)	Loss 0.154 (0.148)
Epoch: [36][40/200]	Time 0.734 (0.824)	Data 0.001 (0.075)	Loss 0.942 (0.270)
Epoch: [36][60/200]	Time 0.740 (0.797)	Data 0.000 (0.051)	Loss 1.046 (0.467)
Epoch: [36][80/200]	Time 0.738 (0.806)	Data 0.001 (0.062)	Loss 0.824 (0.578)
Epoch: [36][100/200]	Time 2.665 (0.814)	Data 1.906 (0.069)	Loss 1.136 (0.648)
Epoch: [36][120/200]	Time 0.735 (0.803)	Data 0.001 (0.058)	Loss 0.976 (0.694)
Epoch: [36][140/200]	Time 0.734 (0.805)	Data 0.002 (0.061)	Loss 1.059 (0.721)
Epoch: [36][160/200]	Time 0.731 (0.797)	Data 0.000 (0.054)	Loss 1.020 (0.743)
Epoch: [36][180/200]	Time 0.735 (0.801)	Data 0.001 (0.057)	Loss 0.894 (0.765)
Epoch: [36][200/200]	Time 0.742 (0.802)	Data 0.001 (0.060)	Loss 0.813 (0.781)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.284)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.246 (0.272)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.33310413360596
==> Statistics for epoch 37: 1052 clusters
Epoch: [37][20/200]	Time 0.840 (0.806)	Data 0.001 (0.059)	Loss 0.092 (0.140)
Epoch: [37][40/200]	Time 0.734 (0.821)	Data 0.001 (0.076)	Loss 0.740 (0.279)
Epoch: [37][60/200]	Time 0.734 (0.797)	Data 0.000 (0.051)	Loss 1.186 (0.483)
Epoch: [37][80/200]	Time 0.740 (0.808)	Data 0.001 (0.061)	Loss 0.847 (0.588)
Epoch: [37][100/200]	Time 0.739 (0.816)	Data 0.002 (0.068)	Loss 0.796 (0.647)
Epoch: [37][120/200]	Time 0.738 (0.803)	Data 0.001 (0.057)	Loss 0.956 (0.694)
Epoch: [37][140/200]	Time 0.744 (0.808)	Data 0.001 (0.061)	Loss 1.126 (0.719)
Epoch: [37][160/200]	Time 0.739 (0.800)	Data 0.000 (0.054)	Loss 1.085 (0.745)
Epoch: [37][180/200]	Time 0.740 (0.805)	Data 0.001 (0.058)	Loss 0.949 (0.773)
Epoch: [37][200/200]	Time 0.739 (0.809)	Data 0.001 (0.061)	Loss 1.019 (0.784)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.285)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.249 (0.270)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.43344569206238
==> Statistics for epoch 38: 1049 clusters
Epoch: [38][20/200]	Time 0.732 (0.805)	Data 0.001 (0.057)	Loss 0.240 (0.156)
Epoch: [38][40/200]	Time 0.733 (0.817)	Data 0.001 (0.074)	Loss 1.074 (0.288)
Epoch: [38][60/200]	Time 0.731 (0.793)	Data 0.000 (0.050)	Loss 1.080 (0.511)
Epoch: [38][80/200]	Time 0.741 (0.803)	Data 0.001 (0.060)	Loss 1.077 (0.605)
Epoch: [38][100/200]	Time 0.737 (0.808)	Data 0.001 (0.065)	Loss 0.678 (0.658)
Epoch: [38][120/200]	Time 0.741 (0.800)	Data 0.001 (0.055)	Loss 0.842 (0.697)
Epoch: [38][140/200]	Time 0.749 (0.804)	Data 0.002 (0.059)	Loss 0.616 (0.723)
Epoch: [38][160/200]	Time 0.737 (0.797)	Data 0.000 (0.052)	Loss 0.970 (0.750)
Epoch: [38][180/200]	Time 0.741 (0.802)	Data 0.001 (0.056)	Loss 0.958 (0.772)
Epoch: [38][200/200]	Time 0.734 (0.805)	Data 0.001 (0.059)	Loss 0.584 (0.776)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.248 (0.286)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.254 (0.272)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.55259990692139
==> Statistics for epoch 39: 1048 clusters
Epoch: [39][20/200]	Time 0.733 (0.795)	Data 0.001 (0.051)	Loss 0.091 (0.159)
Epoch: [39][40/200]	Time 0.739 (0.811)	Data 0.001 (0.071)	Loss 0.810 (0.303)
