==========
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='resnet101', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet101_Market1501/64bs_lr0.0004_ep120_warm20_seed0/resnet101_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/resnet101_cion', pooling_type='gem', use_hard=True)
==========
==> Load unlabeled dataset
=> MSMT17 loaded
Dataset statistics:
  ----------------------------------------
  subset   | # ids | # images | # cameras
  ----------------------------------------
  train    |  1041 |    32621 |        15
  query    |  3060 |    11659 |        15
  gallery  |  3060 |    82161 |        15
  ----------------------------------------
pooling_type: gem
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.238 (0.380)	Data 0.000 (0.022)	
Extract Features: [100/128]	Time 0.168 (0.281)	Data 0.000 (0.011)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.46762943267822
Clustering criterion: eps: 0.700
==> Statistics for epoch 0: 814 clusters
Epoch: [0][20/200]	Time 0.517 (0.997)	Data 0.000 (0.051)	Loss 3.196 (2.989)
Epoch: [0][40/200]	Time 0.523 (0.799)	Data 0.001 (0.066)	Loss 2.224 (2.843)
Epoch: [0][60/200]	Time 0.521 (0.734)	Data 0.001 (0.071)	Loss 1.815 (2.639)
Epoch: [0][80/200]	Time 0.525 (0.701)	Data 0.001 (0.071)	Loss 1.659 (2.459)
Epoch: [0][100/200]	Time 0.522 (0.665)	Data 0.000 (0.057)	Loss 1.859 (2.358)
Epoch: [0][120/200]	Time 0.522 (0.656)	Data 0.000 (0.061)	Loss 1.527 (2.261)
Epoch: [0][140/200]	Time 0.522 (0.648)	Data 0.001 (0.063)	Loss 1.897 (2.196)
Epoch: [0][160/200]	Time 0.520 (0.643)	Data 0.001 (0.065)	Loss 1.426 (2.135)
Epoch: [0][180/200]	Time 0.520 (0.638)	Data 0.001 (0.066)	Loss 1.840 (2.090)
Epoch: [0][200/200]	Time 0.522 (0.627)	Data 0.000 (0.060)	Loss 2.003 (2.042)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.213)	Data 0.000 (0.040)	
Extract Features: [100/128]	Time 0.172 (0.197)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.95147967338562
==> Statistics for epoch 1: 984 clusters
Epoch: [1][20/200]	Time 0.519 (0.577)	Data 0.001 (0.056)	Loss 0.328 (0.446)
Epoch: [1][40/200]	Time 0.522 (0.591)	Data 0.001 (0.066)	Loss 1.807 (0.770)
Epoch: [1][60/200]	Time 0.522 (0.568)	Data 0.000 (0.044)	Loss 1.806 (1.122)
Epoch: [1][80/200]	Time 0.525 (0.580)	Data 0.001 (0.055)	Loss 1.564 (1.316)
Epoch: [1][100/200]	Time 0.527 (0.587)	Data 0.001 (0.060)	Loss 1.335 (1.401)
Epoch: [1][120/200]	Time 0.523 (0.576)	Data 0.000 (0.050)	Loss 1.704 (1.477)
Epoch: [1][140/200]	Time 0.523 (0.583)	Data 0.001 (0.056)	Loss 1.914 (1.517)
Epoch: [1][160/200]	Time 0.520 (0.587)	Data 0.000 (0.059)	Loss 1.742 (1.543)
Epoch: [1][180/200]	Time 0.519 (0.580)	Data 0.000 (0.053)	Loss 2.242 (1.560)
Epoch: [1][200/200]	Time 0.523 (0.583)	Data 0.001 (0.056)	Loss 1.580 (1.567)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.215)	Data 0.000 (0.042)	
Extract Features: [100/128]	Time 0.178 (0.200)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.22932481765747
==> Statistics for epoch 2: 1032 clusters
Epoch: [2][20/200]	Time 0.520 (0.577)	Data 0.001 (0.049)	Loss 0.396 (0.450)
Epoch: [2][40/200]	Time 0.520 (0.591)	Data 0.001 (0.067)	Loss 2.332 (0.735)
Epoch: [2][60/200]	Time 0.521 (0.570)	Data 0.000 (0.045)	Loss 1.865 (1.110)
Epoch: [2][80/200]	Time 0.523 (0.579)	Data 0.001 (0.054)	Loss 1.807 (1.286)
Epoch: [2][100/200]	Time 0.528 (0.584)	Data 0.004 (0.059)	Loss 1.785 (1.412)
Epoch: [2][120/200]	Time 0.525 (0.574)	Data 0.001 (0.049)	Loss 1.880 (1.473)
Epoch: [2][140/200]	Time 0.661 (0.579)	Data 0.001 (0.053)	Loss 1.653 (1.541)
Epoch: [2][160/200]	Time 0.523 (0.572)	Data 0.000 (0.046)	Loss 1.634 (1.579)
Epoch: [2][180/200]	Time 0.524 (0.576)	Data 0.001 (0.050)	Loss 2.058 (1.599)
Epoch: [2][200/200]	Time 0.523 (0.579)	Data 0.002 (0.053)	Loss 2.109 (1.627)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.387 (0.215)	Data 0.218 (0.041)	
Extract Features: [100/128]	Time 0.168 (0.195)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.6497015953064
==> Statistics for epoch 3: 1038 clusters
Epoch: [3][20/200]	Time 0.526 (0.574)	Data 0.001 (0.047)	Loss 0.477 (0.411)
Epoch: [3][40/200]	Time 0.516 (0.586)	Data 0.001 (0.063)	Loss 2.017 (0.728)
Epoch: [3][60/200]	Time 0.519 (0.564)	Data 0.000 (0.042)	Loss 1.981 (1.092)
Epoch: [3][80/200]	Time 0.670 (0.575)	Data 0.001 (0.051)	Loss 2.219 (1.262)
Epoch: [3][100/200]	Time 0.525 (0.581)	Data 0.001 (0.057)	Loss 2.077 (1.364)
Epoch: [3][120/200]	Time 0.522 (0.572)	Data 0.001 (0.048)	Loss 1.985 (1.436)
Epoch: [3][140/200]	Time 0.522 (0.576)	Data 0.001 (0.052)	Loss 1.254 (1.479)
Epoch: [3][160/200]	Time 0.526 (0.570)	Data 0.000 (0.046)	Loss 1.707 (1.505)
Epoch: [3][180/200]	Time 0.523 (0.574)	Data 0.001 (0.049)	Loss 1.897 (1.526)
Epoch: [3][200/200]	Time 0.518 (0.577)	Data 0.001 (0.052)	Loss 1.390 (1.548)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.171 (0.214)	Data 0.000 (0.041)	
Extract Features: [100/128]	Time 0.177 (0.196)	Data 0.001 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.594646692276
==> Statistics for epoch 4: 984 clusters
Epoch: [4][20/200]	Time 0.518 (0.578)	Data 0.001 (0.056)	Loss 0.306 (0.363)
Epoch: [4][40/200]	Time 0.524 (0.592)	Data 0.001 (0.066)	Loss 1.697 (0.701)
Epoch: [4][60/200]	Time 0.518 (0.569)	Data 0.000 (0.045)	Loss 1.669 (1.028)
Epoch: [4][80/200]	Time 0.522 (0.579)	Data 0.001 (0.053)	Loss 1.925 (1.197)
Epoch: [4][100/200]	Time 0.521 (0.583)	Data 0.001 (0.057)	Loss 1.325 (1.292)
Epoch: [4][120/200]	Time 0.518 (0.574)	Data 0.000 (0.048)	Loss 2.204 (1.374)
Epoch: [4][140/200]	Time 0.521 (0.578)	Data 0.001 (0.052)	Loss 1.614 (1.406)
Epoch: [4][160/200]	Time 0.523 (0.582)	Data 0.001 (0.055)	Loss 1.480 (1.435)
Epoch: [4][180/200]	Time 0.523 (0.576)	Data 0.000 (0.049)	Loss 1.782 (1.470)
Epoch: [4][200/200]	Time 0.524 (0.580)	Data 0.001 (0.053)	Loss 1.622 (1.482)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.202 (0.211)	Data 0.026 (0.041)	
Extract Features: [100/128]	Time 0.170 (0.196)	Data 0.001 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.131452798843384
==> Statistics for epoch 5: 1013 clusters
Epoch: [5][20/200]	Time 0.520 (0.582)	Data 0.001 (0.055)	Loss 0.292 (0.361)
Epoch: [5][40/200]	Time 0.518 (0.592)	Data 0.001 (0.069)	Loss 1.861 (0.663)
Epoch: [5][60/200]	Time 0.519 (0.570)	Data 0.000 (0.046)	Loss 1.905 (1.014)
Epoch: [5][80/200]	Time 0.520 (0.581)	Data 0.001 (0.054)	Loss 1.522 (1.176)
Epoch: [5][100/200]	Time 0.522 (0.585)	Data 0.001 (0.060)	Loss 2.254 (1.295)
Epoch: [5][120/200]	Time 0.519 (0.574)	Data 0.000 (0.050)	Loss 1.869 (1.391)
Epoch: [5][140/200]	Time 0.524 (0.578)	Data 0.001 (0.054)	Loss 1.453 (1.436)
Epoch: [5][160/200]	Time 0.522 (0.582)	Data 0.001 (0.057)	Loss 1.621 (1.478)
Epoch: [5][180/200]	Time 0.521 (0.575)	Data 0.000 (0.051)	Loss 1.410 (1.508)
Epoch: [5][200/200]	Time 0.522 (0.578)	Data 0.002 (0.054)	Loss 1.389 (1.539)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.167 (0.218)	Data 0.000 (0.043)	
