==========
Args:Namespace(dataset='market1501', batch_size=256, workers=4, height=256, width=128, num_instances=8, eps=0.6, eps_gap=0.02, k1=30, k2=6, arch='resnet50', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet50_MSMT17/64bs_lr0.0004_ep120_warm20_seed0/resnet50_120.pth', features=0, dropout=0, momentum=0.2, drop_path_rate=0.3, hw_ratio=2, self_norm=True, conv_stem=False, lr=0.00035, weight_decay=0.0005, optimizer='SGD', epochs=50, iters=200, step_size=20, seed=1, print_freq=20, eval_step=10, temp=0.05, data_dir='/home/ma-user/work/Projects/ReIDNet_Finetune/CContrast/data', logs_dir='log/msmt2market/resnet50_cion', pooling_type='gem', use_hard=True)
==========
==> Load unlabeled dataset
=> Market1501 loaded
Dataset statistics:
  ----------------------------------------
  subset   | # ids | # images | # cameras
  ----------------------------------------
  train    |   751 |    12936 |         6
  query    |   750 |     3368 |         6
  gallery  |   751 |    15913 |         6
  ----------------------------------------
pooling_type: gem
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.101 (0.439)	Data 0.009 (0.032)	
Computing jaccard distance...
Jaccard distance computing time cost: 22.33200240135193
Clustering criterion: eps: 0.600
==> Statistics for epoch 0: 526 clusters
Epoch: [0][20/200]	Time 0.254 (0.756)	Data 0.001 (0.102)	Loss 3.348 (3.599)
Epoch: [0][40/200]	Time 0.252 (0.538)	Data 0.001 (0.084)	Loss 2.484 (3.406)
Epoch: [0][60/200]	Time 0.249 (0.467)	Data 0.000 (0.080)	Loss 2.637 (3.176)
Epoch: [0][80/200]	Time 0.251 (0.431)	Data 0.000 (0.076)	Loss 2.216 (3.017)
Epoch: [0][100/200]	Time 0.253 (0.422)	Data 0.001 (0.086)	Loss 1.822 (2.874)
Epoch: [0][120/200]	Time 0.252 (0.405)	Data 0.001 (0.083)	Loss 2.221 (2.787)
Epoch: [0][140/200]	Time 0.249 (0.392)	Data 0.000 (0.080)	Loss 2.567 (2.724)
Epoch: [0][160/200]	Time 0.254 (0.383)	Data 0.000 (0.078)	Loss 2.086 (2.660)
Epoch: [0][180/200]	Time 0.259 (0.384)	Data 0.001 (0.084)	Loss 1.924 (2.605)
Epoch: [0][200/200]	Time 0.270 (0.378)	Data 0.001 (0.082)	Loss 1.729 (2.561)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.145)	Data 0.000 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.450456619262695
==> Statistics for epoch 1: 575 clusters
Epoch: [1][20/200]	Time 0.252 (0.381)	Data 0.001 (0.126)	Loss 1.996 (0.668)
Epoch: [1][40/200]	Time 0.253 (0.353)	Data 0.001 (0.095)	Loss 1.944 (1.336)
Epoch: [1][60/200]	Time 0.253 (0.345)	Data 0.001 (0.088)	Loss 2.100 (1.585)
Epoch: [1][80/200]	Time 0.254 (0.342)	Data 0.000 (0.085)	Loss 2.237 (1.726)
Epoch: [1][100/200]	Time 0.250 (0.340)	Data 0.000 (0.084)	Loss 1.899 (1.782)
Epoch: [1][120/200]	Time 1.841 (0.353)	Data 1.576 (0.096)	Loss 2.210 (1.826)
Epoch: [1][140/200]	Time 0.259 (0.350)	Data 0.001 (0.093)	Loss 2.066 (1.857)
Epoch: [1][160/200]	Time 0.255 (0.348)	Data 0.001 (0.091)	Loss 1.428 (1.865)
Epoch: [1][180/200]	Time 0.253 (0.348)	Data 0.001 (0.090)	Loss 1.879 (1.870)
Epoch: [1][200/200]	Time 0.254 (0.347)	Data 0.000 (0.089)	Loss 1.711 (1.873)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.134)	Data 0.000 (0.037)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.168297290802002
==> Statistics for epoch 2: 588 clusters
Epoch: [2][20/200]	Time 0.264 (0.375)	Data 0.001 (0.117)	Loss 1.812 (0.468)
Epoch: [2][40/200]	Time 0.253 (0.362)	Data 0.001 (0.102)	Loss 1.828 (1.150)
Epoch: [2][60/200]	Time 0.258 (0.347)	Data 0.001 (0.089)	Loss 1.757 (1.389)
Epoch: [2][80/200]	Time 0.264 (0.341)	Data 0.002 (0.082)	Loss 1.426 (1.530)
Epoch: [2][100/200]	Time 0.254 (0.340)	Data 0.001 (0.082)	Loss 1.918 (1.600)
Epoch: [2][120/200]	Time 0.254 (0.339)	Data 0.001 (0.081)	Loss 1.907 (1.640)
Epoch: [2][140/200]	Time 0.252 (0.338)	Data 0.000 (0.080)	Loss 1.608 (1.673)
Epoch: [2][160/200]	Time 0.254 (0.338)	Data 0.000 (0.080)	Loss 1.757 (1.697)
Epoch: [2][180/200]	Time 0.253 (0.338)	Data 0.000 (0.080)	Loss 1.769 (1.713)
Epoch: [2][200/200]	Time 0.256 (0.344)	Data 0.001 (0.086)	Loss 1.698 (1.725)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.094 (0.133)	Data 0.000 (0.035)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.191498279571533
==> Statistics for epoch 3: 579 clusters
Epoch: [3][20/200]	Time 0.252 (0.373)	Data 0.001 (0.113)	Loss 1.775 (0.433)
Epoch: [3][40/200]	Time 0.257 (0.353)	Data 0.001 (0.094)	Loss 1.812 (1.063)
Epoch: [3][60/200]	Time 0.254 (0.347)	Data 0.001 (0.089)	Loss 1.190 (1.308)
Epoch: [3][80/200]	Time 0.253 (0.345)	Data 0.001 (0.087)	Loss 1.358 (1.392)
Epoch: [3][100/200]	Time 0.253 (0.342)	Data 0.001 (0.085)	Loss 2.078 (1.446)
Epoch: [3][120/200]	Time 0.254 (0.341)	Data 0.000 (0.084)	Loss 1.877 (1.502)
Epoch: [3][140/200]	Time 0.253 (0.341)	Data 0.000 (0.083)	Loss 1.965 (1.529)
Epoch: [3][160/200]	Time 0.252 (0.340)	Data 0.000 (0.082)	Loss 1.449 (1.543)
Epoch: [3][180/200]	Time 0.252 (0.339)	Data 0.000 (0.081)	Loss 1.694 (1.565)
Epoch: [3][200/200]	Time 0.257 (0.347)	Data 0.001 (0.088)	Loss 1.555 (1.569)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.097 (0.133)	Data 0.004 (0.037)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.951987266540527
==> Statistics for epoch 4: 593 clusters
Epoch: [4][20/200]	Time 0.258 (0.373)	Data 0.000 (0.118)	Loss 1.699 (0.411)
Epoch: [4][40/200]	Time 0.255 (0.350)	Data 0.001 (0.093)	Loss 1.630 (1.009)
Epoch: [4][60/200]	Time 0.259 (0.343)	Data 0.001 (0.086)	Loss 1.783 (1.216)
Epoch: [4][80/200]	Time 0.256 (0.343)	Data 0.001 (0.084)	Loss 1.897 (1.316)
Epoch: [4][100/200]	Time 0.254 (0.340)	Data 0.001 (0.082)	Loss 1.481 (1.377)
Epoch: [4][120/200]	Time 0.253 (0.340)	Data 0.000 (0.082)	Loss 1.271 (1.399)
Epoch: [4][140/200]	Time 0.250 (0.337)	Data 0.000 (0.079)	Loss 1.965 (1.433)
Epoch: [4][160/200]	Time 0.257 (0.335)	Data 0.001 (0.077)	Loss 1.439 (1.455)
Epoch: [4][180/200]	Time 0.253 (0.336)	Data 0.000 (0.078)	Loss 1.561 (1.476)
Epoch: [4][200/200]	Time 0.260 (0.343)	Data 0.001 (0.085)	Loss 1.452 (1.487)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.122 (0.133)	Data 0.027 (0.033)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.004714965820312
==> Statistics for epoch 5: 592 clusters
Epoch: [5][20/200]	Time 0.254 (0.364)	Data 0.001 (0.102)	Loss 1.339 (0.344)
Epoch: [5][40/200]	Time 0.360 (0.347)	Data 0.001 (0.086)	Loss 1.759 (0.921)