Epoch: [39][60/200]	Time 0.734 (0.790)	Data 0.000 (0.048)	Loss 0.872 (0.489)
Epoch: [39][80/200]	Time 0.736 (0.799)	Data 0.001 (0.059)	Loss 0.917 (0.587)
Epoch: [39][100/200]	Time 0.743 (0.805)	Data 0.003 (0.065)	Loss 0.908 (0.642)
Epoch: [39][120/200]	Time 0.737 (0.794)	Data 0.001 (0.054)	Loss 0.794 (0.692)
Epoch: [39][140/200]	Time 0.735 (0.801)	Data 0.001 (0.059)	Loss 1.019 (0.722)
Epoch: [39][160/200]	Time 0.735 (0.794)	Data 0.000 (0.052)	Loss 0.870 (0.742)
Epoch: [39][180/200]	Time 0.739 (0.799)	Data 0.001 (0.056)	Loss 1.050 (0.761)
Epoch: [39][200/200]	Time 0.738 (0.803)	Data 0.001 (0.059)	Loss 1.039 (0.776)
Extract Features: [50/367]	Time 0.250 (0.291)	Data 0.000 (0.032)	
Extract Features: [100/367]	Time 0.248 (0.274)	Data 0.000 (0.016)	
Extract Features: [150/367]	Time 0.252 (0.268)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.247 (0.266)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.248 (0.264)	Data 0.000 (0.007)	
Extract Features: [300/367]	Time 0.248 (0.262)	Data 0.000 (0.006)	
Extract Features: [350/367]	Time 0.247 (0.261)	Data 0.000 (0.005)	
Mean AP: 73.7%

 * Finished epoch  39  model mAP: 73.7%  best: 73.7% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.244 (0.280)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.247 (0.267)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.00529408454895
==> Statistics for epoch 40: 1052 clusters
Epoch: [40][20/200]	Time 0.735 (0.788)	Data 0.002 (0.053)	Loss 0.110 (0.143)
Epoch: [40][40/200]	Time 0.736 (0.813)	Data 0.001 (0.074)	Loss 0.870 (0.280)
Epoch: [40][60/200]	Time 0.735 (0.788)	Data 0.000 (0.050)	Loss 0.884 (0.470)
Epoch: [40][80/200]	Time 0.736 (0.801)	Data 0.002 (0.062)	Loss 0.747 (0.560)
Epoch: [40][100/200]	Time 0.741 (0.809)	Data 0.002 (0.068)	Loss 0.824 (0.634)
Epoch: [40][120/200]	Time 0.736 (0.797)	Data 0.001 (0.057)	Loss 1.057 (0.681)
Epoch: [40][140/200]	Time 0.736 (0.802)	Data 0.001 (0.061)	Loss 0.851 (0.707)
Epoch: [40][160/200]	Time 0.863 (0.795)	Data 0.000 (0.054)	Loss 1.381 (0.736)
Epoch: [40][180/200]	Time 0.741 (0.799)	Data 0.001 (0.058)	Loss 0.793 (0.749)
Epoch: [40][200/200]	Time 0.737 (0.803)	Data 0.001 (0.061)	Loss 0.736 (0.763)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.284)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.250 (0.272)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.47942090034485
==> Statistics for epoch 41: 1053 clusters
Epoch: [41][20/200]	Time 0.734 (0.806)	Data 0.001 (0.059)	Loss 0.184 (0.154)
Epoch: [41][40/200]	Time 0.738 (0.823)	Data 0.001 (0.074)	Loss 0.760 (0.300)
Epoch: [41][60/200]	Time 0.740 (0.797)	Data 0.000 (0.050)	Loss 0.836 (0.495)
Epoch: [41][80/200]	Time 0.734 (0.809)	Data 0.001 (0.061)	Loss 0.678 (0.601)
Epoch: [41][100/200]	Time 0.738 (0.815)	Data 0.001 (0.067)	Loss 0.857 (0.658)
Epoch: [41][120/200]	Time 0.741 (0.802)	Data 0.001 (0.056)	Loss 0.741 (0.707)
Epoch: [41][140/200]	Time 0.738 (0.808)	Data 0.001 (0.061)	Loss 0.772 (0.728)
Epoch: [41][160/200]	Time 0.847 (0.801)	Data 0.000 (0.054)	Loss 0.953 (0.756)
Epoch: [41][180/200]	Time 0.740 (0.805)	Data 0.001 (0.058)	Loss 0.977 (0.769)