Extract Features: [100/128]	Time 0.168 (0.198)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.39237689971924
==> Statistics for epoch 6: 1024 clusters
Epoch: [6][20/200]	Time 0.517 (0.572)	Data 0.001 (0.047)	Loss 0.323 (0.367)
Epoch: [6][40/200]	Time 0.516 (0.586)	Data 0.000 (0.062)	Loss 1.555 (0.606)
Epoch: [6][60/200]	Time 0.518 (0.564)	Data 0.000 (0.042)	Loss 1.468 (0.947)
Epoch: [6][80/200]	Time 0.523 (0.574)	Data 0.001 (0.052)	Loss 1.199 (1.118)
Epoch: [6][100/200]	Time 0.530 (0.580)	Data 0.002 (0.058)	Loss 1.816 (1.210)
Epoch: [6][120/200]	Time 0.520 (0.570)	Data 0.001 (0.048)	Loss 2.087 (1.276)
Epoch: [6][140/200]	Time 0.523 (0.576)	Data 0.001 (0.054)	Loss 2.297 (1.316)
Epoch: [6][160/200]	Time 0.518 (0.569)	Data 0.000 (0.047)	Loss 1.725 (1.348)
Epoch: [6][180/200]	Time 0.521 (0.574)	Data 0.001 (0.051)	Loss 1.739 (1.371)
Epoch: [6][200/200]	Time 0.633 (0.578)	Data 0.001 (0.054)	Loss 1.970 (1.393)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.212)	Data 0.000 (0.040)	
Extract Features: [100/128]	Time 0.168 (0.197)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.840790033340454
==> Statistics for epoch 7: 1012 clusters
Epoch: [7][20/200]	Time 0.523 (0.572)	Data 0.002 (0.051)	Loss 0.223 (0.315)
Epoch: [7][40/200]	Time 0.517 (0.587)	Data 0.000 (0.065)	Loss 1.607 (0.634)
Epoch: [7][60/200]	Time 0.517 (0.567)	Data 0.000 (0.043)	Loss 1.368 (0.934)
Epoch: [7][80/200]	Time 0.522 (0.575)	Data 0.001 (0.052)	Loss 1.484 (1.095)
Epoch: [7][100/200]	Time 0.520 (0.581)	Data 0.001 (0.058)	Loss 2.253 (1.198)
Epoch: [7][120/200]	Time 0.520 (0.572)	Data 0.000 (0.048)	Loss 1.518 (1.270)
Epoch: [7][140/200]	Time 0.520 (0.578)	Data 0.001 (0.053)	Loss 1.573 (1.325)
Epoch: [7][160/200]	Time 0.522 (0.581)	Data 0.001 (0.056)	Loss 1.345 (1.353)
Epoch: [7][180/200]	Time 0.521 (0.575)	Data 0.000 (0.050)	Loss 1.651 (1.379)
Epoch: [7][200/200]	Time 0.520 (0.577)	Data 0.001 (0.053)	Loss 1.826 (1.399)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.167 (0.214)	Data 0.000 (0.043)	
Extract Features: [100/128]	Time 0.169 (0.196)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.04208660125732
==> Statistics for epoch 8: 1023 clusters
Epoch: [8][20/200]	Time 0.519 (0.574)	Data 0.001 (0.047)	Loss 0.279 (0.316)
Epoch: [8][40/200]	Time 0.636 (0.591)	Data 0.001 (0.064)	Loss 1.603 (0.592)
Epoch: [8][60/200]	Time 0.518 (0.567)	Data 0.000 (0.043)	Loss 1.476 (0.934)
Epoch: [8][80/200]	Time 0.519 (0.577)	Data 0.001 (0.052)	Loss 1.926 (1.116)
Epoch: [8][100/200]	Time 0.524 (0.582)	Data 0.001 (0.058)	Loss 2.173 (1.197)
Epoch: [8][120/200]	Time 0.518 (0.572)	Data 0.000 (0.048)	Loss 1.884 (1.265)
Epoch: [8][140/200]	Time 0.522 (0.575)	Data 0.001 (0.052)	Loss 1.319 (1.306)
Epoch: [8][160/200]	Time 0.522 (0.578)	Data 0.001 (0.055)	Loss 1.338 (1.348)
Epoch: [8][180/200]	Time 0.635 (0.573)	Data 0.000 (0.049)	Loss 1.712 (1.376)
Epoch: [8][200/200]	Time 0.521 (0.575)	Data 0.001 (0.051)	Loss 1.644 (1.396)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.167 (0.211)	Data 0.000 (0.035)	
Extract Features: [100/128]	Time 0.167 (0.194)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.46139931678772
==> Statistics for epoch 9: 1015 clusters
Epoch: [9][20/200]	Time 0.520 (0.572)	Data 0.000 (0.052)	Loss 0.330 (0.288)
Epoch: [9][40/200]	Time 0.524 (0.591)	Data 0.001 (0.068)	Loss 1.561 (0.607)
Epoch: [9][60/200]	Time 0.522 (0.569)	Data 0.000 (0.045)	Loss 1.343 (0.893)
Epoch: [9][80/200]	Time 0.521 (0.578)	Data 0.001 (0.055)	Loss 1.389 (1.055)
Epoch: [9][100/200]	Time 0.523 (0.584)	Data 0.001 (0.060)	Loss 1.841 (1.159)
Epoch: [9][120/200]	Time 0.521 (0.573)	Data 0.000 (0.050)	Loss 1.717 (1.227)
Epoch: [9][140/200]	Time 0.518 (0.578)	Data 0.001 (0.055)	Loss 1.557 (1.271)
Epoch: [9][160/200]	Time 0.523 (0.582)	Data 0.001 (0.058)	Loss 1.580 (1.306)
Epoch: [9][180/200]	Time 0.520 (0.576)	Data 0.000 (0.052)	Loss 1.626 (1.325)
Epoch: [9][200/200]	Time 0.523 (0.578)	Data 0.001 (0.054)	Loss 2.406 (1.351)
Extract Features: [50/367]	Time 0.176 (0.218)	Data 0.008 (0.047)	
Extract Features: [100/367]	Time 0.168 (0.201)	Data 0.000 (0.031)	
Extract Features: [150/367]	Time 0.185 (0.197)	Data 0.017 (0.024)	
Extract Features: [200/367]	Time 0.168 (0.193)	Data 0.000 (0.020)	
Extract Features: [250/367]	Time 0.184 (0.192)	Data 0.001 (0.016)	
Extract Features: [300/367]	Time 0.169 (0.192)	Data 0.000 (0.013)	
Extract Features: [350/367]	Time 0.167 (0.193)	Data 0.000 (0.011)	
Mean AP: 49.4%

 * Finished epoch   9  model mAP: 49.4%  best: 49.4% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.221)	Data 0.000 (0.050)	
Extract Features: [100/128]	Time 0.169 (0.201)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.1287202835083
==> Statistics for epoch 10: 1016 clusters
Epoch: [10][20/200]	Time 0.519 (0.602)	Data 0.001 (0.054)	Loss 0.313 (0.413)
Epoch: [10][40/200]	Time 0.518 (0.607)	Data 0.001 (0.070)	Loss 1.486 (0.650)
Epoch: [10][60/200]	Time 0.519 (0.580)	Data 0.000 (0.047)	Loss 1.940 (0.952)
Epoch: [10][80/200]	Time 0.518 (0.588)	Data 0.001 (0.058)	Loss 1.590 (1.073)
Epoch: [10][100/200]	Time 0.519 (0.592)	Data 0.001 (0.063)	Loss 1.660 (1.178)
Epoch: [10][120/200]	Time 0.521 (0.580)	Data 0.000 (0.053)	Loss 1.788 (1.222)
Epoch: [10][140/200]	Time 0.521 (0.585)	Data 0.001 (0.057)	Loss 1.471 (1.266)
Epoch: [10][160/200]	Time 0.521 (0.587)	Data 0.001 (0.061)	Loss 1.791 (1.292)
Epoch: [10][180/200]	Time 0.522 (0.581)	Data 0.000 (0.054)	Loss 1.643 (1.308)
Epoch: [10][200/200]	Time 0.522 (0.584)	Data 0.001 (0.057)	Loss 1.574 (1.328)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.213)	Data 0.000 (0.042)	
Extract Features: [100/128]	Time 0.179 (0.197)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.86189103126526
==> Statistics for epoch 11: 1001 clusters
Epoch: [11][20/200]	Time 0.518 (0.575)	Data 0.001 (0.051)	Loss 0.323 (0.271)
Epoch: [11][40/200]	Time 0.519 (0.588)	Data 0.001 (0.065)	Loss 1.347 (0.553)
Epoch: [11][60/200]	Time 0.521 (0.566)	Data 0.000 (0.044)	Loss 1.382 (0.859)
Epoch: [11][80/200]	Time 0.520 (0.576)	Data 0.001 (0.054)	Loss 1.319 (1.016)
Epoch: [11][100/200]	Time 0.521 (0.584)	Data 0.001 (0.061)	Loss 1.611 (1.105)
Epoch: [11][120/200]	Time 0.523 (0.574)	Data 0.000 (0.051)	Loss 1.298 (1.168)
Epoch: [11][140/200]	Time 0.527 (0.581)	Data 0.001 (0.056)	Loss 1.333 (1.211)
Epoch: [11][160/200]	Time 0.523 (0.585)	Data 0.001 (0.060)	Loss 1.548 (1.244)
Epoch: [11][180/200]	Time 0.523 (0.579)	Data 0.000 (0.053)	Loss 1.294 (1.269)
Epoch: [11][200/200]	Time 0.521 (0.582)	Data 0.000 (0.056)	Loss 1.560 (1.293)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.216)	Data 0.000 (0.040)	
Extract Features: [100/128]	Time 0.172 (0.197)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.35076832771301
==> Statistics for epoch 12: 1019 clusters
Epoch: [12][20/200]	Time 0.519 (0.582)	Data 0.001 (0.055)	Loss 0.413 (0.297)