Epoch: [5][60/200]	Time 0.254 (0.337)	Data 0.001 (0.079)	Loss 1.312 (1.124)
Epoch: [5][80/200]	Time 0.257 (0.337)	Data 0.001 (0.078)	Loss 1.612 (1.239)
Epoch: [5][100/200]	Time 0.254 (0.334)	Data 0.001 (0.076)	Loss 1.387 (1.324)
Epoch: [5][120/200]	Time 0.253 (0.334)	Data 0.000 (0.076)	Loss 1.517 (1.368)
Epoch: [5][140/200]	Time 0.254 (0.334)	Data 0.000 (0.077)	Loss 1.511 (1.394)
Epoch: [5][160/200]	Time 0.253 (0.334)	Data 0.000 (0.077)	Loss 1.659 (1.415)
Epoch: [5][180/200]	Time 0.252 (0.335)	Data 0.000 (0.077)	Loss 1.630 (1.430)
Epoch: [5][200/200]	Time 0.253 (0.341)	Data 0.001 (0.083)	Loss 1.739 (1.443)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.132)	Data 0.000 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.990028142929077
==> Statistics for epoch 6: 598 clusters
Epoch: [6][20/200]	Time 0.261 (0.369)	Data 0.001 (0.106)	Loss 1.398 (0.337)
Epoch: [6][40/200]	Time 0.254 (0.343)	Data 0.001 (0.084)	Loss 1.180 (0.854)
Epoch: [6][60/200]	Time 0.253 (0.341)	Data 0.001 (0.082)	Loss 1.863 (1.086)
Epoch: [6][80/200]	Time 0.255 (0.335)	Data 0.001 (0.077)	Loss 1.233 (1.196)
Epoch: [6][100/200]	Time 0.251 (0.334)	Data 0.001 (0.077)	Loss 1.443 (1.256)
Epoch: [6][120/200]	Time 0.255 (0.334)	Data 0.000 (0.076)	Loss 1.631 (1.293)
Epoch: [6][140/200]	Time 0.254 (0.333)	Data 0.000 (0.075)	Loss 1.523 (1.317)
Epoch: [6][160/200]	Time 0.253 (0.333)	Data 0.000 (0.075)	Loss 1.155 (1.337)
Epoch: [6][180/200]	Time 0.255 (0.332)	Data 0.000 (0.074)	Loss 1.365 (1.345)
Epoch: [6][200/200]	Time 0.257 (0.338)	Data 0.000 (0.080)	Loss 1.329 (1.352)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.135)	Data 0.000 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.806238651275635
==> Statistics for epoch 7: 599 clusters
Epoch: [7][20/200]	Time 0.255 (0.366)	Data 0.001 (0.107)	Loss 1.405 (0.355)
Epoch: [7][40/200]	Time 0.260 (0.356)	Data 0.001 (0.096)	Loss 1.589 (0.860)
Epoch: [7][60/200]	Time 0.253 (0.347)	Data 0.001 (0.089)	Loss 1.420 (1.065)
Epoch: [7][80/200]	Time 0.254 (0.345)	Data 0.001 (0.086)	Loss 2.043 (1.178)
Epoch: [7][100/200]	Time 0.253 (0.340)	Data 0.001 (0.082)	Loss 1.432 (1.231)
Epoch: [7][120/200]	Time 0.254 (0.338)	Data 0.000 (0.080)	Loss 1.705 (1.271)
Epoch: [7][140/200]	Time 0.253 (0.336)	Data 0.000 (0.078)	Loss 1.592 (1.293)
Epoch: [7][160/200]	Time 0.252 (0.336)	Data 0.000 (0.077)	Loss 1.030 (1.314)
Epoch: [7][180/200]	Time 0.253 (0.334)	Data 0.000 (0.076)	Loss 1.196 (1.328)
Epoch: [7][200/200]	Time 0.258 (0.341)	Data 0.001 (0.083)	Loss 1.509 (1.335)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.132)	Data 0.000 (0.033)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.892050981521606
==> Statistics for epoch 8: 603 clusters
Epoch: [8][20/200]	Time 0.252 (0.360)	Data 0.001 (0.105)	Loss 1.296 (0.355)
Epoch: [8][40/200]	Time 0.256 (0.344)	Data 0.001 (0.086)	Loss 1.672 (0.822)
Epoch: [8][60/200]	Time 0.252 (0.337)	Data 0.001 (0.080)	Loss 1.225 (1.030)
Epoch: [8][80/200]	Time 0.254 (0.333)	Data 0.001 (0.075)	Loss 1.558 (1.115)
Epoch: [8][100/200]	Time 0.256 (0.331)	Data 0.001 (0.074)	Loss 1.564 (1.187)
Epoch: [8][120/200]	Time 0.252 (0.331)	Data 0.000 (0.074)	Loss 1.330 (1.215)
Epoch: [8][140/200]	Time 0.252 (0.330)	Data 0.000 (0.073)	Loss 1.120 (1.246)
Epoch: [8][160/200]	Time 0.338 (0.329)	Data 0.000 (0.072)	Loss 1.516 (1.261)
Epoch: [8][180/200]	Time 0.256 (0.329)	Data 0.000 (0.072)	Loss 1.268 (1.270)
Epoch: [8][200/200]	Time 0.259 (0.338)	Data 0.001 (0.080)	Loss 1.332 (1.281)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.133)	Data 0.000 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.999691247940063
==> Statistics for epoch 9: 603 clusters
Epoch: [9][20/200]	Time 0.254 (0.358)	Data 0.001 (0.098)	Loss 1.121 (0.319)
Epoch: [9][40/200]	Time 0.260 (0.344)	Data 0.002 (0.087)	Loss 1.774 (0.813)
Epoch: [9][60/200]	Time 0.253 (0.343)	Data 0.001 (0.085)	Loss 1.540 (1.008)
Epoch: [9][80/200]	Time 0.252 (0.339)	Data 0.001 (0.081)	Loss 1.542 (1.095)
Epoch: [9][100/200]	Time 0.269 (0.336)	Data 0.001 (0.078)	Loss 1.282 (1.148)
Epoch: [9][120/200]	Time 0.253 (0.334)	Data 0.000 (0.076)	Loss 1.195 (1.171)
Epoch: [9][140/200]	Time 0.252 (0.333)	Data 0.000 (0.075)	Loss 1.567 (1.195)
Epoch: [9][160/200]	Time 0.250 (0.331)	Data 0.000 (0.074)	Loss 1.185 (1.205)
Epoch: [9][180/200]	Time 0.251 (0.332)	Data 0.000 (0.075)	Loss 1.164 (1.226)
Epoch: [9][200/200]	Time 0.259 (0.338)	Data 0.001 (0.081)	Loss 1.480 (1.246)
Extract Features: [50/76]	Time 0.130 (0.134)	Data 0.037 (0.038)	
Mean AP: 89.2%

 * Finished epoch   9  model mAP: 89.2%  best: 89.2% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.130)	Data 0.000 (0.030)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.50595736503601
==> Statistics for epoch 10: 606 clusters
Epoch: [10][20/200]	Time 0.255 (0.402)	Data 0.001 (0.123)	Loss 1.367 (0.308)
Epoch: [10][40/200]	Time 0.254 (0.366)	Data 0.001 (0.097)	Loss 1.128 (0.764)
Epoch: [10][60/200]	Time 0.252 (0.354)	Data 0.001 (0.089)	Loss 1.019 (0.928)
Epoch: [10][80/200]	Time 0.254 (0.348)	Data 0.001 (0.085)	Loss 1.443 (1.018)
Epoch: [10][100/200]	Time 0.255 (0.345)	Data 0.001 (0.082)	Loss 1.915 (1.085)
Epoch: [10][120/200]	Time 0.358 (0.343)	Data 0.000 (0.080)	Loss 1.633 (1.127)
Epoch: [10][140/200]	Time 0.255 (0.341)	Data 0.000 (0.079)	Loss 1.138 (1.152)
Epoch: [10][160/200]	Time 0.256 (0.341)	Data 0.000 (0.079)	Loss 1.372 (1.173)
Epoch: [10][180/200]	Time 0.253 (0.340)	Data 0.000 (0.079)	Loss 1.146 (1.180)
Epoch: [10][200/200]	Time 0.256 (0.347)	Data 0.001 (0.086)	Loss 1.411 (1.190)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.137)	Data 0.000 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.66493272781372
==> Statistics for epoch 11: 604 clusters
Epoch: [11][20/200]	Time 0.255 (0.380)	Data 0.001 (0.122)	Loss 1.072 (0.279)
Epoch: [11][40/200]	Time 0.254 (0.356)	Data 0.001 (0.096)	Loss 1.284 (0.729)
Epoch: [11][60/200]	Time 0.254 (0.344)	Data 0.001 (0.086)	Loss 1.337 (0.908)
Epoch: [11][80/200]	Time 0.255 (0.343)	Data 0.001 (0.084)	Loss 1.270 (1.009)
Epoch: [11][100/200]	Time 0.263 (0.339)	Data 0.001 (0.080)	Loss 0.898 (1.051)
Epoch: [11][120/200]	Time 0.254 (0.339)	Data 0.000 (0.079)	Loss 1.212 (1.095)
Epoch: [11][140/200]	Time 0.253 (0.337)	Data 0.000 (0.078)	Loss 1.271 (1.120)
Epoch: [11][160/200]	Time 0.255 (0.335)	Data 0.000 (0.077)	Loss 1.235 (1.135)
