==========
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='resnet_ibn152a', pretrained_path='/home/ma-user/work/Projects/ReIDNets_Checkpoints_TransReID/ResNet152_IBN.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/market/resnet152_ibn_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
checkpoint loaded!
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.203 (0.610)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 21.870579481124878
Clustering criterion: eps: 0.600
==> Statistics for epoch 0: 599 clusters
Epoch: [0][20/200]	Time 0.548 (1.103)	Data 0.001 (0.107)	Loss 4.476 (2.795)
Epoch: [0][40/200]	Time 0.549 (0.876)	Data 0.002 (0.085)	Loss 2.977 (3.253)
Epoch: [0][60/200]	Time 0.536 (0.796)	Data 0.001 (0.077)	Loss 2.450 (3.145)
Epoch: [0][80/200]	Time 0.567 (0.760)	Data 0.002 (0.075)	Loss 2.553 (3.024)
Epoch: [0][100/200]	Time 0.637 (0.741)	Data 0.001 (0.077)	Loss 2.321 (2.913)
Epoch: [0][120/200]	Time 0.557 (0.723)	Data 0.000 (0.074)	Loss 2.826 (2.835)
Epoch: [0][140/200]	Time 0.554 (0.713)	Data 0.000 (0.073)	Loss 2.713 (2.772)
Epoch: [0][160/200]	Time 0.534 (0.704)	Data 0.000 (0.071)	Loss 2.416 (2.697)
Epoch: [0][180/200]	Time 0.569 (0.697)	Data 0.000 (0.070)	Loss 1.711 (2.639)
Epoch: [0][200/200]	Time 0.681 (0.700)	Data 0.001 (0.076)	Loss 2.348 (2.597)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.191 (0.251)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.52091407775879
==> Statistics for epoch 1: 611 clusters
Epoch: [1][20/200]	Time 2.099 (0.694)	Data 1.479 (0.114)	Loss 2.171 (0.547)
Epoch: [1][40/200]	Time 0.730 (0.679)	Data 0.001 (0.095)	Loss 1.834 (1.259)
Epoch: [1][60/200]	Time 0.569 (0.670)	Data 0.001 (0.087)	Loss 1.940 (1.513)
Epoch: [1][80/200]	Time 0.544 (0.665)	Data 0.001 (0.082)	Loss 1.920 (1.609)
Epoch: [1][100/200]	Time 0.569 (0.665)	Data 0.001 (0.082)	Loss 2.185 (1.672)
Epoch: [1][120/200]	Time 0.578 (0.664)	Data 0.001 (0.081)	Loss 1.850 (1.695)
Epoch: [1][140/200]	Time 0.559 (0.663)	Data 0.001 (0.080)	Loss 2.040 (1.707)
Epoch: [1][160/200]	Time 0.565 (0.663)	Data 0.001 (0.079)	Loss 1.412 (1.696)
Epoch: [1][180/200]	Time 0.556 (0.662)	Data 0.002 (0.078)	Loss 2.070 (1.702)
Epoch: [1][200/200]	Time 0.563 (0.662)	Data 0.002 (0.078)	Loss 1.425 (1.697)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.201 (0.248)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.704001665115356
==> Statistics for epoch 2: 586 clusters
Epoch: [2][20/200]	Time 0.569 (0.706)	Data 0.001 (0.123)	Loss 1.646 (0.491)
Epoch: [2][40/200]	Time 0.681 (0.686)	Data 0.002 (0.101)	Loss 1.749 (1.056)
Epoch: [2][60/200]	Time 0.544 (0.675)	Data 0.001 (0.094)	Loss 1.289 (1.228)
Epoch: [2][80/200]	Time 0.648 (0.671)	Data 0.002 (0.089)	Loss 1.753 (1.331)
Epoch: [2][100/200]	Time 0.564 (0.668)	Data 0.001 (0.087)	Loss 1.612 (1.374)
Epoch: [2][120/200]	Time 0.565 (0.664)	Data 0.000 (0.083)	Loss 2.069 (1.400)
Epoch: [2][140/200]	Time 0.562 (0.664)	Data 0.000 (0.082)	Loss 1.740 (1.416)
Epoch: [2][160/200]	Time 0.565 (0.665)	Data 0.000 (0.083)	Loss 1.616 (1.428)
Epoch: [2][180/200]	Time 0.629 (0.664)	Data 0.000 (0.082)	Loss 0.910 (1.427)
Epoch: [2][200/200]	Time 0.585 (0.669)	Data 0.001 (0.087)	Loss 1.251 (1.422)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.197 (0.248)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.810527324676514
==> Statistics for epoch 3: 587 clusters
Epoch: [3][20/200]	Time 0.670 (0.695)	Data 0.001 (0.116)	Loss 1.367 (0.433)