Epoch: [12][40/200]	Time 0.523 (0.593)	Data 0.001 (0.069)	Loss 1.680 (0.545)
Epoch: [12][60/200]	Time 0.518 (0.571)	Data 0.000 (0.046)	Loss 1.452 (0.830)
Epoch: [12][80/200]	Time 0.522 (0.580)	Data 0.001 (0.056)	Loss 1.286 (0.975)
Epoch: [12][100/200]	Time 0.523 (0.588)	Data 0.001 (0.063)	Loss 1.457 (1.074)
Epoch: [12][120/200]	Time 0.521 (0.577)	Data 0.000 (0.052)	Loss 1.578 (1.130)
Epoch: [12][140/200]	Time 0.525 (0.581)	Data 0.001 (0.057)	Loss 1.559 (1.178)
Epoch: [12][160/200]	Time 0.524 (0.586)	Data 0.001 (0.061)	Loss 1.493 (1.202)
Epoch: [12][180/200]	Time 0.521 (0.579)	Data 0.000 (0.054)	Loss 1.341 (1.226)
Epoch: [12][200/200]	Time 0.524 (0.583)	Data 0.001 (0.058)	Loss 1.682 (1.250)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.216)	Data 0.000 (0.044)	
Extract Features: [100/128]	Time 0.172 (0.200)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.171996116638184
==> Statistics for epoch 13: 1004 clusters
Epoch: [13][20/200]	Time 0.654 (0.582)	Data 0.001 (0.055)	Loss 0.215 (0.304)
Epoch: [13][40/200]	Time 0.518 (0.596)	Data 0.001 (0.071)	Loss 1.176 (0.549)
Epoch: [13][60/200]	Time 0.522 (0.572)	Data 0.000 (0.048)	Loss 1.587 (0.842)
Epoch: [13][80/200]	Time 0.523 (0.584)	Data 0.001 (0.058)	Loss 1.524 (1.029)
Epoch: [13][100/200]	Time 0.527 (0.590)	Data 0.001 (0.064)	Loss 1.777 (1.109)
Epoch: [13][120/200]	Time 0.524 (0.579)	Data 0.000 (0.054)	Loss 1.379 (1.171)
Epoch: [13][140/200]	Time 0.521 (0.583)	Data 0.001 (0.058)	Loss 1.317 (1.206)
Epoch: [13][160/200]	Time 0.522 (0.587)	Data 0.001 (0.062)	Loss 1.364 (1.236)
Epoch: [13][180/200]	Time 0.521 (0.581)	Data 0.000 (0.055)	Loss 1.521 (1.261)
Epoch: [13][200/200]	Time 0.520 (0.584)	Data 0.001 (0.059)	Loss 1.466 (1.294)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.215)	Data 0.000 (0.041)	
Extract Features: [100/128]	Time 0.172 (0.198)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.53668427467346
==> Statistics for epoch 14: 1004 clusters
Epoch: [14][20/200]	Time 0.517 (0.578)	Data 0.001 (0.056)	Loss 0.191 (0.276)
Epoch: [14][40/200]	Time 0.521 (0.594)	Data 0.001 (0.073)	Loss 1.407 (0.527)
Epoch: [14][60/200]	Time 0.518 (0.571)	Data 0.000 (0.049)	Loss 1.402 (0.816)
Epoch: [14][80/200]	Time 0.520 (0.580)	Data 0.001 (0.057)	Loss 1.492 (0.964)
Epoch: [14][100/200]	Time 0.530 (0.587)	Data 0.002 (0.062)	Loss 1.078 (1.047)
Epoch: [14][120/200]	Time 0.523 (0.576)	Data 0.000 (0.052)	Loss 1.226 (1.105)
Epoch: [14][140/200]	Time 0.521 (0.582)	Data 0.001 (0.057)	Loss 1.669 (1.132)
Epoch: [14][160/200]	Time 0.521 (0.586)	Data 0.001 (0.061)	Loss 1.060 (1.174)
Epoch: [14][180/200]	Time 0.520 (0.579)	Data 0.000 (0.054)	Loss 1.518 (1.198)
Epoch: [14][200/200]	Time 0.518 (0.582)	Data 0.001 (0.057)	Loss 1.372 (1.220)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.210)	Data 0.000 (0.037)	
Extract Features: [100/128]	Time 0.177 (0.196)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.01202178001404
==> Statistics for epoch 15: 998 clusters
Epoch: [15][20/200]	Time 0.520 (0.582)	Data 0.001 (0.053)	Loss 0.222 (0.247)
Epoch: [15][40/200]	Time 0.621 (0.600)	Data 0.001 (0.072)	Loss 1.518 (0.502)
Epoch: [15][60/200]	Time 0.519 (0.574)	Data 0.000 (0.048)	Loss 1.431 (0.777)
Epoch: [15][80/200]	Time 0.531 (0.584)	Data 0.001 (0.058)	Loss 1.732 (0.911)
Epoch: [15][100/200]	Time 0.521 (0.590)	Data 0.001 (0.064)	Loss 1.051 (0.980)
Epoch: [15][120/200]	Time 0.520 (0.579)	Data 0.000 (0.054)	Loss 0.980 (1.046)
Epoch: [15][140/200]	Time 0.523 (0.583)	Data 0.001 (0.058)	Loss 1.194 (1.095)
Epoch: [15][160/200]	Time 0.526 (0.587)	Data 0.001 (0.062)	Loss 1.605 (1.127)
Epoch: [15][180/200]	Time 0.528 (0.580)	Data 0.000 (0.055)	Loss 1.279 (1.146)
Epoch: [15][200/200]	Time 0.523 (0.583)	Data 0.001 (0.058)	Loss 1.640 (1.172)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.171 (0.224)	Data 0.000 (0.047)	
Extract Features: [100/128]	Time 0.170 (0.202)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.45658588409424
==> Statistics for epoch 16: 1020 clusters
Epoch: [16][20/200]	Time 0.519 (0.575)	Data 0.001 (0.056)	Loss 0.201 (0.271)
Epoch: [16][40/200]	Time 0.520 (0.595)	Data 0.001 (0.074)	Loss 1.357 (0.504)
Epoch: [16][60/200]	Time 0.524 (0.570)	Data 0.000 (0.049)	Loss 1.119 (0.783)
Epoch: [16][80/200]	Time 0.525 (0.583)	Data 0.001 (0.060)	Loss 1.398 (0.919)
Epoch: [16][100/200]	Time 0.518 (0.588)	Data 0.001 (0.064)	Loss 2.012 (1.028)
Epoch: [16][120/200]	Time 0.518 (0.578)	Data 0.000 (0.054)	Loss 1.401 (1.078)
Epoch: [16][140/200]	Time 0.521 (0.583)	Data 0.001 (0.059)	Loss 1.193 (1.123)
Epoch: [16][160/200]	Time 0.534 (0.586)	Data 0.001 (0.062)	Loss 1.527 (1.158)
Epoch: [16][180/200]	Time 0.525 (0.580)	Data 0.000 (0.055)	Loss 1.390 (1.184)
Epoch: [16][200/200]	Time 0.524 (0.584)	Data 0.001 (0.059)	Loss 1.149 (1.207)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.214)	Data 0.000 (0.042)	
Extract Features: [100/128]	Time 0.169 (0.196)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.517682790756226
==> Statistics for epoch 17: 1040 clusters
Epoch: [17][20/200]	Time 0.520 (0.572)	Data 0.001 (0.052)	Loss 0.166 (0.248)
Epoch: [17][40/200]	Time 0.520 (0.589)	Data 0.001 (0.066)	Loss 1.431 (0.440)
Epoch: [17][60/200]	Time 0.520 (0.567)	Data 0.000 (0.044)	Loss 0.972 (0.718)
Epoch: [17][80/200]	Time 0.519 (0.579)	Data 0.001 (0.056)	Loss 1.456 (0.888)
Epoch: [17][100/200]	Time 0.529 (0.586)	Data 0.001 (0.062)	Loss 1.059 (0.990)
Epoch: [17][120/200]	Time 0.523 (0.575)	Data 0.001 (0.052)	Loss 1.227 (1.050)
Epoch: [17][140/200]	Time 0.520 (0.580)	Data 0.001 (0.056)	Loss 1.569 (1.098)
Epoch: [17][160/200]	Time 0.520 (0.572)	Data 0.000 (0.049)	Loss 1.326 (1.129)
Epoch: [17][180/200]	Time 0.524 (0.577)	Data 0.001 (0.053)	Loss 1.380 (1.153)
Epoch: [17][200/200]	Time 0.525 (0.580)	Data 0.001 (0.056)	Loss 1.498 (1.178)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.219)	Data 0.000 (0.045)	
Extract Features: [100/128]	Time 0.174 (0.201)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.7523729801178
==> Statistics for epoch 18: 1029 clusters
Epoch: [18][20/200]	Time 0.531 (0.583)	Data 0.001 (0.056)	Loss 0.278 (0.294)
Epoch: [18][40/200]	Time 0.518 (0.597)	Data 0.001 (0.073)	Loss 1.539 (0.488)
Epoch: [18][60/200]	Time 0.520 (0.575)	Data 0.000 (0.049)	Loss 1.327 (0.775)
Epoch: [18][80/200]	Time 0.521 (0.585)	Data 0.001 (0.059)	Loss 0.898 (0.918)
Epoch: [18][100/200]	Time 0.524 (0.589)	Data 0.001 (0.064)	Loss 1.224 (1.001)
Epoch: [18][120/200]	Time 0.523 (0.579)	Data 0.001 (0.053)	Loss 1.351 (1.056)
Epoch: [18][140/200]	Time 0.528 (0.583)	Data 0.001 (0.057)	Loss 1.262 (1.093)
Epoch: [18][160/200]	Time 0.520 (0.575)	Data 0.000 (0.050)	Loss 1.592 (1.124)
Epoch: [18][180/200]	Time 0.521 (0.580)	Data 0.001 (0.055)	Loss 1.196 (1.148)
Epoch: [18][200/200]	Time 0.523 (0.583)	Data 0.001 (0.058)	Loss 1.025 (1.164)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.223 (0.215)	Data 0.054 (0.042)	