Epoch: [11][180/200]	Time 0.252 (0.335)	Data 0.000 (0.077)	Loss 1.172 (1.146)
Epoch: [11][200/200]	Time 0.256 (0.340)	Data 0.001 (0.081)	Loss 1.210 (1.157)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.095 (0.133)	Data 0.000 (0.033)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.949729204177856
==> Statistics for epoch 12: 603 clusters
Epoch: [12][20/200]	Time 0.259 (0.368)	Data 0.001 (0.106)	Loss 1.047 (0.298)
Epoch: [12][40/200]	Time 0.254 (0.348)	Data 0.001 (0.090)	Loss 1.706 (0.750)
Epoch: [12][60/200]	Time 0.251 (0.344)	Data 0.001 (0.087)	Loss 1.027 (0.926)
Epoch: [12][80/200]	Time 0.256 (0.339)	Data 0.001 (0.082)	Loss 1.149 (1.010)
Epoch: [12][100/200]	Time 0.256 (0.336)	Data 0.001 (0.078)	Loss 1.147 (1.051)
Epoch: [12][120/200]	Time 0.255 (0.335)	Data 0.000 (0.077)	Loss 1.218 (1.079)
Epoch: [12][140/200]	Time 0.254 (0.335)	Data 0.000 (0.077)	Loss 1.255 (1.096)
Epoch: [12][160/200]	Time 0.253 (0.335)	Data 0.000 (0.077)	Loss 0.929 (1.110)
Epoch: [12][180/200]	Time 0.252 (0.334)	Data 0.000 (0.075)	Loss 1.108 (1.122)
Epoch: [12][200/200]	Time 0.259 (0.341)	Data 0.001 (0.083)	Loss 1.110 (1.135)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.139)	Data 0.000 (0.041)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.9231858253479
==> Statistics for epoch 13: 599 clusters
Epoch: [13][20/200]	Time 0.255 (0.379)	Data 0.001 (0.114)	Loss 0.898 (0.260)
Epoch: [13][40/200]	Time 0.329 (0.355)	Data 0.001 (0.093)	Loss 1.354 (0.713)
Epoch: [13][60/200]	Time 0.253 (0.348)	Data 0.001 (0.089)	Loss 1.164 (0.890)
Epoch: [13][80/200]	Time 0.253 (0.345)	Data 0.001 (0.085)	Loss 1.050 (0.956)
Epoch: [13][100/200]	Time 0.373 (0.342)	Data 0.002 (0.083)	Loss 1.333 (1.007)
Epoch: [13][120/200]	Time 0.254 (0.341)	Data 0.000 (0.082)	Loss 1.120 (1.038)
Epoch: [13][140/200]	Time 0.253 (0.339)	Data 0.000 (0.080)	Loss 1.006 (1.056)
Epoch: [13][160/200]	Time 0.255 (0.339)	Data 0.000 (0.079)	Loss 1.368 (1.076)
Epoch: [13][180/200]	Time 0.253 (0.338)	Data 0.000 (0.079)	Loss 1.430 (1.088)
Epoch: [13][200/200]	Time 0.254 (0.345)	Data 0.001 (0.086)	Loss 1.274 (1.095)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.138)	Data 0.000 (0.040)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.715959310531616
==> Statistics for epoch 14: 603 clusters
Epoch: [14][20/200]	Time 0.260 (0.384)	Data 0.000 (0.121)	Loss 0.960 (0.264)
Epoch: [14][40/200]	Time 0.395 (0.355)	Data 0.001 (0.093)	Loss 0.943 (0.713)
Epoch: [14][60/200]	Time 0.253 (0.346)	Data 0.001 (0.085)	Loss 1.084 (0.846)
Epoch: [14][80/200]	Time 0.254 (0.342)	Data 0.001 (0.084)	Loss 1.400 (0.944)
Epoch: [14][100/200]	Time 0.254 (0.338)	Data 0.001 (0.081)	Loss 0.769 (0.972)
Epoch: [14][120/200]	Time 0.255 (0.336)	Data 0.000 (0.079)	Loss 1.167 (1.007)
Epoch: [14][140/200]	Time 0.251 (0.335)	Data 0.000 (0.078)	Loss 1.238 (1.030)
Epoch: [14][160/200]	Time 0.253 (0.333)	Data 0.000 (0.076)	Loss 1.103 (1.050)
Epoch: [14][180/200]	Time 0.252 (0.332)	Data 0.000 (0.075)	Loss 1.031 (1.060)
Epoch: [14][200/200]	Time 0.254 (0.339)	Data 0.000 (0.081)	Loss 1.408 (1.073)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.153 (0.134)	Data 0.062 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.674717664718628
==> Statistics for epoch 15: 603 clusters
Epoch: [15][20/200]	Time 0.254 (0.363)	Data 0.001 (0.105)	Loss 1.036 (0.270)
Epoch: [15][40/200]	Time 0.360 (0.348)	Data 0.001 (0.088)	Loss 1.252 (0.682)
Epoch: [15][60/200]	Time 0.253 (0.340)	Data 0.001 (0.082)	Loss 0.856 (0.829)
Epoch: [15][80/200]	Time 0.257 (0.338)	Data 0.001 (0.080)	Loss 1.201 (0.900)
Epoch: [15][100/200]	Time 0.256 (0.336)	Data 0.001 (0.078)	Loss 1.031 (0.952)
Epoch: [15][120/200]	Time 0.252 (0.336)	Data 0.000 (0.078)	Loss 1.175 (0.990)
Epoch: [15][140/200]	Time 0.252 (0.335)	Data 0.000 (0.077)	Loss 1.056 (1.006)
Epoch: [15][160/200]	Time 0.252 (0.334)	Data 0.000 (0.076)	Loss 1.419 (1.025)
Epoch: [15][180/200]	Time 0.251 (0.333)	Data 0.000 (0.075)	Loss 1.289 (1.039)
Epoch: [15][200/200]	Time 0.253 (0.340)	Data 0.001 (0.083)	Loss 1.172 (1.045)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.137)	Data 0.000 (0.039)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.592423915863037
==> Statistics for epoch 16: 602 clusters
Epoch: [16][20/200]	Time 0.255 (0.364)	Data 0.001 (0.100)	Loss 1.165 (0.285)
Epoch: [16][40/200]	Time 0.255 (0.346)	Data 0.001 (0.085)	Loss 1.252 (0.711)
Epoch: [16][60/200]	Time 0.253 (0.339)	Data 0.001 (0.081)	Loss 1.315 (0.856)
Epoch: [16][80/200]	Time 0.253 (0.336)	Data 0.001 (0.077)	Loss 1.077 (0.925)
Epoch: [16][100/200]	Time 0.351 (0.337)	Data 0.001 (0.078)	Loss 1.095 (0.963)
Epoch: [16][120/200]	Time 0.253 (0.336)	Data 0.000 (0.077)	Loss 1.259 (0.989)
Epoch: [16][140/200]	Time 0.253 (0.334)	Data 0.000 (0.075)	Loss 1.180 (1.004)
Epoch: [16][160/200]	Time 0.253 (0.332)	Data 0.000 (0.074)	Loss 1.024 (1.021)
Epoch: [16][180/200]	Time 0.253 (0.332)	Data 0.000 (0.073)	Loss 1.016 (1.030)
Epoch: [16][200/200]	Time 0.256 (0.338)	Data 0.001 (0.080)	Loss 1.107 (1.035)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.130)	Data 0.000 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.005675792694092
==> Statistics for epoch 17: 597 clusters
Epoch: [17][20/200]	Time 0.304 (0.360)	Data 0.001 (0.099)	Loss 1.192 (0.277)
Epoch: [17][40/200]	Time 0.260 (0.344)	Data 0.001 (0.085)	Loss 0.785 (0.670)
Epoch: [17][60/200]	Time 0.254 (0.340)	Data 0.001 (0.081)	Loss 1.028 (0.812)
Epoch: [17][80/200]	Time 0.354 (0.337)	Data 0.001 (0.077)	Loss 1.038 (0.880)
Epoch: [17][100/200]	Time 0.254 (0.333)	Data 0.001 (0.075)	Loss 1.141 (0.923)
Epoch: [17][120/200]	Time 0.252 (0.332)	Data 0.000 (0.073)	Loss 1.116 (0.939)
Epoch: [17][140/200]	Time 0.253 (0.331)	Data 0.000 (0.073)	Loss 1.358 (0.966)
Epoch: [17][160/200]	Time 0.253 (0.330)	Data 0.000 (0.072)	Loss 0.593 (0.979)
Epoch: [17][180/200]	Time 0.252 (0.331)	Data 0.000 (0.073)	Loss 0.991 (0.991)
Epoch: [17][200/200]	Time 0.258 (0.339)	Data 0.001 (0.081)	Loss 1.118 (1.003)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.134)	Data 0.000 (0.037)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.746254444122314
==> Statistics for epoch 18: 601 clusters
Epoch: [18][20/200]	Time 0.254 (0.367)	Data 0.001 (0.111)	Loss 1.072 (0.253)
Epoch: [18][40/200]	Time 0.253 (0.352)	Data 0.001 (0.097)	Loss 1.113 (0.648)