Epoch: [41][200/200]	Time 0.736 (0.810)	Data 0.001 (0.062)	Loss 0.871 (0.777)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.250 (0.284)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.251 (0.269)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.5210223197937
==> Statistics for epoch 42: 1055 clusters
Epoch: [42][20/200]	Time 0.730 (0.806)	Data 0.001 (0.060)	Loss 0.173 (0.153)
Epoch: [42][40/200]	Time 0.740 (0.813)	Data 0.001 (0.074)	Loss 0.921 (0.292)
Epoch: [42][60/200]	Time 0.740 (0.789)	Data 0.000 (0.049)	Loss 0.862 (0.476)
Epoch: [42][80/200]	Time 0.738 (0.800)	Data 0.001 (0.058)	Loss 0.770 (0.591)
Epoch: [42][100/200]	Time 0.736 (0.804)	Data 0.001 (0.064)	Loss 1.070 (0.651)
Epoch: [42][120/200]	Time 0.741 (0.794)	Data 0.001 (0.053)	Loss 1.107 (0.702)
Epoch: [42][140/200]	Time 0.736 (0.801)	Data 0.001 (0.059)	Loss 0.741 (0.729)
Epoch: [42][160/200]	Time 0.733 (0.793)	Data 0.000 (0.052)	Loss 1.092 (0.752)
Epoch: [42][180/200]	Time 0.733 (0.796)	Data 0.001 (0.055)	Loss 0.645 (0.768)
Epoch: [42][200/200]	Time 0.733 (0.799)	Data 0.001 (0.059)	Loss 0.960 (0.776)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.248 (0.282)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.247 (0.268)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.23799180984497
==> Statistics for epoch 43: 1048 clusters
Epoch: [43][20/200]	Time 0.732 (0.796)	Data 0.001 (0.054)	Loss 0.127 (0.143)
Epoch: [43][40/200]	Time 0.734 (0.814)	Data 0.001 (0.072)	Loss 1.176 (0.283)
Epoch: [43][60/200]	Time 0.738 (0.790)	Data 0.000 (0.048)	Loss 0.957 (0.484)
Epoch: [43][80/200]	Time 0.737 (0.801)	Data 0.001 (0.060)	Loss 0.882 (0.588)
Epoch: [43][100/200]	Time 0.738 (0.807)	Data 0.001 (0.065)	Loss 0.974 (0.642)
Epoch: [43][120/200]	Time 0.742 (0.796)	Data 0.001 (0.054)	Loss 0.925 (0.689)
Epoch: [43][140/200]	Time 0.740 (0.802)	Data 0.001 (0.060)	Loss 0.815 (0.713)
Epoch: [43][160/200]	Time 0.739 (0.795)	Data 0.000 (0.052)	Loss 1.006 (0.732)
Epoch: [43][180/200]	Time 0.746 (0.799)	Data 0.001 (0.057)	Loss 0.830 (0.748)
Epoch: [43][200/200]	Time 0.739 (0.802)	Data 0.001 (0.060)	Loss 0.721 (0.761)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.247 (0.285)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.250 (0.271)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.38425016403198
==> Statistics for epoch 44: 1052 clusters
Epoch: [44][20/200]	Time 0.726 (0.798)	Data 0.001 (0.055)	Loss 0.147 (0.152)
Epoch: [44][40/200]	Time 0.864 (0.813)	Data 0.001 (0.071)	Loss 0.555 (0.301)
Epoch: [44][60/200]	Time 0.735 (0.790)	Data 0.000 (0.048)	Loss 0.786 (0.498)
Epoch: [44][80/200]	Time 0.740 (0.801)	Data 0.001 (0.059)	Loss 0.715 (0.599)
Epoch: [44][100/200]	Time 0.746 (0.809)	Data 0.001 (0.067)	Loss 0.910 (0.653)
Epoch: [44][120/200]	Time 0.729 (0.798)	Data 0.001 (0.056)	Loss 0.981 (0.689)
Epoch: [44][140/200]	Time 0.728 (0.802)	Data 0.001 (0.059)	Loss 0.806 (0.717)
Epoch: [44][160/200]	Time 0.737 (0.793)	Data 0.000 (0.052)	Loss 0.654 (0.738)
Epoch: [44][180/200]	Time 0.735 (0.797)	Data 0.001 (0.057)	Loss 0.846 (0.750)