Extract Features: [100/128]	Time 0.184 (0.197)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.18766760826111
==> Statistics for epoch 19: 1027 clusters
Epoch: [19][20/200]	Time 0.515 (0.579)	Data 0.001 (0.053)	Loss 0.322 (0.270)
Epoch: [19][40/200]	Time 0.522 (0.594)	Data 0.001 (0.068)	Loss 1.559 (0.458)
Epoch: [19][60/200]	Time 0.522 (0.572)	Data 0.000 (0.046)	Loss 1.437 (0.727)
Epoch: [19][80/200]	Time 0.522 (0.581)	Data 0.001 (0.056)	Loss 1.616 (0.901)
Epoch: [19][100/200]	Time 0.521 (0.588)	Data 0.001 (0.063)	Loss 1.222 (0.989)
Epoch: [19][120/200]	Time 0.521 (0.578)	Data 0.001 (0.052)	Loss 1.582 (1.050)
Epoch: [19][140/200]	Time 0.522 (0.584)	Data 0.001 (0.058)	Loss 1.721 (1.109)
Epoch: [19][160/200]	Time 0.522 (0.576)	Data 0.000 (0.051)	Loss 1.075 (1.135)
Epoch: [19][180/200]	Time 0.524 (0.581)	Data 0.001 (0.055)	Loss 0.960 (1.156)
Epoch: [19][200/200]	Time 0.521 (0.584)	Data 0.001 (0.058)	Loss 1.211 (1.176)
Extract Features: [50/367]	Time 0.170 (0.225)	Data 0.000 (0.051)	
Extract Features: [100/367]	Time 0.186 (0.203)	Data 0.001 (0.028)	
Extract Features: [150/367]	Time 0.173 (0.196)	Data 0.005 (0.020)	
Extract Features: [200/367]	Time 0.227 (0.193)	Data 0.050 (0.018)	
Extract Features: [250/367]	Time 0.219 (0.191)	Data 0.044 (0.016)	
Extract Features: [300/367]	Time 0.171 (0.189)	Data 0.000 (0.015)	
Extract Features: [350/367]	Time 0.168 (0.188)	Data 0.000 (0.014)	
Mean AP: 64.1%

 * Finished epoch  19  model mAP: 64.1%  best: 64.1% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.277 (0.219)	Data 0.109 (0.049)	
Extract Features: [100/128]	Time 0.170 (0.200)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.02153491973877
==> Statistics for epoch 20: 1017 clusters
Epoch: [20][20/200]	Time 0.519 (0.571)	Data 0.001 (0.051)	Loss 0.245 (0.230)
Epoch: [20][40/200]	Time 0.520 (0.591)	Data 0.001 (0.072)	Loss 1.702 (0.441)
Epoch: [20][60/200]	Time 0.518 (0.570)	Data 0.000 (0.048)	Loss 1.018 (0.721)
Epoch: [20][80/200]	Time 0.520 (0.580)	Data 0.001 (0.057)	Loss 1.627 (0.847)
Epoch: [20][100/200]	Time 0.522 (0.586)	Data 0.001 (0.062)	Loss 1.038 (0.926)
Epoch: [20][120/200]	Time 0.521 (0.577)	Data 0.000 (0.052)	Loss 1.231 (0.996)
Epoch: [20][140/200]	Time 0.522 (0.582)	Data 0.001 (0.057)	Loss 1.653 (1.033)
Epoch: [20][160/200]	Time 0.521 (0.586)	Data 0.001 (0.061)	Loss 1.330 (1.058)
Epoch: [20][180/200]	Time 0.518 (0.579)	Data 0.000 (0.055)	Loss 1.426 (1.086)
Epoch: [20][200/200]	Time 0.531 (0.582)	Data 0.001 (0.058)	Loss 1.057 (1.105)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.217)	Data 0.000 (0.045)	
Extract Features: [100/128]	Time 0.173 (0.201)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.751983404159546
==> Statistics for epoch 21: 1046 clusters
Epoch: [21][20/200]	Time 0.518 (0.570)	Data 0.001 (0.050)	Loss 0.199 (0.237)
Epoch: [21][40/200]	Time 0.529 (0.589)	Data 0.001 (0.068)	Loss 0.957 (0.430)
Epoch: [21][60/200]	Time 0.520 (0.568)	Data 0.000 (0.045)	Loss 1.503 (0.709)
Epoch: [21][80/200]	Time 0.524 (0.578)	Data 0.001 (0.055)	Loss 1.120 (0.840)
Epoch: [21][100/200]	Time 0.521 (0.585)	Data 0.001 (0.061)	Loss 1.215 (0.936)
Epoch: [21][120/200]	Time 0.524 (0.575)	Data 0.001 (0.051)	Loss 0.909 (0.993)
Epoch: [21][140/200]	Time 0.522 (0.582)	Data 0.001 (0.057)	Loss 1.396 (1.047)
Epoch: [21][160/200]	Time 0.519 (0.575)	Data 0.000 (0.050)	Loss 1.137 (1.074)
Epoch: [21][180/200]	Time 0.523 (0.580)	Data 0.001 (0.055)	Loss 1.211 (1.100)
Epoch: [21][200/200]	Time 0.521 (0.583)	Data 0.001 (0.058)	Loss 1.304 (1.131)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.165 (0.213)	Data 0.000 (0.041)	
Extract Features: [100/128]	Time 0.233 (0.196)	Data 0.069 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.24391222000122
==> Statistics for epoch 22: 1044 clusters
Epoch: [22][20/200]	Time 0.519 (0.573)	Data 0.001 (0.053)	Loss 0.178 (0.223)
Epoch: [22][40/200]	Time 0.519 (0.587)	Data 0.001 (0.067)	Loss 1.758 (0.423)
Epoch: [22][60/200]	Time 0.520 (0.567)	Data 0.000 (0.045)	Loss 1.052 (0.714)
Epoch: [22][80/200]	Time 0.521 (0.578)	Data 0.001 (0.054)	Loss 1.502 (0.861)
Epoch: [22][100/200]	Time 0.527 (0.585)	Data 0.003 (0.061)	Loss 2.108 (0.942)
Epoch: [22][120/200]	Time 0.525 (0.575)	Data 0.001 (0.051)	Loss 1.416 (0.997)
Epoch: [22][140/200]	Time 0.523 (0.581)	Data 0.001 (0.057)	Loss 1.046 (1.040)
Epoch: [22][160/200]	Time 0.522 (0.573)	Data 0.000 (0.050)	Loss 1.379 (1.065)
Epoch: [22][180/200]	Time 0.524 (0.577)	Data 0.001 (0.054)	Loss 1.519 (1.091)
Epoch: [22][200/200]	Time 0.523 (0.581)	Data 0.001 (0.057)	Loss 0.969 (1.111)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.298 (0.216)	Data 0.122 (0.044)	
Extract Features: [100/128]	Time 0.169 (0.197)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.36131191253662
==> Statistics for epoch 23: 1033 clusters
Epoch: [23][20/200]	Time 0.518 (0.578)	Data 0.001 (0.058)	Loss 0.234 (0.223)
Epoch: [23][40/200]	Time 0.520 (0.594)	Data 0.001 (0.072)	Loss 1.124 (0.422)
Epoch: [23][60/200]	Time 0.518 (0.571)	Data 0.000 (0.048)	Loss 1.216 (0.724)
Epoch: [23][80/200]	Time 0.521 (0.581)	Data 0.001 (0.057)	Loss 0.726 (0.855)
Epoch: [23][100/200]	Time 0.525 (0.587)	Data 0.001 (0.064)	Loss 0.836 (0.939)
Epoch: [23][120/200]	Time 0.521 (0.576)	Data 0.001 (0.053)	Loss 1.570 (1.006)
Epoch: [23][140/200]	Time 0.521 (0.582)	Data 0.001 (0.058)	Loss 1.001 (1.044)
Epoch: [23][160/200]	Time 0.521 (0.574)	Data 0.000 (0.050)	Loss 1.789 (1.073)
Epoch: [23][180/200]	Time 0.526 (0.578)	Data 0.001 (0.054)	Loss 1.113 (1.096)
Epoch: [23][200/200]	Time 0.522 (0.582)	Data 0.001 (0.057)	Loss 1.192 (1.112)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.216)	Data 0.000 (0.047)	
Extract Features: [100/128]	Time 0.170 (0.198)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.21994495391846
==> Statistics for epoch 24: 1039 clusters
Epoch: [24][20/200]	Time 0.518 (0.586)	Data 0.001 (0.057)	Loss 0.185 (0.214)
Epoch: [24][40/200]	Time 0.518 (0.599)	Data 0.001 (0.072)	Loss 1.330 (0.420)
Epoch: [24][60/200]	Time 0.519 (0.573)	Data 0.000 (0.048)	Loss 1.047 (0.691)
Epoch: [24][80/200]	Time 0.522 (0.583)	Data 0.001 (0.059)	Loss 1.289 (0.833)
Epoch: [24][100/200]	Time 0.521 (0.589)	Data 0.001 (0.065)	Loss 1.415 (0.915)
Epoch: [24][120/200]	Time 0.526 (0.579)	Data 0.001 (0.054)	Loss 0.943 (0.977)
Epoch: [24][140/200]	Time 0.656 (0.585)	Data 0.001 (0.059)	Loss 1.104 (1.026)
Epoch: [24][160/200]	Time 0.520 (0.577)	Data 0.000 (0.051)	Loss 1.170 (1.056)
Epoch: [24][180/200]	Time 0.526 (0.581)	Data 0.001 (0.055)	Loss 1.463 (1.087)
Epoch: [24][200/200]	Time 0.523 (0.586)	Data 0.001 (0.059)	Loss 1.400 (1.101)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.216)	Data 0.000 (0.042)	
Extract Features: [100/128]	Time 0.169 (0.198)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.43289613723755
==> Statistics for epoch 25: 1031 clusters
Epoch: [25][20/200]	Time 0.517 (0.573)	Data 0.001 (0.053)	Loss 0.160 (0.234)