Epoch: [18][60/200]	Time 0.254 (0.348)	Data 0.001 (0.091)	Loss 1.363 (0.800)
Epoch: [18][80/200]	Time 0.255 (0.342)	Data 0.001 (0.084)	Loss 1.024 (0.868)
Epoch: [18][100/200]	Time 0.253 (0.340)	Data 0.000 (0.083)	Loss 1.010 (0.912)
Epoch: [18][120/200]	Time 0.253 (0.340)	Data 0.000 (0.082)	Loss 0.917 (0.944)
Epoch: [18][140/200]	Time 0.252 (0.337)	Data 0.000 (0.080)	Loss 0.634 (0.965)
Epoch: [18][160/200]	Time 0.252 (0.336)	Data 0.000 (0.078)	Loss 1.032 (0.980)
Epoch: [18][180/200]	Time 0.251 (0.335)	Data 0.000 (0.077)	Loss 0.991 (0.986)
Epoch: [18][200/200]	Time 0.259 (0.341)	Data 0.001 (0.083)	Loss 1.036 (0.995)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.163 (0.132)	Data 0.071 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.144349098205566
==> Statistics for epoch 19: 596 clusters
Epoch: [19][20/200]	Time 0.252 (0.378)	Data 0.001 (0.114)	Loss 1.035 (0.236)
Epoch: [19][40/200]	Time 0.253 (0.358)	Data 0.001 (0.094)	Loss 0.873 (0.623)
Epoch: [19][60/200]	Time 0.252 (0.348)	Data 0.001 (0.087)	Loss 0.960 (0.761)
Epoch: [19][80/200]	Time 0.371 (0.346)	Data 0.001 (0.085)	Loss 0.867 (0.835)
Epoch: [19][100/200]	Time 0.253 (0.343)	Data 0.001 (0.083)	Loss 0.898 (0.879)
Epoch: [19][120/200]	Time 0.254 (0.344)	Data 0.001 (0.083)	Loss 0.979 (0.907)
Epoch: [19][140/200]	Time 0.253 (0.342)	Data 0.000 (0.082)	Loss 1.070 (0.920)
Epoch: [19][160/200]	Time 0.253 (0.340)	Data 0.000 (0.081)	Loss 1.317 (0.944)
Epoch: [19][180/200]	Time 0.254 (0.339)	Data 0.000 (0.079)	Loss 1.080 (0.960)
Epoch: [19][200/200]	Time 0.254 (0.346)	Data 0.001 (0.086)	Loss 1.021 (0.972)
Extract Features: [50/76]	Time 0.093 (0.135)	Data 0.000 (0.037)	
Mean AP: 90.1%

 * Finished epoch  19  model mAP: 90.1%  best: 90.1% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.135)	Data 0.000 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.71184539794922
==> Statistics for epoch 20: 592 clusters
Epoch: [20][20/200]	Time 0.253 (0.362)	Data 0.001 (0.106)	Loss 1.024 (0.233)
Epoch: [20][40/200]	Time 0.253 (0.344)	Data 0.000 (0.088)	Loss 0.915 (0.601)
Epoch: [20][60/200]	Time 0.253 (0.339)	Data 0.001 (0.081)	Loss 1.445 (0.735)
Epoch: [20][80/200]	Time 0.254 (0.334)	Data 0.001 (0.077)	Loss 1.050 (0.815)
Epoch: [20][100/200]	Time 0.256 (0.335)	Data 0.001 (0.077)	Loss 1.207 (0.869)
Epoch: [20][120/200]	Time 0.257 (0.334)	Data 0.000 (0.076)	Loss 0.957 (0.888)
Epoch: [20][140/200]	Time 0.253 (0.333)	Data 0.000 (0.075)	Loss 0.800 (0.909)
Epoch: [20][160/200]	Time 0.250 (0.332)	Data 0.000 (0.074)	Loss 0.848 (0.914)
Epoch: [20][180/200]	Time 0.250 (0.330)	Data 0.000 (0.073)	Loss 0.984 (0.927)
Epoch: [20][200/200]	Time 0.255 (0.339)	Data 0.001 (0.080)	Loss 0.918 (0.934)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.132)	Data 0.000 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.150730848312378
==> Statistics for epoch 21: 593 clusters
Epoch: [21][20/200]	Time 0.263 (0.364)	Data 0.001 (0.105)	Loss 1.331 (0.236)
Epoch: [21][40/200]	Time 0.253 (0.346)	Data 0.001 (0.088)	Loss 1.158 (0.602)
Epoch: [21][60/200]	Time 0.255 (0.338)	Data 0.001 (0.081)	Loss 0.967 (0.741)
Epoch: [21][80/200]	Time 0.254 (0.336)	Data 0.001 (0.079)	Loss 0.761 (0.811)
Epoch: [21][100/200]	Time 0.258 (0.334)	Data 0.001 (0.077)	Loss 0.999 (0.847)
Epoch: [21][120/200]	Time 0.252 (0.333)	Data 0.000 (0.076)	Loss 0.687 (0.864)
Epoch: [21][140/200]	Time 0.251 (0.331)	Data 0.000 (0.074)	Loss 1.239 (0.883)
Epoch: [21][160/200]	Time 0.255 (0.330)	Data 0.000 (0.073)	Loss 0.908 (0.900)
Epoch: [21][180/200]	Time 0.344 (0.329)	Data 0.000 (0.072)	Loss 0.773 (0.910)
Epoch: [21][200/200]	Time 0.259 (0.335)	Data 0.001 (0.078)	Loss 0.885 (0.921)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.135)	Data 0.000 (0.037)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.7700252532959
==> Statistics for epoch 22: 592 clusters
Epoch: [22][20/200]	Time 0.254 (0.377)	Data 0.001 (0.115)	Loss 0.876 (0.213)
Epoch: [22][40/200]	Time 0.254 (0.354)	Data 0.001 (0.094)	Loss 0.904 (0.576)
Epoch: [22][60/200]	Time 0.253 (0.348)	Data 0.001 (0.087)	Loss 0.821 (0.723)
Epoch: [22][80/200]	Time 0.253 (0.346)	Data 0.001 (0.085)	Loss 0.775 (0.794)
Epoch: [22][100/200]	Time 0.253 (0.343)	Data 0.001 (0.084)	Loss 1.243 (0.839)
Epoch: [22][120/200]	Time 0.252 (0.343)	Data 0.000 (0.083)	Loss 1.283 (0.865)
Epoch: [22][140/200]	Time 0.259 (0.342)	Data 0.000 (0.082)	Loss 1.051 (0.885)
Epoch: [22][160/200]	Time 0.253 (0.341)	Data 0.000 (0.081)	Loss 0.722 (0.907)
Epoch: [22][180/200]	Time 0.255 (0.340)	Data 0.000 (0.080)	Loss 1.015 (0.918)
Epoch: [22][200/200]	Time 0.258 (0.346)	Data 0.001 (0.086)	Loss 0.892 (0.928)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.138)	Data 0.000 (0.040)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.19438099861145
==> Statistics for epoch 23: 594 clusters
Epoch: [23][20/200]	Time 0.257 (0.373)	Data 0.001 (0.108)	Loss 0.896 (0.207)
Epoch: [23][40/200]	Time 0.252 (0.351)	Data 0.001 (0.088)	Loss 0.923 (0.572)
Epoch: [23][60/200]	Time 0.254 (0.340)	Data 0.001 (0.080)	Loss 0.699 (0.693)
Epoch: [23][80/200]	Time 0.366 (0.337)	Data 0.001 (0.077)	Loss 0.994 (0.769)
Epoch: [23][100/200]	Time 0.255 (0.334)	Data 0.001 (0.075)	Loss 1.087 (0.823)
Epoch: [23][120/200]	Time 0.252 (0.332)	Data 0.000 (0.073)	Loss 1.174 (0.848)
Epoch: [23][140/200]	Time 0.251 (0.331)	Data 0.000 (0.072)	Loss 0.938 (0.874)
Epoch: [23][160/200]	Time 0.253 (0.330)	Data 0.000 (0.072)	Loss 0.720 (0.891)
Epoch: [23][180/200]	Time 0.252 (0.330)	Data 0.000 (0.072)	Loss 0.727 (0.898)
Epoch: [23][200/200]	Time 0.253 (0.336)	Data 0.001 (0.078)	Loss 1.144 (0.900)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.109 (0.132)	Data 0.017 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.50299644470215
==> Statistics for epoch 24: 595 clusters
Epoch: [24][20/200]	Time 0.260 (0.367)	Data 0.001 (0.109)	Loss 0.710 (0.210)
Epoch: [24][40/200]	Time 0.253 (0.344)	Data 0.001 (0.088)	Loss 0.914 (0.586)
Epoch: [24][60/200]	Time 0.255 (0.338)	Data 0.001 (0.080)	Loss 1.047 (0.709)
Epoch: [24][80/200]	Time 0.257 (0.335)	Data 0.001 (0.076)	Loss 0.765 (0.774)
Epoch: [24][100/200]	Time 0.252 (0.332)	Data 0.001 (0.074)	Loss 1.304 (0.815)
Epoch: [24][120/200]	Time 0.254 (0.330)	Data 0.000 (0.072)	Loss 1.171 (0.833)