Epoch: [3][40/200]	Time 0.563 (0.679)	Data 0.002 (0.102)	Loss 2.134 (0.956)
Epoch: [3][60/200]	Time 0.557 (0.672)	Data 0.002 (0.093)	Loss 1.258 (1.092)
Epoch: [3][80/200]	Time 0.557 (0.667)	Data 0.001 (0.088)	Loss 1.766 (1.176)
Epoch: [3][100/200]	Time 0.557 (0.663)	Data 0.001 (0.084)	Loss 1.327 (1.220)
Epoch: [3][120/200]	Time 0.556 (0.662)	Data 0.000 (0.082)	Loss 1.503 (1.252)
Epoch: [3][140/200]	Time 0.573 (0.660)	Data 0.000 (0.080)	Loss 1.619 (1.274)
Epoch: [3][160/200]	Time 0.676 (0.661)	Data 0.000 (0.079)	Loss 1.163 (1.277)
Epoch: [3][180/200]	Time 0.540 (0.660)	Data 0.000 (0.077)	Loss 1.303 (1.281)
Epoch: [3][200/200]	Time 0.704 (0.669)	Data 0.001 (0.085)	Loss 1.315 (1.287)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.201 (0.243)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.672794818878174
==> Statistics for epoch 4: 594 clusters
Epoch: [4][20/200]	Time 0.554 (0.691)	Data 0.001 (0.107)	Loss 0.990 (0.328)
Epoch: [4][40/200]	Time 0.562 (0.670)	Data 0.001 (0.088)	Loss 0.846 (0.807)
Epoch: [4][60/200]	Time 0.542 (0.664)	Data 0.001 (0.082)	Loss 1.316 (0.958)
Epoch: [4][80/200]	Time 0.566 (0.661)	Data 0.001 (0.079)	Loss 1.269 (1.044)
Epoch: [4][100/200]	Time 0.556 (0.659)	Data 0.002 (0.076)	Loss 1.235 (1.075)
Epoch: [4][120/200]	Time 0.570 (0.661)	Data 0.000 (0.078)	Loss 0.850 (1.100)
Epoch: [4][140/200]	Time 0.565 (0.659)	Data 0.000 (0.076)	Loss 1.718 (1.131)
Epoch: [4][160/200]	Time 0.560 (0.657)	Data 0.000 (0.075)	Loss 1.548 (1.149)
Epoch: [4][180/200]	Time 0.635 (0.655)	Data 0.000 (0.074)	Loss 1.231 (1.158)
Epoch: [4][200/200]	Time 0.541 (0.660)	Data 0.001 (0.080)	Loss 1.294 (1.160)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.301 (0.234)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.026527881622314
==> Statistics for epoch 5: 595 clusters
Epoch: [5][20/200]	Time 0.544 (0.707)	Data 0.002 (0.113)	Loss 1.120 (0.337)
Epoch: [5][40/200]	Time 0.678 (0.680)	Data 0.002 (0.089)	Loss 1.232 (0.776)
Epoch: [5][60/200]	Time 0.569 (0.675)	Data 0.006 (0.085)	Loss 1.168 (0.900)
Epoch: [5][80/200]	Time 0.629 (0.669)	Data 0.001 (0.080)	Loss 0.958 (0.960)
Epoch: [5][100/200]	Time 0.551 (0.665)	Data 0.001 (0.078)	Loss 0.901 (1.003)
Epoch: [5][120/200]	Time 0.569 (0.662)	Data 0.000 (0.076)	Loss 1.232 (1.034)
Epoch: [5][140/200]	Time 0.564 (0.662)	Data 0.000 (0.075)	Loss 0.888 (1.049)
Epoch: [5][160/200]	Time 0.563 (0.660)	Data 0.000 (0.074)	Loss 1.018 (1.058)
Epoch: [5][180/200]	Time 0.642 (0.658)	Data 0.000 (0.073)	Loss 1.146 (1.064)
Epoch: [5][200/200]	Time 0.567 (0.664)	Data 0.001 (0.079)	Loss 0.854 (1.073)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.309 (0.242)	Data 0.000 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.61916756629944
==> Statistics for epoch 6: 596 clusters
Epoch: [6][20/200]	Time 0.565 (0.686)	Data 0.001 (0.096)	Loss 1.108 (0.273)
Epoch: [6][40/200]	Time 0.662 (0.673)	Data 0.001 (0.083)	Loss 1.124 (0.651)
Epoch: [6][60/200]	Time 0.549 (0.666)	Data 0.002 (0.080)	Loss 0.783 (0.794)
Epoch: [6][80/200]	Time 0.663 (0.665)	Data 0.001 (0.078)	Loss 1.051 (0.872)
Epoch: [6][100/200]	Time 0.565 (0.663)	Data 0.001 (0.078)	Loss 1.281 (0.925)
Epoch: [6][120/200]	Time 0.561 (0.664)	Data 0.000 (0.078)	Loss 1.068 (0.954)
Epoch: [6][140/200]	Time 0.559 (0.662)	Data 0.000 (0.076)	Loss 0.885 (0.965)
Epoch: [6][160/200]	Time 0.562 (0.660)	Data 0.000 (0.075)	Loss 0.913 (0.982)
Epoch: [6][180/200]	Time 0.562 (0.660)	Data 0.000 (0.076)	Loss 1.223 (0.991)