Epoch: [25][40/200]	Time 0.518 (0.591)	Data 0.001 (0.068)	Loss 1.096 (0.423)
Epoch: [25][60/200]	Time 0.521 (0.568)	Data 0.000 (0.046)	Loss 1.392 (0.717)
Epoch: [25][80/200]	Time 0.523 (0.577)	Data 0.001 (0.055)	Loss 1.440 (0.863)
Epoch: [25][100/200]	Time 0.523 (0.584)	Data 0.001 (0.062)	Loss 1.434 (0.931)
Epoch: [25][120/200]	Time 0.525 (0.574)	Data 0.001 (0.052)	Loss 1.292 (0.989)
Epoch: [25][140/200]	Time 0.522 (0.580)	Data 0.001 (0.057)	Loss 1.631 (1.046)
Epoch: [25][160/200]	Time 0.523 (0.572)	Data 0.000 (0.050)	Loss 0.985 (1.076)
Epoch: [25][180/200]	Time 0.522 (0.577)	Data 0.001 (0.054)	Loss 1.399 (1.107)
Epoch: [25][200/200]	Time 0.521 (0.580)	Data 0.001 (0.056)	Loss 1.183 (1.126)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.166 (0.216)	Data 0.000 (0.045)	
Extract Features: [100/128]	Time 0.169 (0.199)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.65904760360718
==> Statistics for epoch 26: 1035 clusters
Epoch: [26][20/200]	Time 0.513 (0.580)	Data 0.001 (0.054)	Loss 0.179 (0.229)
Epoch: [26][40/200]	Time 0.519 (0.595)	Data 0.001 (0.069)	Loss 1.483 (0.454)
Epoch: [26][60/200]	Time 0.522 (0.570)	Data 0.000 (0.046)	Loss 1.444 (0.711)
Epoch: [26][80/200]	Time 0.520 (0.581)	Data 0.001 (0.056)	Loss 1.369 (0.870)
Epoch: [26][100/200]	Time 0.524 (0.586)	Data 0.001 (0.062)	Loss 1.463 (0.946)
Epoch: [26][120/200]	Time 0.522 (0.577)	Data 0.001 (0.052)	Loss 1.247 (1.006)
Epoch: [26][140/200]	Time 0.528 (0.581)	Data 0.002 (0.056)	Loss 1.101 (1.035)
Epoch: [26][160/200]	Time 0.523 (0.575)	Data 0.000 (0.049)	Loss 1.282 (1.063)
Epoch: [26][180/200]	Time 0.526 (0.580)	Data 0.001 (0.054)	Loss 0.987 (1.081)
Epoch: [26][200/200]	Time 0.522 (0.584)	Data 0.001 (0.057)	Loss 1.315 (1.103)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.215)	Data 0.000 (0.043)	
Extract Features: [100/128]	Time 0.167 (0.198)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.15665626525879
==> Statistics for epoch 27: 1046 clusters
Epoch: [27][20/200]	Time 0.519 (0.570)	Data 0.001 (0.049)	Loss 0.171 (0.217)
Epoch: [27][40/200]	Time 0.520 (0.593)	Data 0.001 (0.068)	Loss 1.069 (0.413)
Epoch: [27][60/200]	Time 0.520 (0.569)	Data 0.000 (0.046)	Loss 1.158 (0.675)
Epoch: [27][80/200]	Time 0.521 (0.580)	Data 0.001 (0.057)	Loss 1.218 (0.829)
Epoch: [27][100/200]	Time 0.659 (0.588)	Data 0.001 (0.063)	Loss 1.309 (0.915)
Epoch: [27][120/200]	Time 0.534 (0.578)	Data 0.001 (0.053)	Loss 1.130 (0.987)
Epoch: [27][140/200]	Time 0.518 (0.583)	Data 0.001 (0.058)	Loss 1.363 (1.028)
Epoch: [27][160/200]	Time 0.523 (0.577)	Data 0.000 (0.051)	Loss 1.200 (1.054)
Epoch: [27][180/200]	Time 0.520 (0.580)	Data 0.001 (0.055)	Loss 1.315 (1.075)
Epoch: [27][200/200]	Time 0.528 (0.585)	Data 0.001 (0.059)	Loss 1.275 (1.090)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.172 (0.213)	Data 0.000 (0.040)	
Extract Features: [100/128]	Time 0.169 (0.195)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.12247705459595
==> Statistics for epoch 28: 1042 clusters
Epoch: [28][20/200]	Time 0.519 (0.573)	Data 0.001 (0.052)	Loss 0.206 (0.218)
Epoch: [28][40/200]	Time 0.522 (0.589)	Data 0.001 (0.068)	Loss 1.215 (0.441)
Epoch: [28][60/200]	Time 0.519 (0.567)	Data 0.000 (0.046)	Loss 1.206 (0.691)
Epoch: [28][80/200]	Time 0.519 (0.579)	Data 0.001 (0.057)	Loss 1.209 (0.831)
Epoch: [28][100/200]	Time 0.525 (0.587)	Data 0.001 (0.064)	Loss 1.391 (0.917)
Epoch: [28][120/200]	Time 0.664 (0.578)	Data 0.001 (0.053)	Loss 1.241 (0.975)
Epoch: [28][140/200]	Time 0.522 (0.583)	Data 0.001 (0.058)	Loss 1.189 (1.020)
Epoch: [28][160/200]	Time 0.521 (0.575)	Data 0.000 (0.051)	Loss 1.349 (1.050)
Epoch: [28][180/200]	Time 0.520 (0.580)	Data 0.001 (0.055)	Loss 1.114 (1.073)
Epoch: [28][200/200]	Time 0.522 (0.584)	Data 0.001 (0.059)	Loss 1.370 (1.096)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.217)	Data 0.000 (0.041)	
Extract Features: [100/128]	Time 0.168 (0.199)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.23783874511719
==> Statistics for epoch 29: 1043 clusters
Epoch: [29][20/200]	Time 0.519 (0.584)	Data 0.001 (0.057)	Loss 0.187 (0.210)
Epoch: [29][40/200]	Time 0.532 (0.597)	Data 0.001 (0.072)	Loss 1.214 (0.393)
Epoch: [29][60/200]	Time 0.522 (0.572)	Data 0.000 (0.048)	Loss 1.490 (0.692)
Epoch: [29][80/200]	Time 0.522 (0.584)	Data 0.001 (0.058)	Loss 1.214 (0.823)
Epoch: [29][100/200]	Time 0.522 (0.589)	Data 0.001 (0.064)	Loss 1.231 (0.911)
Epoch: [29][120/200]	Time 0.520 (0.579)	Data 0.001 (0.054)	Loss 1.065 (0.982)
Epoch: [29][140/200]	Time 0.526 (0.585)	Data 0.001 (0.059)	Loss 1.067 (1.012)
Epoch: [29][160/200]	Time 0.524 (0.578)	Data 0.000 (0.052)	Loss 0.998 (1.059)
Epoch: [29][180/200]	Time 0.523 (0.582)	Data 0.000 (0.056)	Loss 1.046 (1.083)
Epoch: [29][200/200]	Time 0.525 (0.585)	Data 0.001 (0.059)	Loss 1.427 (1.102)
Extract Features: [50/367]	Time 0.166 (0.223)	Data 0.000 (0.047)	
Extract Features: [100/367]	Time 0.165 (0.201)	Data 0.000 (0.027)	
Extract Features: [150/367]	Time 0.224 (0.196)	Data 0.049 (0.021)	
Extract Features: [200/367]	Time 0.169 (0.192)	Data 0.000 (0.018)	
Extract Features: [250/367]	Time 0.245 (0.191)	Data 0.077 (0.017)	
Extract Features: [300/367]	Time 0.168 (0.189)	Data 0.000 (0.016)	
Extract Features: [350/367]	Time 0.170 (0.188)	Data 0.000 (0.015)	
Mean AP: 67.1%

 * Finished epoch  29  model mAP: 67.1%  best: 67.1% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.167 (0.214)	Data 0.000 (0.041)	
Extract Features: [100/128]	Time 0.174 (0.196)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.025585889816284
==> Statistics for epoch 30: 1038 clusters
Epoch: [30][20/200]	Time 0.518 (0.576)	Data 0.001 (0.055)	Loss 0.247 (0.226)
Epoch: [30][40/200]	Time 0.521 (0.594)	Data 0.001 (0.069)	Loss 1.141 (0.418)
Epoch: [30][60/200]	Time 0.522 (0.569)	Data 0.000 (0.046)	Loss 0.912 (0.667)
Epoch: [30][80/200]	Time 0.520 (0.580)	Data 0.001 (0.056)	Loss 1.445 (0.826)
Epoch: [30][100/200]	Time 0.523 (0.586)	Data 0.001 (0.062)	Loss 1.254 (0.923)
Epoch: [30][120/200]	Time 0.522 (0.576)	Data 0.001 (0.052)	Loss 1.176 (0.985)
Epoch: [30][140/200]	Time 0.521 (0.582)	Data 0.001 (0.057)	Loss 1.337 (1.027)
Epoch: [30][160/200]	Time 0.521 (0.575)	Data 0.000 (0.050)	Loss 0.978 (1.051)
Epoch: [30][180/200]	Time 0.523 (0.579)	Data 0.001 (0.055)	Loss 1.215 (1.078)
Epoch: [30][200/200]	Time 0.521 (0.583)	Data 0.001 (0.057)	Loss 1.444 (1.101)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.166 (0.216)	Data 0.000 (0.042)	
Extract Features: [100/128]	Time 0.168 (0.197)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.12562918663025
==> Statistics for epoch 31: 1056 clusters
Epoch: [31][20/200]	Time 0.519 (0.584)	Data 0.001 (0.055)	Loss 0.170 (0.217)
Epoch: [31][40/200]	Time 0.520 (0.598)	Data 0.001 (0.072)	Loss 1.305 (0.376)
Epoch: [31][60/200]	Time 0.521 (0.576)	Data 0.000 (0.048)	Loss 1.636 (0.667)
Epoch: [31][80/200]	Time 0.526 (0.584)	Data 0.001 (0.058)	Loss 1.048 (0.817)