Epoch: [24][140/200]	Time 0.253 (0.330)	Data 0.000 (0.072)	Loss 1.175 (0.860)
Epoch: [24][160/200]	Time 0.253 (0.329)	Data 0.000 (0.071)	Loss 0.976 (0.876)
Epoch: [24][180/200]	Time 0.255 (0.329)	Data 0.000 (0.070)	Loss 0.813 (0.888)
Epoch: [24][200/200]	Time 0.261 (0.336)	Data 0.001 (0.078)	Loss 0.781 (0.897)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.139)	Data 0.000 (0.041)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.79431986808777
==> Statistics for epoch 25: 594 clusters
Epoch: [25][20/200]	Time 0.256 (0.380)	Data 0.000 (0.116)	Loss 1.053 (0.201)
Epoch: [25][40/200]	Time 0.256 (0.357)	Data 0.001 (0.097)	Loss 1.030 (0.555)
Epoch: [25][60/200]	Time 0.255 (0.349)	Data 0.001 (0.089)	Loss 0.842 (0.708)
Epoch: [25][80/200]	Time 0.255 (0.346)	Data 0.000 (0.085)	Loss 1.071 (0.783)
Epoch: [25][100/200]	Time 0.256 (0.343)	Data 0.001 (0.083)	Loss 1.210 (0.827)
Epoch: [25][120/200]	Time 0.253 (0.338)	Data 0.000 (0.079)	Loss 0.746 (0.847)
Epoch: [25][140/200]	Time 0.253 (0.338)	Data 0.000 (0.078)	Loss 1.181 (0.859)
Epoch: [25][160/200]	Time 0.254 (0.337)	Data 0.000 (0.077)	Loss 0.881 (0.876)
Epoch: [25][180/200]	Time 0.252 (0.336)	Data 0.000 (0.076)	Loss 0.905 (0.886)
Epoch: [25][200/200]	Time 0.256 (0.341)	Data 0.001 (0.082)	Loss 1.049 (0.893)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.131)	Data 0.000 (0.032)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.13063144683838
==> Statistics for epoch 26: 593 clusters
Epoch: [26][20/200]	Time 0.254 (0.359)	Data 0.001 (0.101)	Loss 0.739 (0.217)
Epoch: [26][40/200]	Time 0.254 (0.346)	Data 0.001 (0.085)	Loss 0.997 (0.595)
Epoch: [26][60/200]	Time 0.254 (0.337)	Data 0.001 (0.078)	Loss 1.368 (0.709)
Epoch: [26][80/200]	Time 0.253 (0.334)	Data 0.001 (0.076)	Loss 1.061 (0.792)
Epoch: [26][100/200]	Time 0.255 (0.333)	Data 0.001 (0.075)	Loss 1.018 (0.824)
Epoch: [26][120/200]	Time 0.253 (0.332)	Data 0.000 (0.073)	Loss 1.193 (0.850)
Epoch: [26][140/200]	Time 0.254 (0.331)	Data 0.000 (0.073)	Loss 0.801 (0.875)
Epoch: [26][160/200]	Time 0.252 (0.331)	Data 0.000 (0.073)	Loss 0.901 (0.891)
Epoch: [26][180/200]	Time 0.252 (0.331)	Data 0.000 (0.073)	Loss 1.128 (0.904)
Epoch: [26][200/200]	Time 0.256 (0.337)	Data 0.001 (0.079)	Loss 0.765 (0.909)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.132)	Data 0.000 (0.034)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.541415691375732
==> Statistics for epoch 27: 593 clusters
Epoch: [27][20/200]	Time 0.254 (0.365)	Data 0.001 (0.108)	Loss 0.891 (0.213)
Epoch: [27][40/200]	Time 0.253 (0.344)	Data 0.001 (0.086)	Loss 1.027 (0.550)
Epoch: [27][60/200]	Time 0.253 (0.337)	Data 0.001 (0.079)	Loss 0.903 (0.707)
Epoch: [27][80/200]	Time 0.253 (0.334)	Data 0.001 (0.076)	Loss 0.693 (0.761)
Epoch: [27][100/200]	Time 0.256 (0.330)	Data 0.001 (0.073)	Loss 0.967 (0.797)
Epoch: [27][120/200]	Time 0.253 (0.331)	Data 0.000 (0.073)	Loss 1.029 (0.832)
Epoch: [27][140/200]	Time 0.252 (0.331)	Data 0.000 (0.072)	Loss 0.875 (0.862)
Epoch: [27][160/200]	Time 0.252 (0.330)	Data 0.000 (0.072)	Loss 0.948 (0.873)
Epoch: [27][180/200]	Time 0.255 (0.330)	Data 0.000 (0.071)	Loss 1.010 (0.877)
Epoch: [27][200/200]	Time 0.287 (0.337)	Data 0.002 (0.078)	Loss 1.193 (0.884)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.133)	Data 0.000 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.86924409866333
==> Statistics for epoch 28: 596 clusters
Epoch: [28][20/200]	Time 0.255 (0.366)	Data 0.001 (0.108)	Loss 0.952 (0.207)
Epoch: [28][40/200]	Time 0.367 (0.346)	Data 0.001 (0.086)	Loss 0.938 (0.576)
Epoch: [28][60/200]	Time 0.253 (0.338)	Data 0.001 (0.080)	Loss 0.947 (0.703)
Epoch: [28][80/200]	Time 0.256 (0.334)	Data 0.001 (0.077)	Loss 1.020 (0.774)
Epoch: [28][100/200]	Time 0.254 (0.332)	Data 0.001 (0.075)	Loss 0.903 (0.815)
Epoch: [28][120/200]	Time 0.251 (0.331)	Data 0.000 (0.073)	Loss 0.882 (0.840)
Epoch: [28][140/200]	Time 0.253 (0.329)	Data 0.000 (0.072)	Loss 0.670 (0.854)
Epoch: [28][160/200]	Time 0.253 (0.329)	Data 0.000 (0.072)	Loss 0.780 (0.872)
Epoch: [28][180/200]	Time 0.251 (0.329)	Data 0.000 (0.072)	Loss 0.905 (0.882)
Epoch: [28][200/200]	Time 0.256 (0.336)	Data 0.001 (0.078)	Loss 1.173 (0.887)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.135)	Data 0.000 (0.039)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.69760298728943
==> Statistics for epoch 29: 596 clusters
Epoch: [29][20/200]	Time 0.254 (0.358)	Data 0.001 (0.101)	Loss 0.817 (0.212)
Epoch: [29][40/200]	Time 0.253 (0.345)	Data 0.001 (0.087)	Loss 1.102 (0.581)
Epoch: [29][60/200]	Time 0.258 (0.339)	Data 0.001 (0.079)	Loss 0.612 (0.708)
Epoch: [29][80/200]	Time 0.254 (0.336)	Data 0.001 (0.077)	Loss 0.794 (0.783)
Epoch: [29][100/200]	Time 0.253 (0.334)	Data 0.001 (0.076)	Loss 1.401 (0.828)
Epoch: [29][120/200]	Time 0.252 (0.332)	Data 0.000 (0.074)	Loss 1.156 (0.857)
Epoch: [29][140/200]	Time 0.251 (0.333)	Data 0.000 (0.075)	Loss 0.819 (0.876)
Epoch: [29][160/200]	Time 0.253 (0.333)	Data 0.000 (0.075)	Loss 0.639 (0.884)
Epoch: [29][180/200]	Time 0.254 (0.332)	Data 0.000 (0.074)	Loss 1.095 (0.892)
Epoch: [29][200/200]	Time 0.255 (0.338)	Data 0.001 (0.080)	Loss 0.953 (0.900)
Extract Features: [50/76]	Time 0.139 (0.132)	Data 0.047 (0.034)	
Mean AP: 90.6%

 * Finished epoch  29  model mAP: 90.6%  best: 90.6% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.134)	Data 0.000 (0.037)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.225260257720947
==> Statistics for epoch 30: 596 clusters
Epoch: [30][20/200]	Time 0.254 (0.366)	Data 0.001 (0.109)	Loss 0.823 (0.198)
Epoch: [30][40/200]	Time 0.253 (0.347)	Data 0.001 (0.089)	Loss 0.891 (0.544)
Epoch: [30][60/200]	Time 0.253 (0.341)	Data 0.001 (0.083)	Loss 0.823 (0.664)
Epoch: [30][80/200]	Time 0.258 (0.336)	Data 0.001 (0.078)	Loss 1.129 (0.739)
Epoch: [30][100/200]	Time 0.254 (0.335)	Data 0.001 (0.076)	Loss 0.816 (0.789)
Epoch: [30][120/200]	Time 0.255 (0.334)	Data 0.001 (0.075)	Loss 1.094 (0.807)
Epoch: [30][140/200]	Time 0.255 (0.333)	Data 0.000 (0.075)	Loss 0.916 (0.827)
Epoch: [30][160/200]	Time 0.342 (0.332)	Data 0.000 (0.073)	Loss 1.047 (0.846)
Epoch: [30][180/200]	Time 0.253 (0.331)	Data 0.000 (0.073)	Loss 0.866 (0.859)
Epoch: [30][200/200]	Time 0.255 (0.337)	Data 0.001 (0.079)	Loss 0.807 (0.869)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.141)	Data 0.000 (0.043)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.39584469795227