Epoch: [44][200/200]	Time 0.732 (0.800)	Data 0.001 (0.059)	Loss 0.829 (0.772)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.280)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.256 (0.268)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.98229479789734
==> Statistics for epoch 45: 1058 clusters
Epoch: [45][20/200]	Time 0.728 (0.803)	Data 0.001 (0.057)	Loss 0.173 (0.147)
Epoch: [45][40/200]	Time 0.739 (0.814)	Data 0.001 (0.072)	Loss 1.078 (0.284)
Epoch: [45][60/200]	Time 0.738 (0.792)	Data 0.000 (0.048)	Loss 0.706 (0.476)
Epoch: [45][80/200]	Time 0.887 (0.804)	Data 0.001 (0.059)	Loss 1.073 (0.595)
Epoch: [45][100/200]	Time 2.569 (0.810)	Data 1.783 (0.065)	Loss 1.130 (0.639)
Epoch: [45][120/200]	Time 0.737 (0.799)	Data 0.001 (0.055)	Loss 0.928 (0.689)
Epoch: [45][140/200]	Time 0.737 (0.805)	Data 0.001 (0.061)	Loss 0.876 (0.713)
Epoch: [45][160/200]	Time 0.731 (0.797)	Data 0.000 (0.053)	Loss 1.219 (0.743)
Epoch: [45][180/200]	Time 0.864 (0.801)	Data 0.001 (0.057)	Loss 0.752 (0.765)
Epoch: [45][200/200]	Time 0.747 (0.805)	Data 0.001 (0.061)	Loss 0.787 (0.782)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.246 (0.282)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.252 (0.269)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.347100019454956
==> Statistics for epoch 46: 1055 clusters
Epoch: [46][20/200]	Time 0.729 (0.794)	Data 0.001 (0.051)	Loss 0.151 (0.148)
Epoch: [46][40/200]	Time 0.728 (0.808)	Data 0.001 (0.066)	Loss 0.714 (0.296)
Epoch: [46][60/200]	Time 0.820 (0.787)	Data 0.000 (0.045)	Loss 0.801 (0.477)
Epoch: [46][80/200]	Time 0.735 (0.796)	Data 0.001 (0.056)	Loss 1.006 (0.562)
Epoch: [46][100/200]	Time 0.733 (0.802)	Data 0.002 (0.062)	Loss 0.790 (0.625)
Epoch: [46][120/200]	Time 0.747 (0.792)	Data 0.001 (0.052)	Loss 0.922 (0.664)
Epoch: [46][140/200]	Time 0.740 (0.798)	Data 0.001 (0.057)	Loss 1.086 (0.700)
Epoch: [46][160/200]	Time 0.738 (0.792)	Data 0.000 (0.050)	Loss 0.614 (0.722)
Epoch: [46][180/200]	Time 0.740 (0.796)	Data 0.001 (0.054)	Loss 0.984 (0.744)
Epoch: [46][200/200]	Time 0.737 (0.801)	Data 0.001 (0.058)	Loss 0.881 (0.763)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.250 (0.287)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.250 (0.272)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.82338762283325
==> Statistics for epoch 47: 1052 clusters
Epoch: [47][20/200]	Time 0.739 (0.792)	Data 0.001 (0.056)	Loss 0.178 (0.145)
Epoch: [47][40/200]	Time 0.848 (0.808)	Data 0.001 (0.071)	Loss 1.050 (0.302)
Epoch: [47][60/200]	Time 0.735 (0.784)	Data 0.000 (0.047)	Loss 0.684 (0.487)
Epoch: [47][80/200]	Time 0.737 (0.796)	Data 0.001 (0.057)	Loss 1.389 (0.574)
Epoch: [47][100/200]	Time 0.737 (0.804)	Data 0.001 (0.062)	Loss 1.087 (0.645)
Epoch: [47][120/200]	Time 0.862 (0.794)	Data 0.001 (0.052)	Loss 0.880 (0.678)
Epoch: [47][140/200]	Time 0.742 (0.799)	Data 0.001 (0.056)	Loss 0.725 (0.708)
Epoch: [47][160/200]	Time 0.736 (0.792)	Data 0.000 (0.049)	Loss 1.027 (0.733)
Epoch: [47][180/200]	Time 0.740 (0.798)	Data 0.001 (0.055)	Loss 0.694 (0.752)
Epoch: [47][200/200]	Time 0.738 (0.802)	Data 0.001 (0.058)	Loss 0.897 (0.764)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.248 (0.284)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.250 (0.271)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.0897536277771