Epoch: [31][100/200]	Time 2.379 (0.592)	Data 1.838 (0.065)	Loss 1.219 (0.907)
Epoch: [31][120/200]	Time 0.524 (0.582)	Data 0.001 (0.054)	Loss 1.134 (0.969)
Epoch: [31][140/200]	Time 0.523 (0.586)	Data 0.001 (0.059)	Loss 1.683 (1.018)
Epoch: [31][160/200]	Time 0.523 (0.579)	Data 0.000 (0.052)	Loss 1.345 (1.046)
Epoch: [31][180/200]	Time 0.524 (0.583)	Data 0.001 (0.056)	Loss 0.906 (1.074)
Epoch: [31][200/200]	Time 0.524 (0.586)	Data 0.001 (0.059)	Loss 1.304 (1.090)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.319 (0.219)	Data 0.153 (0.049)	
Extract Features: [100/128]	Time 0.170 (0.200)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.18740940093994
==> Statistics for epoch 32: 1054 clusters
Epoch: [32][20/200]	Time 0.518 (0.575)	Data 0.001 (0.057)	Loss 0.343 (0.230)
Epoch: [32][40/200]	Time 0.520 (0.591)	Data 0.001 (0.070)	Loss 0.919 (0.416)
Epoch: [32][60/200]	Time 0.524 (0.569)	Data 0.000 (0.047)	Loss 1.207 (0.688)
Epoch: [32][80/200]	Time 0.521 (0.578)	Data 0.001 (0.056)	Loss 1.025 (0.844)
Epoch: [32][100/200]	Time 0.524 (0.585)	Data 0.001 (0.062)	Loss 1.024 (0.921)
Epoch: [32][120/200]	Time 0.518 (0.575)	Data 0.001 (0.051)	Loss 0.963 (0.984)
Epoch: [32][140/200]	Time 0.519 (0.579)	Data 0.001 (0.056)	Loss 1.410 (1.014)
Epoch: [32][160/200]	Time 0.517 (0.572)	Data 0.000 (0.049)	Loss 0.745 (1.040)
Epoch: [32][180/200]	Time 0.522 (0.576)	Data 0.001 (0.053)	Loss 1.182 (1.064)
Epoch: [32][200/200]	Time 0.523 (0.580)	Data 0.001 (0.056)	Loss 1.356 (1.088)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.171 (0.220)	Data 0.001 (0.047)	
Extract Features: [100/128]	Time 0.182 (0.200)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.59739828109741
==> Statistics for epoch 33: 1046 clusters
Epoch: [33][20/200]	Time 0.674 (0.582)	Data 0.002 (0.053)	Loss 0.132 (0.207)
Epoch: [33][40/200]	Time 0.524 (0.597)	Data 0.001 (0.071)	Loss 1.378 (0.410)
Epoch: [33][60/200]	Time 0.521 (0.574)	Data 0.000 (0.047)	Loss 1.248 (0.681)
Epoch: [33][80/200]	Time 0.525 (0.584)	Data 0.001 (0.058)	Loss 1.779 (0.824)
Epoch: [33][100/200]	Time 0.522 (0.590)	Data 0.001 (0.064)	Loss 1.798 (0.894)
Epoch: [33][120/200]	Time 0.520 (0.578)	Data 0.002 (0.054)	Loss 1.105 (0.956)
Epoch: [33][140/200]	Time 0.523 (0.584)	Data 0.001 (0.059)	Loss 1.653 (0.996)
Epoch: [33][160/200]	Time 0.520 (0.576)	Data 0.000 (0.052)	Loss 1.077 (1.029)
Epoch: [33][180/200]	Time 0.521 (0.581)	Data 0.001 (0.056)	Loss 1.270 (1.061)
Epoch: [33][200/200]	Time 0.521 (0.584)	Data 0.001 (0.059)	Loss 1.585 (1.087)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.165 (0.211)	Data 0.000 (0.037)	
Extract Features: [100/128]	Time 0.176 (0.198)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.29113960266113
==> Statistics for epoch 34: 1051 clusters
Epoch: [34][20/200]	Time 0.519 (0.573)	Data 0.001 (0.049)	Loss 0.264 (0.234)
Epoch: [34][40/200]	Time 0.523 (0.594)	Data 0.002 (0.071)	Loss 1.063 (0.409)
Epoch: [34][60/200]	Time 0.520 (0.570)	Data 0.000 (0.048)	Loss 1.384 (0.691)
Epoch: [34][80/200]	Time 0.520 (0.584)	Data 0.001 (0.059)	Loss 1.095 (0.831)
Epoch: [34][100/200]	Time 0.524 (0.589)	Data 0.001 (0.064)	Loss 1.040 (0.927)
Epoch: [34][120/200]	Time 0.528 (0.579)	Data 0.001 (0.053)	Loss 1.211 (0.987)
Epoch: [34][140/200]	Time 0.531 (0.585)	Data 0.001 (0.058)	Loss 1.602 (1.027)
Epoch: [34][160/200]	Time 0.527 (0.577)	Data 0.000 (0.050)	Loss 1.172 (1.057)
Epoch: [34][180/200]	Time 0.524 (0.582)	Data 0.001 (0.055)	Loss 1.003 (1.092)
Epoch: [34][200/200]	Time 0.531 (0.585)	Data 0.001 (0.059)	Loss 1.474 (1.110)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.210 (0.213)	Data 0.042 (0.042)	
Extract Features: [100/128]	Time 0.180 (0.197)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.836796283721924
==> Statistics for epoch 35: 1056 clusters
Epoch: [35][20/200]	Time 0.523 (0.573)	Data 0.001 (0.051)	Loss 0.177 (0.218)
Epoch: [35][40/200]	Time 0.642 (0.597)	Data 0.001 (0.072)	Loss 1.360 (0.402)
Epoch: [35][60/200]	Time 0.522 (0.572)	Data 0.000 (0.048)	Loss 1.181 (0.676)
Epoch: [35][80/200]	Time 0.525 (0.582)	Data 0.001 (0.057)	Loss 1.220 (0.807)
Epoch: [35][100/200]	Time 2.228 (0.587)	Data 1.665 (0.062)	Loss 1.275 (0.893)
Epoch: [35][120/200]	Time 0.523 (0.577)	Data 0.001 (0.052)	Loss 1.620 (0.962)
Epoch: [35][140/200]	Time 0.521 (0.582)	Data 0.001 (0.057)	Loss 1.546 (1.009)
Epoch: [35][160/200]	Time 0.523 (0.575)	Data 0.000 (0.050)	Loss 1.288 (1.040)
Epoch: [35][180/200]	Time 0.666 (0.580)	Data 0.001 (0.055)	Loss 1.301 (1.070)
Epoch: [35][200/200]	Time 0.526 (0.583)	Data 0.001 (0.058)	Loss 1.202 (1.093)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.218)	Data 0.000 (0.044)	
Extract Features: [100/128]	Time 0.169 (0.198)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.957799673080444
==> Statistics for epoch 36: 1037 clusters
Epoch: [36][20/200]	Time 0.519 (0.579)	Data 0.001 (0.052)	Loss 0.230 (0.201)
Epoch: [36][40/200]	Time 0.518 (0.588)	Data 0.001 (0.065)	Loss 1.133 (0.381)
Epoch: [36][60/200]	Time 0.520 (0.567)	Data 0.000 (0.044)	Loss 1.331 (0.640)
Epoch: [36][80/200]	Time 0.521 (0.576)	Data 0.001 (0.053)	Loss 1.323 (0.795)
Epoch: [36][100/200]	Time 0.530 (0.584)	Data 0.001 (0.060)	Loss 1.614 (0.882)
Epoch: [36][120/200]	Time 0.526 (0.574)	Data 0.001 (0.050)	Loss 1.500 (0.943)
Epoch: [36][140/200]	Time 0.655 (0.580)	Data 0.001 (0.055)	Loss 1.596 (0.979)
Epoch: [36][160/200]	Time 0.521 (0.573)	Data 0.000 (0.048)	Loss 1.339 (1.009)
Epoch: [36][180/200]	Time 0.522 (0.578)	Data 0.001 (0.053)	Loss 1.279 (1.028)
Epoch: [36][200/200]	Time 0.522 (0.581)	Data 0.001 (0.057)	Loss 1.262 (1.042)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.200 (0.211)	Data 0.022 (0.041)	
Extract Features: [100/128]	Time 0.170 (0.196)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.23762845993042
==> Statistics for epoch 37: 1048 clusters
Epoch: [37][20/200]	Time 0.517 (0.582)	Data 0.001 (0.053)	Loss 0.135 (0.205)
Epoch: [37][40/200]	Time 0.520 (0.599)	Data 0.001 (0.070)	Loss 1.203 (0.408)
Epoch: [37][60/200]	Time 0.521 (0.573)	Data 0.000 (0.047)	Loss 1.485 (0.685)
Epoch: [37][80/200]	Time 0.526 (0.584)	Data 0.001 (0.058)	Loss 1.391 (0.826)
Epoch: [37][100/200]	Time 0.530 (0.590)	Data 0.001 (0.065)	Loss 1.102 (0.902)
Epoch: [37][120/200]	Time 0.521 (0.580)	Data 0.001 (0.054)	Loss 1.196 (0.963)
Epoch: [37][140/200]	Time 0.524 (0.584)	Data 0.001 (0.059)	Loss 1.033 (1.006)
Epoch: [37][160/200]	Time 0.523 (0.577)	Data 0.000 (0.051)	Loss 1.158 (1.040)
Epoch: [37][180/200]	Time 0.533 (0.582)	Data 0.001 (0.056)	Loss 1.414 (1.070)
Epoch: [37][200/200]	Time 0.522 (0.585)	Data 0.001 (0.059)	Loss 1.566 (1.083)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.171 (0.220)	Data 0.000 (0.044)	
Extract Features: [100/128]	Time 0.170 (0.201)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.556376934051514
==> Statistics for epoch 38: 1053 clusters
Epoch: [38][20/200]	Time 0.520 (0.577)	Data 0.001 (0.057)	Loss 0.248 (0.215)