==> Statistics for epoch 31: 596 clusters
Epoch: [31][20/200]	Time 0.253 (0.362)	Data 0.001 (0.104)	Loss 0.894 (0.195)
Epoch: [31][40/200]	Time 0.252 (0.343)	Data 0.001 (0.085)	Loss 1.049 (0.551)
Epoch: [31][60/200]	Time 0.253 (0.338)	Data 0.001 (0.080)	Loss 1.203 (0.687)
Epoch: [31][80/200]	Time 0.253 (0.335)	Data 0.001 (0.076)	Loss 0.964 (0.753)
Epoch: [31][100/200]	Time 0.252 (0.332)	Data 0.001 (0.074)	Loss 0.998 (0.795)
Epoch: [31][120/200]	Time 0.253 (0.331)	Data 0.000 (0.074)	Loss 0.918 (0.826)
Epoch: [31][140/200]	Time 0.254 (0.331)	Data 0.000 (0.074)	Loss 0.768 (0.846)
Epoch: [31][160/200]	Time 0.252 (0.331)	Data 0.000 (0.074)	Loss 0.928 (0.866)
Epoch: [31][180/200]	Time 0.251 (0.330)	Data 0.000 (0.073)	Loss 0.734 (0.868)
Epoch: [31][200/200]	Time 0.256 (0.336)	Data 0.000 (0.078)	Loss 1.062 (0.880)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.135)	Data 0.000 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.51813268661499
==> Statistics for epoch 32: 596 clusters
Epoch: [32][20/200]	Time 0.262 (0.368)	Data 0.001 (0.114)	Loss 0.761 (0.198)
Epoch: [32][40/200]	Time 0.254 (0.353)	Data 0.001 (0.095)	Loss 1.042 (0.536)
Epoch: [32][60/200]	Time 0.255 (0.344)	Data 0.001 (0.087)	Loss 0.930 (0.688)
Epoch: [32][80/200]	Time 0.269 (0.338)	Data 0.001 (0.081)	Loss 1.326 (0.763)
Epoch: [32][100/200]	Time 0.255 (0.335)	Data 0.001 (0.078)	Loss 1.058 (0.797)
Epoch: [32][120/200]	Time 0.253 (0.335)	Data 0.000 (0.077)	Loss 0.900 (0.824)
Epoch: [32][140/200]	Time 0.255 (0.332)	Data 0.000 (0.075)	Loss 0.690 (0.847)
Epoch: [32][160/200]	Time 0.252 (0.333)	Data 0.000 (0.075)	Loss 0.852 (0.851)
Epoch: [32][180/200]	Time 0.255 (0.333)	Data 0.000 (0.075)	Loss 1.137 (0.862)
Epoch: [32][200/200]	Time 0.256 (0.339)	Data 0.001 (0.081)	Loss 0.994 (0.869)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.140)	Data 0.000 (0.045)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.678762912750244
==> Statistics for epoch 33: 596 clusters
Epoch: [33][20/200]	Time 0.410 (0.375)	Data 0.001 (0.111)	Loss 0.793 (0.194)
Epoch: [33][40/200]	Time 0.255 (0.348)	Data 0.001 (0.089)	Loss 1.186 (0.576)
Epoch: [33][60/200]	Time 0.254 (0.342)	Data 0.001 (0.081)	Loss 0.877 (0.686)
Epoch: [33][80/200]	Time 0.255 (0.337)	Data 0.001 (0.078)	Loss 1.180 (0.750)
Epoch: [33][100/200]	Time 0.256 (0.335)	Data 0.001 (0.076)	Loss 1.002 (0.792)
Epoch: [33][120/200]	Time 0.254 (0.332)	Data 0.000 (0.074)	Loss 0.744 (0.816)
Epoch: [33][140/200]	Time 0.251 (0.332)	Data 0.000 (0.074)	Loss 0.782 (0.831)
Epoch: [33][160/200]	Time 0.253 (0.331)	Data 0.000 (0.073)	Loss 1.005 (0.855)
Epoch: [33][180/200]	Time 0.251 (0.330)	Data 0.000 (0.072)	Loss 1.077 (0.878)
Epoch: [33][200/200]	Time 0.259 (0.338)	Data 0.001 (0.080)	Loss 1.220 (0.883)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.134)	Data 0.000 (0.039)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.76665949821472
==> Statistics for epoch 34: 594 clusters
Epoch: [34][20/200]	Time 0.270 (0.361)	Data 0.001 (0.102)	Loss 0.691 (0.217)
Epoch: [34][40/200]	Time 0.253 (0.346)	Data 0.001 (0.084)	Loss 1.031 (0.586)
Epoch: [34][60/200]	Time 0.255 (0.340)	Data 0.001 (0.078)	Loss 1.048 (0.717)
Epoch: [34][80/200]	Time 0.255 (0.334)	Data 0.001 (0.075)	Loss 1.008 (0.781)
Epoch: [34][100/200]	Time 0.252 (0.334)	Data 0.001 (0.074)	Loss 0.700 (0.808)
Epoch: [34][120/200]	Time 0.254 (0.332)	Data 0.000 (0.073)	Loss 1.059 (0.830)
Epoch: [34][140/200]	Time 0.251 (0.331)	Data 0.000 (0.073)	Loss 0.712 (0.853)
Epoch: [34][160/200]	Time 0.252 (0.331)	Data 0.000 (0.073)	Loss 1.210 (0.868)
Epoch: [34][180/200]	Time 0.253 (0.330)	Data 0.000 (0.072)	Loss 0.641 (0.874)
Epoch: [34][200/200]	Time 0.254 (0.336)	Data 0.001 (0.078)	Loss 0.750 (0.873)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.165 (0.137)	Data 0.073 (0.041)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.75723433494568
==> Statistics for epoch 35: 595 clusters
Epoch: [35][20/200]	Time 0.255 (0.361)	Data 0.001 (0.104)	Loss 1.093 (0.194)
Epoch: [35][40/200]	Time 0.277 (0.349)	Data 0.001 (0.092)	Loss 0.927 (0.523)
Epoch: [35][60/200]	Time 0.258 (0.341)	Data 0.001 (0.082)	Loss 0.906 (0.653)
Epoch: [35][80/200]	Time 0.254 (0.338)	Data 0.001 (0.080)	Loss 0.881 (0.727)
Epoch: [35][100/200]	Time 0.256 (0.336)	Data 0.001 (0.078)	Loss 1.022 (0.773)
Epoch: [35][120/200]	Time 0.255 (0.334)	Data 0.000 (0.075)	Loss 1.054 (0.803)
Epoch: [35][140/200]	Time 0.254 (0.334)	Data 0.000 (0.075)	Loss 0.971 (0.826)
Epoch: [35][160/200]	Time 0.253 (0.333)	Data 0.000 (0.074)	Loss 0.850 (0.837)
Epoch: [35][180/200]	Time 0.253 (0.332)	Data 0.000 (0.074)	Loss 0.798 (0.851)
Epoch: [35][200/200]	Time 0.258 (0.340)	Data 0.001 (0.081)	Loss 1.168 (0.863)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.134)	Data 0.000 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.878342866897583
==> Statistics for epoch 36: 596 clusters
Epoch: [36][20/200]	Time 0.257 (0.361)	Data 0.001 (0.103)	Loss 0.864 (0.208)
Epoch: [36][40/200]	Time 0.257 (0.346)	Data 0.001 (0.086)	Loss 0.898 (0.537)
Epoch: [36][60/200]	Time 0.255 (0.340)	Data 0.001 (0.081)	Loss 0.552 (0.655)
Epoch: [36][80/200]	Time 0.254 (0.337)	Data 0.001 (0.078)	Loss 1.049 (0.737)
Epoch: [36][100/200]	Time 0.255 (0.335)	Data 0.001 (0.076)	Loss 1.250 (0.776)
Epoch: [36][120/200]	Time 0.256 (0.334)	Data 0.000 (0.075)	Loss 0.945 (0.809)
Epoch: [36][140/200]	Time 0.253 (0.333)	Data 0.000 (0.074)	Loss 1.069 (0.831)
Epoch: [36][160/200]	Time 0.251 (0.332)	Data 0.000 (0.073)	Loss 0.986 (0.844)
Epoch: [36][180/200]	Time 0.253 (0.331)	Data 0.000 (0.072)	Loss 1.099 (0.854)
Epoch: [36][200/200]	Time 0.267 (0.338)	Data 0.001 (0.079)	Loss 0.690 (0.862)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.132)	Data 0.000 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.809399843215942
==> Statistics for epoch 37: 596 clusters
Epoch: [37][20/200]	Time 0.253 (0.364)	Data 0.001 (0.106)	Loss 0.863 (0.198)
Epoch: [37][40/200]	Time 0.253 (0.345)	Data 0.001 (0.087)	Loss 1.477 (0.570)
Epoch: [37][60/200]	Time 0.254 (0.336)	Data 0.001 (0.079)	Loss 1.179 (0.697)
Epoch: [37][80/200]	Time 0.253 (0.335)	Data 0.001 (0.078)	Loss 0.803 (0.778)
Epoch: [37][100/200]	Time 0.252 (0.332)	Data 0.000 (0.075)	Loss 0.780 (0.811)