==> Statistics for epoch 48: 1049 clusters
Epoch: [48][20/200]	Time 0.737 (0.800)	Data 0.001 (0.052)	Loss 0.198 (0.151)
Epoch: [48][40/200]	Time 0.734 (0.814)	Data 0.001 (0.072)	Loss 0.982 (0.289)
Epoch: [48][60/200]	Time 0.733 (0.790)	Data 0.000 (0.048)	Loss 1.175 (0.480)
Epoch: [48][80/200]	Time 0.845 (0.799)	Data 0.001 (0.057)	Loss 0.905 (0.551)
Epoch: [48][100/200]	Time 0.733 (0.803)	Data 0.001 (0.063)	Loss 0.738 (0.625)
Epoch: [48][120/200]	Time 0.735 (0.792)	Data 0.001 (0.053)	Loss 1.206 (0.667)
Epoch: [48][140/200]	Time 0.732 (0.797)	Data 0.001 (0.057)	Loss 0.995 (0.697)
Epoch: [48][160/200]	Time 0.735 (0.789)	Data 0.000 (0.050)	Loss 1.004 (0.725)
Epoch: [48][180/200]	Time 0.742 (0.794)	Data 0.001 (0.054)	Loss 0.570 (0.748)
Epoch: [48][200/200]	Time 0.843 (0.797)	Data 0.001 (0.057)	Loss 0.922 (0.764)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.248 (0.282)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.247 (0.269)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.810118198394775
==> Statistics for epoch 49: 1054 clusters
Epoch: [49][20/200]	Time 0.730 (0.790)	Data 0.001 (0.057)	Loss 0.078 (0.131)
Epoch: [49][40/200]	Time 0.734 (0.812)	Data 0.001 (0.076)	Loss 0.882 (0.270)
Epoch: [49][60/200]	Time 0.732 (0.786)	Data 0.000 (0.051)	Loss 1.003 (0.491)
Epoch: [49][80/200]	Time 0.737 (0.798)	Data 0.001 (0.060)	Loss 0.618 (0.579)
Epoch: [49][100/200]	Time 0.740 (0.804)	Data 0.001 (0.064)	Loss 0.642 (0.649)
Epoch: [49][120/200]	Time 0.740 (0.793)	Data 0.001 (0.054)	Loss 0.803 (0.694)
Epoch: [49][140/200]	Time 0.733 (0.800)	Data 0.001 (0.059)	Loss 0.939 (0.729)
Epoch: [49][160/200]	Time 0.738 (0.793)	Data 0.000 (0.052)	Loss 0.983 (0.751)
Epoch: [49][180/200]	Time 0.740 (0.798)	Data 0.001 (0.056)	Loss 1.512 (0.769)
Epoch: [49][200/200]	Time 0.740 (0.802)	Data 0.001 (0.060)	Loss 0.765 (0.781)
Extract Features: [50/367]	Time 0.246 (0.290)	Data 0.000 (0.033)	
Extract Features: [100/367]	Time 0.251 (0.274)	Data 0.000 (0.017)	
Extract Features: [150/367]	Time 0.254 (0.269)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.249 (0.267)	Data 0.000 (0.009)	
Extract Features: [250/367]	Time 0.249 (0.265)	Data 0.000 (0.007)	
Extract Features: [300/367]	Time 0.249 (0.264)	Data 0.000 (0.006)	
Extract Features: [350/367]	Time 0.248 (0.263)	Data 0.000 (0.005)	
Mean AP: 73.7%

 * Finished epoch  49  model mAP: 73.7%  best: 73.7%

==> Test with the best model:
=> Loaded checkpoint 'log/market2msmt/resnet152_ibn_cion/model_best.pth.tar'
Extract Features: [50/367]	Time 0.253 (0.283)	Data 0.000 (0.029)	
Extract Features: [100/367]	Time 0.247 (0.270)	Data 0.000 (0.015)	
Extract Features: [150/367]	Time 0.250 (0.266)	Data 0.001 (0.010)	
Extract Features: [200/367]	Time 0.247 (0.265)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.247 (0.264)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.248 (0.263)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.248 (0.263)	Data 0.000 (0.005)	
Mean AP: 73.7%
CMC Scores:
  top-1          89.2%
  top-5          94.5%
  top-10         95.6%
Total running time:  4:12:50.090242