Epoch: [38][40/200]	Time 0.522 (0.594)	Data 0.001 (0.071)	Loss 1.158 (0.401)
Epoch: [38][60/200]	Time 0.520 (0.571)	Data 0.000 (0.048)	Loss 1.553 (0.687)
Epoch: [38][80/200]	Time 0.522 (0.582)	Data 0.001 (0.058)	Loss 1.396 (0.827)
Epoch: [38][100/200]	Time 0.525 (0.589)	Data 0.001 (0.064)	Loss 1.523 (0.897)
Epoch: [38][120/200]	Time 0.522 (0.580)	Data 0.001 (0.053)	Loss 1.233 (0.957)
Epoch: [38][140/200]	Time 0.519 (0.585)	Data 0.001 (0.058)	Loss 1.346 (1.007)
Epoch: [38][160/200]	Time 0.522 (0.577)	Data 0.000 (0.051)	Loss 1.444 (1.035)
Epoch: [38][180/200]	Time 0.523 (0.582)	Data 0.001 (0.056)	Loss 0.975 (1.057)
Epoch: [38][200/200]	Time 0.524 (0.585)	Data 0.001 (0.059)	Loss 1.368 (1.077)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.215)	Data 0.000 (0.043)	
Extract Features: [100/128]	Time 0.167 (0.196)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.73034739494324
==> Statistics for epoch 39: 1061 clusters
Epoch: [39][20/200]	Time 0.522 (0.570)	Data 0.001 (0.051)	Loss 0.185 (0.214)
Epoch: [39][40/200]	Time 0.517 (0.588)	Data 0.001 (0.068)	Loss 1.414 (0.382)
Epoch: [39][60/200]	Time 0.518 (0.565)	Data 0.000 (0.046)	Loss 1.439 (0.668)
Epoch: [39][80/200]	Time 0.524 (0.577)	Data 0.001 (0.056)	Loss 1.182 (0.803)
Epoch: [39][100/200]	Time 2.297 (0.584)	Data 1.745 (0.063)	Loss 0.868 (0.891)
Epoch: [39][120/200]	Time 0.519 (0.575)	Data 0.001 (0.052)	Loss 1.590 (0.967)
Epoch: [39][140/200]	Time 0.521 (0.581)	Data 0.001 (0.058)	Loss 0.929 (0.995)
Epoch: [39][160/200]	Time 0.530 (0.574)	Data 0.000 (0.051)	Loss 1.288 (1.038)
Epoch: [39][180/200]	Time 0.519 (0.579)	Data 0.001 (0.055)	Loss 1.595 (1.070)
Epoch: [39][200/200]	Time 0.526 (0.583)	Data 0.001 (0.059)	Loss 1.091 (1.096)
Extract Features: [50/367]	Time 0.168 (0.219)	Data 0.000 (0.048)	
Extract Features: [100/367]	Time 0.166 (0.199)	Data 0.000 (0.028)	
Extract Features: [150/367]	Time 0.168 (0.194)	Data 0.000 (0.022)	
Extract Features: [200/367]	Time 0.217 (0.191)	Data 0.049 (0.020)	
Extract Features: [250/367]	Time 0.167 (0.189)	Data 0.000 (0.019)	
Extract Features: [300/367]	Time 0.238 (0.189)	Data 0.071 (0.018)	
Extract Features: [350/367]	Time 0.352 (0.188)	Data 0.000 (0.017)	
Mean AP: 67.5%

 * Finished epoch  39  model mAP: 67.5%  best: 67.5% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.164 (0.217)	Data 0.000 (0.050)	
Extract Features: [100/128]	Time 0.171 (0.201)	Data 0.000 (0.029)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.780961751937866
==> Statistics for epoch 40: 1048 clusters
Epoch: [40][20/200]	Time 0.517 (0.576)	Data 0.001 (0.056)	Loss 0.160 (0.217)
Epoch: [40][40/200]	Time 0.523 (0.595)	Data 0.001 (0.074)	Loss 1.210 (0.411)
Epoch: [40][60/200]	Time 0.522 (0.573)	Data 0.000 (0.049)	Loss 1.085 (0.701)
Epoch: [40][80/200]	Time 0.519 (0.582)	Data 0.001 (0.058)	Loss 1.057 (0.817)
Epoch: [40][100/200]	Time 0.530 (0.587)	Data 0.001 (0.062)	Loss 1.086 (0.892)
Epoch: [40][120/200]	Time 0.531 (0.577)	Data 0.001 (0.052)	Loss 1.169 (0.959)
Epoch: [40][140/200]	Time 0.523 (0.582)	Data 0.001 (0.058)	Loss 1.326 (0.998)
Epoch: [40][160/200]	Time 0.523 (0.575)	Data 0.000 (0.051)	Loss 1.164 (1.028)
Epoch: [40][180/200]	Time 0.525 (0.579)	Data 0.001 (0.054)	Loss 1.242 (1.056)
Epoch: [40][200/200]	Time 0.526 (0.583)	Data 0.001 (0.058)	Loss 1.602 (1.081)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.219)	Data 0.000 (0.046)	
Extract Features: [100/128]	Time 0.170 (0.198)	Data 0.001 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.352914333343506
==> Statistics for epoch 41: 1045 clusters
Epoch: [41][20/200]	Time 0.520 (0.571)	Data 0.001 (0.052)	Loss 0.154 (0.207)
Epoch: [41][40/200]	Time 0.518 (0.589)	Data 0.001 (0.067)	Loss 1.119 (0.392)
Epoch: [41][60/200]	Time 0.519 (0.568)	Data 0.000 (0.045)	Loss 0.915 (0.643)
Epoch: [41][80/200]	Time 0.523 (0.578)	Data 0.001 (0.055)	Loss 1.128 (0.794)
Epoch: [41][100/200]	Time 0.521 (0.585)	Data 0.001 (0.062)	Loss 1.084 (0.896)
Epoch: [41][120/200]	Time 0.526 (0.575)	Data 0.001 (0.051)	Loss 0.908 (0.948)
Epoch: [41][140/200]	Time 0.524 (0.580)	Data 0.001 (0.056)	Loss 1.266 (0.995)
Epoch: [41][160/200]	Time 0.519 (0.573)	Data 0.000 (0.049)	Loss 1.184 (1.031)
Epoch: [41][180/200]	Time 0.523 (0.579)	Data 0.001 (0.054)	Loss 1.358 (1.052)
Epoch: [41][200/200]	Time 0.667 (0.583)	Data 0.001 (0.058)	Loss 1.318 (1.080)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.407 (0.226)	Data 0.234 (0.052)	
Extract Features: [100/128]	Time 0.173 (0.206)	Data 0.000 (0.033)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.4617395401001
==> Statistics for epoch 42: 1060 clusters
Epoch: [42][20/200]	Time 0.520 (0.573)	Data 0.001 (0.049)	Loss 0.206 (0.212)
Epoch: [42][40/200]	Time 0.520 (0.590)	Data 0.001 (0.067)	Loss 1.143 (0.388)
Epoch: [42][60/200]	Time 0.525 (0.570)	Data 0.000 (0.045)	Loss 1.648 (0.670)
Epoch: [42][80/200]	Time 0.527 (0.583)	Data 0.001 (0.056)	Loss 1.210 (0.796)
Epoch: [42][100/200]	Time 2.372 (0.590)	Data 1.830 (0.063)	Loss 1.373 (0.891)
Epoch: [42][120/200]	Time 0.526 (0.580)	Data 0.001 (0.053)	Loss 1.440 (0.944)
Epoch: [42][140/200]	Time 0.525 (0.584)	Data 0.002 (0.057)	Loss 1.543 (0.988)
Epoch: [42][160/200]	Time 0.523 (0.577)	Data 0.000 (0.050)	Loss 1.260 (1.018)
Epoch: [42][180/200]	Time 0.524 (0.582)	Data 0.001 (0.055)	Loss 1.324 (1.044)
Epoch: [42][200/200]	Time 0.524 (0.585)	Data 0.001 (0.058)	Loss 1.433 (1.069)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.221)	Data 0.000 (0.047)	
Extract Features: [100/128]	Time 0.173 (0.199)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.10596823692322
==> Statistics for epoch 43: 1051 clusters
Epoch: [43][20/200]	Time 0.518 (0.577)	Data 0.001 (0.056)	Loss 0.180 (0.192)
Epoch: [43][40/200]	Time 0.519 (0.598)	Data 0.000 (0.073)	Loss 1.203 (0.374)
Epoch: [43][60/200]	Time 0.519 (0.572)	Data 0.000 (0.049)	Loss 0.980 (0.645)
Epoch: [43][80/200]	Time 0.521 (0.580)	Data 0.001 (0.058)	Loss 1.005 (0.764)
Epoch: [43][100/200]	Time 0.636 (0.587)	Data 0.001 (0.064)	Loss 1.581 (0.875)
Epoch: [43][120/200]	Time 0.523 (0.577)	Data 0.001 (0.053)	Loss 1.226 (0.933)
Epoch: [43][140/200]	Time 0.528 (0.582)	Data 0.001 (0.058)	Loss 1.427 (0.982)
Epoch: [43][160/200]	Time 0.524 (0.575)	Data 0.000 (0.051)	Loss 1.484 (1.014)
Epoch: [43][180/200]	Time 0.521 (0.580)	Data 0.001 (0.056)	Loss 1.209 (1.037)
Epoch: [43][200/200]	Time 0.523 (0.583)	Data 0.001 (0.059)	Loss 0.852 (1.053)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.169 (0.219)	Data 0.000 (0.046)	
Extract Features: [100/128]	Time 0.173 (0.200)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.51131534576416
==> Statistics for epoch 44: 1049 clusters
Epoch: [44][20/200]	Time 0.519 (0.571)	Data 0.001 (0.049)	Loss 0.072 (0.213)
Epoch: [44][40/200]	Time 0.521 (0.593)	Data 0.001 (0.069)	Loss 1.245 (0.377)
Epoch: [44][60/200]	Time 0.523 (0.570)	Data 0.000 (0.046)	Loss 1.249 (0.655)