Epoch: [37][120/200]	Time 0.346 (0.330)	Data 0.000 (0.073)	Loss 0.797 (0.830)
Epoch: [37][140/200]	Time 0.251 (0.330)	Data 0.000 (0.073)	Loss 0.776 (0.839)
Epoch: [37][160/200]	Time 0.252 (0.329)	Data 0.000 (0.072)	Loss 0.928 (0.851)
Epoch: [37][180/200]	Time 0.252 (0.328)	Data 0.000 (0.071)	Loss 0.963 (0.861)
Epoch: [37][200/200]	Time 0.263 (0.337)	Data 0.001 (0.079)	Loss 1.037 (0.868)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.179 (0.137)	Data 0.000 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.983020544052124
==> Statistics for epoch 38: 596 clusters
Epoch: [38][20/200]	Time 0.253 (0.362)	Data 0.001 (0.106)	Loss 0.736 (0.197)
Epoch: [38][40/200]	Time 0.254 (0.342)	Data 0.001 (0.086)	Loss 0.804 (0.537)
Epoch: [38][60/200]	Time 0.253 (0.337)	Data 0.001 (0.079)	Loss 0.848 (0.658)
Epoch: [38][80/200]	Time 0.261 (0.334)	Data 0.001 (0.077)	Loss 0.806 (0.729)
Epoch: [38][100/200]	Time 0.258 (0.333)	Data 0.000 (0.075)	Loss 0.959 (0.765)
Epoch: [38][120/200]	Time 0.254 (0.331)	Data 0.000 (0.073)	Loss 1.055 (0.797)
Epoch: [38][140/200]	Time 0.254 (0.332)	Data 0.000 (0.073)	Loss 1.006 (0.820)
Epoch: [38][160/200]	Time 0.254 (0.332)	Data 0.000 (0.073)	Loss 0.741 (0.832)
Epoch: [38][180/200]	Time 0.252 (0.331)	Data 0.000 (0.073)	Loss 0.754 (0.847)
Epoch: [38][200/200]	Time 0.260 (0.339)	Data 0.000 (0.080)	Loss 1.019 (0.860)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.136)	Data 0.000 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.87479591369629
==> Statistics for epoch 39: 595 clusters
Epoch: [39][20/200]	Time 0.370 (0.363)	Data 0.001 (0.101)	Loss 0.810 (0.209)
Epoch: [39][40/200]	Time 0.255 (0.342)	Data 0.001 (0.082)	Loss 0.563 (0.572)
Epoch: [39][60/200]	Time 0.359 (0.336)	Data 0.001 (0.076)	Loss 0.729 (0.697)
Epoch: [39][80/200]	Time 0.254 (0.332)	Data 0.001 (0.074)	Loss 0.825 (0.752)
Epoch: [39][100/200]	Time 0.252 (0.330)	Data 0.001 (0.073)	Loss 1.112 (0.789)
Epoch: [39][120/200]	Time 0.253 (0.330)	Data 0.000 (0.072)	Loss 1.092 (0.815)
Epoch: [39][140/200]	Time 0.253 (0.329)	Data 0.000 (0.071)	Loss 1.049 (0.821)
Epoch: [39][160/200]	Time 0.253 (0.329)	Data 0.000 (0.071)	Loss 1.097 (0.832)
Epoch: [39][180/200]	Time 0.251 (0.330)	Data 0.000 (0.072)	Loss 1.075 (0.837)
Epoch: [39][200/200]	Time 0.264 (0.337)	Data 0.001 (0.079)	Loss 1.038 (0.850)
Extract Features: [50/76]	Time 0.101 (0.138)	Data 0.008 (0.041)	
Mean AP: 90.6%

 * Finished epoch  39  model mAP: 90.6%  best: 90.6% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.132)	Data 0.000 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.00593614578247
==> Statistics for epoch 40: 596 clusters
Epoch: [40][20/200]	Time 0.254 (0.378)	Data 0.001 (0.115)	Loss 0.868 (0.199)
Epoch: [40][40/200]	Time 0.254 (0.353)	Data 0.001 (0.091)	Loss 1.027 (0.549)
Epoch: [40][60/200]	Time 0.252 (0.343)	Data 0.001 (0.084)	Loss 0.978 (0.674)
Epoch: [40][80/200]	Time 0.254 (0.342)	Data 0.001 (0.082)	Loss 1.028 (0.732)
Epoch: [40][100/200]	Time 0.254 (0.341)	Data 0.001 (0.082)	Loss 1.288 (0.766)
Epoch: [40][120/200]	Time 0.256 (0.338)	Data 0.000 (0.080)	Loss 0.787 (0.787)
Epoch: [40][140/200]	Time 0.256 (0.337)	Data 0.000 (0.078)	Loss 0.955 (0.814)
Epoch: [40][160/200]	Time 0.254 (0.336)	Data 0.000 (0.078)	Loss 1.262 (0.832)
Epoch: [40][180/200]	Time 0.253 (0.337)	Data 0.000 (0.078)	Loss 0.655 (0.840)
Epoch: [40][200/200]	Time 0.253 (0.344)	Data 0.001 (0.085)	Loss 0.910 (0.852)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.091 (0.136)	Data 0.000 (0.039)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.089808702468872
==> Statistics for epoch 41: 596 clusters
Epoch: [41][20/200]	Time 0.255 (0.380)	Data 0.001 (0.123)	Loss 0.784 (0.201)
Epoch: [41][40/200]	Time 0.256 (0.360)	Data 0.001 (0.102)	Loss 0.844 (0.536)
Epoch: [41][60/200]	Time 0.257 (0.352)	Data 0.001 (0.094)	Loss 1.036 (0.660)
Epoch: [41][80/200]	Time 0.256 (0.350)	Data 0.001 (0.090)	Loss 0.938 (0.740)
Epoch: [41][100/200]	Time 0.255 (0.346)	Data 0.001 (0.087)	Loss 1.030 (0.788)
Epoch: [41][120/200]	Time 0.255 (0.345)	Data 0.000 (0.085)	Loss 1.048 (0.807)
Epoch: [41][140/200]	Time 0.255 (0.343)	Data 0.000 (0.084)	Loss 0.994 (0.827)
Epoch: [41][160/200]	Time 0.255 (0.342)	Data 0.000 (0.083)	Loss 0.695 (0.838)
Epoch: [41][180/200]	Time 0.253 (0.342)	Data 0.000 (0.083)	Loss 0.993 (0.844)
Epoch: [41][200/200]	Time 0.255 (0.349)	Data 0.001 (0.089)	Loss 0.730 (0.856)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.136)	Data 0.000 (0.038)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.91349506378174
==> Statistics for epoch 42: 596 clusters
Epoch: [42][20/200]	Time 0.255 (0.383)	Data 0.001 (0.117)	Loss 0.929 (0.207)
Epoch: [42][40/200]	Time 0.255 (0.357)	Data 0.001 (0.096)	Loss 1.084 (0.553)
Epoch: [42][60/200]	Time 0.258 (0.350)	Data 0.001 (0.090)	Loss 1.264 (0.683)
Epoch: [42][80/200]	Time 0.253 (0.346)	Data 0.001 (0.087)	Loss 0.792 (0.756)
Epoch: [42][100/200]	Time 0.254 (0.344)	Data 0.001 (0.084)	Loss 0.754 (0.788)
Epoch: [42][120/200]	Time 0.254 (0.341)	Data 0.000 (0.082)	Loss 0.839 (0.805)
Epoch: [42][140/200]	Time 0.254 (0.340)	Data 0.000 (0.081)	Loss 1.007 (0.823)
Epoch: [42][160/200]	Time 0.254 (0.339)	Data 0.000 (0.080)	Loss 0.841 (0.838)
Epoch: [42][180/200]	Time 0.253 (0.339)	Data 0.000 (0.080)	Loss 0.804 (0.852)
Epoch: [42][200/200]	Time 0.255 (0.346)	Data 0.001 (0.087)	Loss 1.007 (0.864)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.093 (0.139)	Data 0.000 (0.040)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.783504962921143
==> Statistics for epoch 43: 595 clusters
Epoch: [43][20/200]	Time 0.259 (0.377)	Data 0.001 (0.112)	Loss 1.094 (0.208)
Epoch: [43][40/200]	Time 0.253 (0.348)	Data 0.000 (0.088)	Loss 0.831 (0.510)
Epoch: [43][60/200]	Time 0.253 (0.341)	Data 0.001 (0.081)	Loss 0.787 (0.633)
Epoch: [43][80/200]	Time 0.254 (0.336)	Data 0.001 (0.078)	Loss 0.844 (0.716)
Epoch: [43][100/200]	Time 0.253 (0.334)	Data 0.001 (0.075)	Loss 0.877 (0.759)
Epoch: [43][120/200]	Time 0.253 (0.332)	Data 0.000 (0.074)	Loss 0.916 (0.781)
Epoch: [43][140/200]	Time 0.252 (0.331)	Data 0.000 (0.073)	Loss 1.029 (0.810)
Epoch: [43][160/200]	Time 0.252 (0.330)	Data 0.000 (0.072)	Loss 1.340 (0.828)
Epoch: [43][180/200]	Time 0.251 (0.329)	Data 0.000 (0.072)	Loss 0.994 (0.845)