Epoch: [44][80/200]	Time 0.522 (0.582)	Data 0.001 (0.057)	Loss 1.502 (0.800)
Epoch: [44][100/200]	Time 0.524 (0.587)	Data 0.002 (0.063)	Loss 1.331 (0.879)
Epoch: [44][120/200]	Time 0.522 (0.578)	Data 0.001 (0.053)	Loss 1.592 (0.935)
Epoch: [44][140/200]	Time 0.523 (0.582)	Data 0.001 (0.058)	Loss 1.356 (0.973)
Epoch: [44][160/200]	Time 0.522 (0.575)	Data 0.000 (0.051)	Loss 0.977 (1.004)
Epoch: [44][180/200]	Time 0.524 (0.580)	Data 0.001 (0.055)	Loss 1.355 (1.025)
Epoch: [44][200/200]	Time 0.523 (0.584)	Data 0.001 (0.059)	Loss 1.002 (1.043)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.291 (0.213)	Data 0.122 (0.039)	
Extract Features: [100/128]	Time 0.332 (0.199)	Data 0.001 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.485477685928345
==> Statistics for epoch 45: 1045 clusters
Epoch: [45][20/200]	Time 0.515 (0.578)	Data 0.001 (0.057)	Loss 0.143 (0.219)
Epoch: [45][40/200]	Time 0.527 (0.595)	Data 0.001 (0.074)	Loss 1.282 (0.411)
Epoch: [45][60/200]	Time 0.522 (0.573)	Data 0.000 (0.050)	Loss 1.531 (0.684)
Epoch: [45][80/200]	Time 0.522 (0.582)	Data 0.001 (0.059)	Loss 1.132 (0.816)
Epoch: [45][100/200]	Time 0.526 (0.590)	Data 0.002 (0.065)	Loss 1.263 (0.905)
Epoch: [45][120/200]	Time 0.524 (0.579)	Data 0.001 (0.055)	Loss 1.138 (0.958)
Epoch: [45][140/200]	Time 0.524 (0.584)	Data 0.001 (0.059)	Loss 0.957 (0.986)
Epoch: [45][160/200]	Time 0.522 (0.577)	Data 0.000 (0.052)	Loss 1.194 (1.026)
Epoch: [45][180/200]	Time 0.526 (0.581)	Data 0.002 (0.056)	Loss 1.075 (1.055)
Epoch: [45][200/200]	Time 0.523 (0.585)	Data 0.001 (0.059)	Loss 1.537 (1.077)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.220)	Data 0.000 (0.047)	
Extract Features: [100/128]	Time 0.173 (0.199)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.528759717941284
==> Statistics for epoch 46: 1049 clusters
Epoch: [46][20/200]	Time 0.519 (0.577)	Data 0.001 (0.056)	Loss 0.154 (0.213)
Epoch: [46][40/200]	Time 0.529 (0.596)	Data 0.001 (0.074)	Loss 1.101 (0.405)
Epoch: [46][60/200]	Time 0.520 (0.574)	Data 0.000 (0.049)	Loss 0.869 (0.666)
Epoch: [46][80/200]	Time 0.519 (0.585)	Data 0.000 (0.059)	Loss 1.274 (0.794)
Epoch: [46][100/200]	Time 0.522 (0.591)	Data 0.001 (0.065)	Loss 1.044 (0.872)
Epoch: [46][120/200]	Time 0.523 (0.579)	Data 0.001 (0.055)	Loss 1.145 (0.927)
Epoch: [46][140/200]	Time 0.524 (0.584)	Data 0.001 (0.059)	Loss 1.551 (0.979)
Epoch: [46][160/200]	Time 0.521 (0.576)	Data 0.000 (0.052)	Loss 1.085 (1.003)
Epoch: [46][180/200]	Time 0.518 (0.580)	Data 0.001 (0.056)	Loss 1.613 (1.036)
Epoch: [46][200/200]	Time 0.522 (0.583)	Data 0.001 (0.059)	Loss 0.989 (1.052)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.168 (0.215)	Data 0.000 (0.043)	
Extract Features: [100/128]	Time 0.306 (0.198)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.24325132369995
==> Statistics for epoch 47: 1053 clusters
Epoch: [47][20/200]	Time 0.524 (0.576)	Data 0.001 (0.058)	Loss 0.138 (0.203)
Epoch: [47][40/200]	Time 0.521 (0.593)	Data 0.000 (0.071)	Loss 0.969 (0.378)
Epoch: [47][60/200]	Time 0.517 (0.568)	Data 0.000 (0.047)	Loss 1.145 (0.654)
Epoch: [47][80/200]	Time 0.518 (0.577)	Data 0.001 (0.056)	Loss 1.018 (0.790)
Epoch: [47][100/200]	Time 0.525 (0.585)	Data 0.001 (0.063)	Loss 1.244 (0.893)
Epoch: [47][120/200]	Time 0.527 (0.575)	Data 0.001 (0.053)	Loss 1.148 (0.937)
Epoch: [47][140/200]	Time 0.520 (0.580)	Data 0.001 (0.058)	Loss 1.040 (0.985)
Epoch: [47][160/200]	Time 0.521 (0.573)	Data 0.000 (0.051)	Loss 1.446 (1.022)
Epoch: [47][180/200]	Time 0.520 (0.577)	Data 0.001 (0.055)	Loss 0.936 (1.044)
Epoch: [47][200/200]	Time 0.522 (0.582)	Data 0.001 (0.058)	Loss 1.165 (1.071)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.218)	Data 0.000 (0.046)	
Extract Features: [100/128]	Time 0.170 (0.199)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.08405590057373
==> Statistics for epoch 48: 1045 clusters
Epoch: [48][20/200]	Time 0.520 (0.572)	Data 0.001 (0.050)	Loss 0.308 (0.212)
Epoch: [48][40/200]	Time 0.523 (0.592)	Data 0.001 (0.069)	Loss 1.404 (0.400)
Epoch: [48][60/200]	Time 0.519 (0.572)	Data 0.000 (0.046)	Loss 1.239 (0.679)
Epoch: [48][80/200]	Time 0.520 (0.584)	Data 0.001 (0.058)	Loss 1.256 (0.813)
Epoch: [48][100/200]	Time 0.529 (0.591)	Data 0.001 (0.064)	Loss 1.202 (0.898)
Epoch: [48][120/200]	Time 0.521 (0.580)	Data 0.001 (0.054)	Loss 1.170 (0.948)
Epoch: [48][140/200]	Time 0.526 (0.585)	Data 0.001 (0.058)	Loss 1.083 (0.976)
Epoch: [48][160/200]	Time 0.524 (0.577)	Data 0.000 (0.051)	Loss 1.490 (1.007)
Epoch: [48][180/200]	Time 0.527 (0.582)	Data 0.001 (0.054)	Loss 0.969 (1.036)
Epoch: [48][200/200]	Time 0.534 (0.585)	Data 0.002 (0.058)	Loss 1.404 (1.059)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.170 (0.219)	Data 0.000 (0.042)	
Extract Features: [100/128]	Time 0.176 (0.198)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.97821593284607
==> Statistics for epoch 49: 1045 clusters
Epoch: [49][20/200]	Time 0.598 (0.579)	Data 0.001 (0.056)	Loss 0.189 (0.184)
Epoch: [49][40/200]	Time 0.516 (0.593)	Data 0.000 (0.072)	Loss 1.356 (0.375)
Epoch: [49][60/200]	Time 0.521 (0.568)	Data 0.000 (0.048)	Loss 1.254 (0.637)
Epoch: [49][80/200]	Time 0.521 (0.578)	Data 0.001 (0.059)	Loss 1.407 (0.778)
Epoch: [49][100/200]	Time 0.526 (0.586)	Data 0.001 (0.064)	Loss 1.630 (0.869)
Epoch: [49][120/200]	Time 0.524 (0.576)	Data 0.001 (0.053)	Loss 1.505 (0.932)
Epoch: [49][140/200]	Time 0.520 (0.581)	Data 0.001 (0.058)	Loss 1.201 (0.975)
Epoch: [49][160/200]	Time 0.584 (0.574)	Data 0.000 (0.051)	Loss 1.135 (1.013)
Epoch: [49][180/200]	Time 0.519 (0.578)	Data 0.001 (0.055)	Loss 1.028 (1.031)
Epoch: [49][200/200]	Time 0.520 (0.582)	Data 0.001 (0.058)	Loss 0.959 (1.058)
Extract Features: [50/367]	Time 0.168 (0.223)	Data 0.000 (0.051)	
Extract Features: [100/367]	Time 0.170 (0.205)	Data 0.000 (0.032)	
Extract Features: [150/367]	Time 0.175 (0.200)	Data 0.000 (0.027)	
Extract Features: [200/367]	Time 0.172 (0.197)	Data 0.000 (0.023)	
Extract Features: [250/367]	Time 0.173 (0.194)	Data 0.000 (0.021)	
Extract Features: [300/367]	Time 0.167 (0.193)	Data 0.000 (0.020)	
Extract Features: [350/367]	Time 0.169 (0.191)	Data 0.000 (0.019)	
Mean AP: 67.6%

 * Finished epoch  49  model mAP: 67.6%  best: 67.6% *

==> Test with the best model:
=> Loaded checkpoint 'log/market2msmt/resnet101_cion/model_best.pth.tar'
Extract Features: [50/367]	Time 0.167 (0.221)	Data 0.000 (0.048)	
Extract Features: [100/367]	Time 0.167 (0.202)	Data 0.000 (0.031)	
Extract Features: [150/367]	Time 0.165 (0.195)	Data 0.000 (0.024)	
Extract Features: [200/367]	Time 0.238 (0.194)	Data 0.070 (0.023)	
Extract Features: [250/367]	Time 0.178 (0.192)	Data 0.000 (0.020)	
Extract Features: [300/367]	Time 0.249 (0.190)	Data 0.080 (0.019)	
Extract Features: [350/367]	Time 0.171 (0.190)	Data 0.000 (0.018)	
Mean AP: 67.6%
CMC Scores:
  top-1          86.9%
  top-5          92.5%
  top-10         93.9%
Total running time:  3:26:48.226986