Epoch: [43][200/200]	Time 0.254 (0.336)	Data 0.001 (0.078)	Loss 0.651 (0.851)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.131)	Data 0.000 (0.035)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.82331156730652
==> Statistics for epoch 44: 596 clusters
Epoch: [44][20/200]	Time 0.254 (0.355)	Data 0.001 (0.096)	Loss 0.542 (0.201)
Epoch: [44][40/200]	Time 0.256 (0.346)	Data 0.001 (0.084)	Loss 0.874 (0.542)
Epoch: [44][60/200]	Time 0.254 (0.339)	Data 0.001 (0.080)	Loss 0.784 (0.649)
Epoch: [44][80/200]	Time 0.253 (0.336)	Data 0.001 (0.078)	Loss 0.865 (0.733)
Epoch: [44][100/200]	Time 0.253 (0.334)	Data 0.001 (0.075)	Loss 0.946 (0.767)
Epoch: [44][120/200]	Time 0.253 (0.334)	Data 0.000 (0.075)	Loss 0.896 (0.810)
Epoch: [44][140/200]	Time 0.253 (0.333)	Data 0.000 (0.075)	Loss 0.692 (0.820)
Epoch: [44][160/200]	Time 0.252 (0.332)	Data 0.000 (0.074)	Loss 0.780 (0.833)
Epoch: [44][180/200]	Time 0.253 (0.331)	Data 0.000 (0.073)	Loss 1.116 (0.842)
Epoch: [44][200/200]	Time 0.258 (0.338)	Data 0.001 (0.080)	Loss 0.758 (0.848)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.135 (0.133)	Data 0.043 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.674880027770996
==> Statistics for epoch 45: 597 clusters
Epoch: [45][20/200]	Time 0.253 (0.359)	Data 0.001 (0.103)	Loss 0.758 (0.197)
Epoch: [45][40/200]	Time 0.255 (0.339)	Data 0.001 (0.084)	Loss 0.765 (0.540)
Epoch: [45][60/200]	Time 0.270 (0.337)	Data 0.001 (0.080)	Loss 1.020 (0.660)
Epoch: [45][80/200]	Time 0.253 (0.335)	Data 0.001 (0.078)	Loss 0.860 (0.712)
Epoch: [45][100/200]	Time 0.252 (0.333)	Data 0.001 (0.077)	Loss 1.251 (0.772)
Epoch: [45][120/200]	Time 0.254 (0.332)	Data 0.000 (0.076)	Loss 0.896 (0.793)
Epoch: [45][140/200]	Time 0.252 (0.331)	Data 0.000 (0.074)	Loss 0.680 (0.816)
Epoch: [45][160/200]	Time 0.252 (0.331)	Data 0.000 (0.074)	Loss 0.924 (0.829)
Epoch: [45][180/200]	Time 0.255 (0.331)	Data 0.000 (0.074)	Loss 1.170 (0.844)
Epoch: [45][200/200]	Time 0.256 (0.338)	Data 0.001 (0.080)	Loss 0.719 (0.858)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.140)	Data 0.000 (0.044)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.97493052482605
==> Statistics for epoch 46: 595 clusters
Epoch: [46][20/200]	Time 0.254 (0.356)	Data 0.001 (0.098)	Loss 0.754 (0.194)
Epoch: [46][40/200]	Time 0.259 (0.344)	Data 0.001 (0.084)	Loss 1.059 (0.540)
Epoch: [46][60/200]	Time 0.254 (0.336)	Data 0.001 (0.078)	Loss 0.780 (0.661)
Epoch: [46][80/200]	Time 0.259 (0.334)	Data 0.001 (0.075)	Loss 0.776 (0.716)
Epoch: [46][100/200]	Time 0.255 (0.331)	Data 0.001 (0.072)	Loss 1.047 (0.760)
Epoch: [46][120/200]	Time 0.254 (0.330)	Data 0.000 (0.071)	Loss 0.968 (0.791)
Epoch: [46][140/200]	Time 0.251 (0.329)	Data 0.000 (0.070)	Loss 0.842 (0.804)
Epoch: [46][160/200]	Time 0.254 (0.329)	Data 0.000 (0.070)	Loss 0.805 (0.826)
Epoch: [46][180/200]	Time 0.256 (0.330)	Data 0.000 (0.070)	Loss 1.087 (0.835)
Epoch: [46][200/200]	Time 0.256 (0.336)	Data 0.001 (0.077)	Loss 0.965 (0.848)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.092 (0.130)	Data 0.000 (0.032)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.550461053848267
==> Statistics for epoch 47: 596 clusters
Epoch: [47][20/200]	Time 0.253 (0.388)	Data 0.001 (0.124)	Loss 0.616 (0.169)
Epoch: [47][40/200]	Time 0.280 (0.358)	Data 0.001 (0.098)	Loss 1.220 (0.529)
Epoch: [47][60/200]	Time 0.254 (0.349)	Data 0.002 (0.090)	Loss 0.906 (0.652)
Epoch: [47][80/200]	Time 0.250 (0.345)	Data 0.001 (0.086)	Loss 1.018 (0.717)
Epoch: [47][100/200]	Time 0.253 (0.340)	Data 0.001 (0.082)	Loss 0.965 (0.758)
Epoch: [47][120/200]	Time 0.254 (0.338)	Data 0.000 (0.080)	Loss 1.104 (0.789)
Epoch: [47][140/200]	Time 0.254 (0.337)	Data 0.000 (0.079)	Loss 0.992 (0.815)
Epoch: [47][160/200]	Time 0.252 (0.335)	Data 0.000 (0.077)	Loss 0.855 (0.822)
Epoch: [47][180/200]	Time 0.250 (0.334)	Data 0.000 (0.076)	Loss 0.674 (0.833)
Epoch: [47][200/200]	Time 0.255 (0.340)	Data 0.001 (0.081)	Loss 0.995 (0.841)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.107 (0.140)	Data 0.015 (0.042)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.072064876556396
==> Statistics for epoch 48: 596 clusters
Epoch: [48][20/200]	Time 0.255 (0.365)	Data 0.001 (0.108)	Loss 0.830 (0.192)
Epoch: [48][40/200]	Time 0.255 (0.345)	Data 0.001 (0.089)	Loss 0.943 (0.520)
Epoch: [48][60/200]	Time 0.255 (0.343)	Data 0.001 (0.085)	Loss 1.077 (0.645)
Epoch: [48][80/200]	Time 0.253 (0.339)	Data 0.001 (0.082)	Loss 0.881 (0.712)
Epoch: [48][100/200]	Time 0.254 (0.337)	Data 0.001 (0.079)	Loss 0.870 (0.762)
Epoch: [48][120/200]	Time 0.256 (0.336)	Data 0.000 (0.078)	Loss 0.948 (0.793)
Epoch: [48][140/200]	Time 0.253 (0.334)	Data 0.000 (0.077)	Loss 1.025 (0.805)
Epoch: [48][160/200]	Time 0.251 (0.334)	Data 0.000 (0.075)	Loss 1.234 (0.820)
Epoch: [48][180/200]	Time 0.253 (0.332)	Data 0.000 (0.074)	Loss 1.138 (0.833)
Epoch: [48][200/200]	Time 0.257 (0.338)	Data 0.001 (0.080)	Loss 1.100 (0.842)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.166 (0.134)	Data 0.074 (0.036)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.767109155654907
==> Statistics for epoch 49: 596 clusters
Epoch: [49][20/200]	Time 0.254 (0.375)	Data 0.001 (0.112)	Loss 0.656 (0.181)
Epoch: [49][40/200]	Time 0.253 (0.348)	Data 0.001 (0.089)	Loss 0.998 (0.525)
Epoch: [49][60/200]	Time 0.254 (0.341)	Data 0.001 (0.082)	Loss 1.211 (0.684)
Epoch: [49][80/200]	Time 0.253 (0.337)	Data 0.001 (0.078)	Loss 0.968 (0.778)
Epoch: [49][100/200]	Time 0.252 (0.333)	Data 0.001 (0.075)	Loss 1.185 (0.815)
Epoch: [49][120/200]	Time 0.251 (0.333)	Data 0.000 (0.074)	Loss 0.982 (0.827)
Epoch: [49][140/200]	Time 0.251 (0.332)	Data 0.000 (0.073)	Loss 0.770 (0.839)
Epoch: [49][160/200]	Time 0.257 (0.330)	Data 0.000 (0.072)	Loss 0.962 (0.853)
Epoch: [49][180/200]	Time 0.253 (0.330)	Data 0.000 (0.071)	Loss 0.847 (0.861)
Epoch: [49][200/200]	Time 0.256 (0.337)	Data 0.001 (0.078)	Loss 0.823 (0.868)
Extract Features: [50/76]	Time 0.093 (0.133)	Data 0.000 (0.034)	
Mean AP: 90.6%

 * Finished epoch  49  model mAP: 90.6%  best: 90.6%

==> Test with the best model:
=> Loaded checkpoint 'log/msmt2market/resnet50_cion/model_best.pth.tar'
Extract Features: [50/76]	Time 0.093 (0.134)	Data 0.000 (0.038)	
Mean AP: 90.6%
CMC Scores:
  top-1          96.1%
  top-5          98.4%
  top-10         99.0%
Total running time:  1:23:56.797187