Epoch: [6][200/200]	Time 0.565 (0.668)	Data 0.001 (0.084)	Loss 0.924 (0.986)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.199 (0.244)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.664329767227173
==> Statistics for epoch 7: 603 clusters
Epoch: [7][20/200]	Time 0.656 (0.692)	Data 0.001 (0.096)	Loss 1.071 (0.297)
Epoch: [7][40/200]	Time 0.569 (0.671)	Data 0.002 (0.083)	Loss 0.722 (0.659)
Epoch: [7][60/200]	Time 0.629 (0.667)	Data 0.001 (0.078)	Loss 0.818 (0.781)
Epoch: [7][80/200]	Time 0.553 (0.663)	Data 0.001 (0.075)	Loss 1.283 (0.843)
Epoch: [7][100/200]	Time 0.639 (0.658)	Data 0.001 (0.073)	Loss 1.135 (0.884)
Epoch: [7][120/200]	Time 0.569 (0.659)	Data 0.000 (0.073)	Loss 1.200 (0.923)
Epoch: [7][140/200]	Time 0.560 (0.656)	Data 0.000 (0.072)	Loss 1.271 (0.934)
Epoch: [7][160/200]	Time 0.551 (0.655)	Data 0.000 (0.072)	Loss 0.993 (0.935)
Epoch: [7][180/200]	Time 0.551 (0.653)	Data 0.000 (0.071)	Loss 1.095 (0.945)
Epoch: [7][200/200]	Time 0.557 (0.661)	Data 0.001 (0.077)	Loss 1.178 (0.953)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.202 (0.242)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.6989643573761
==> Statistics for epoch 8: 605 clusters
Epoch: [8][20/200]	Time 0.570 (0.704)	Data 0.001 (0.120)	Loss 1.008 (0.280)
Epoch: [8][40/200]	Time 0.687 (0.685)	Data 0.001 (0.098)	Loss 0.642 (0.606)
Epoch: [8][60/200]	Time 0.562 (0.677)	Data 0.001 (0.091)	Loss 0.991 (0.729)
Epoch: [8][80/200]	Time 0.565 (0.673)	Data 0.001 (0.087)	Loss 0.979 (0.788)
Epoch: [8][100/200]	Time 0.546 (0.672)	Data 0.001 (0.086)	Loss 1.153 (0.834)
Epoch: [8][120/200]	Time 0.561 (0.670)	Data 0.000 (0.085)	Loss 0.875 (0.854)
Epoch: [8][140/200]	Time 0.565 (0.669)	Data 0.000 (0.084)	Loss 1.443 (0.870)
Epoch: [8][160/200]	Time 0.566 (0.667)	Data 0.000 (0.083)	Loss 1.092 (0.889)
Epoch: [8][180/200]	Time 0.662 (0.667)	Data 0.000 (0.082)	Loss 1.348 (0.896)
Epoch: [8][200/200]	Time 0.546 (0.674)	Data 0.001 (0.089)	Loss 1.055 (0.904)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.277 (0.243)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.421695709228516
==> Statistics for epoch 9: 608 clusters
Epoch: [9][20/200]	Time 2.263 (0.697)	Data 1.505 (0.109)	Loss 1.062 (0.223)
Epoch: [9][40/200]	Time 0.552 (0.677)	Data 0.001 (0.092)	Loss 1.166 (0.588)
Epoch: [9][60/200]	Time 0.687 (0.675)	Data 0.001 (0.088)	Loss 1.054 (0.698)
Epoch: [9][80/200]	Time 0.565 (0.670)	Data 0.002 (0.084)	Loss 1.083 (0.774)
Epoch: [9][100/200]	Time 0.546 (0.669)	Data 0.001 (0.082)	Loss 0.886 (0.807)
Epoch: [9][120/200]	Time 0.577 (0.670)	Data 0.001 (0.083)	Loss 0.673 (0.834)
Epoch: [9][140/200]	Time 0.565 (0.668)	Data 0.001 (0.082)	Loss 0.726 (0.853)
Epoch: [9][160/200]	Time 0.668 (0.667)	Data 0.001 (0.081)	Loss 0.516 (0.861)
Epoch: [9][180/200]	Time 0.563 (0.667)	Data 0.001 (0.081)	Loss 0.625 (0.865)
Epoch: [9][200/200]	Time 0.563 (0.665)	Data 0.001 (0.080)	Loss 0.663 (0.869)
Extract Features: [50/76]	Time 0.205 (0.248)	Data 0.000 (0.025)	
Mean AP: 93.4%

 * Finished epoch   9  model mAP: 93.4%  best: 93.4% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.202 (0.252)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.742141008377075
==> Statistics for epoch 10: 609 clusters
Epoch: [10][20/200]	Time 2.165 (0.793)	Data 1.544 (0.114)	Loss 0.785 (0.181)
Epoch: [10][40/200]	Time 0.565 (0.719)	Data 0.001 (0.095)	Loss 0.920 (0.544)
Epoch: [10][60/200]	Time 0.547 (0.697)	Data 0.001 (0.087)	Loss 0.930 (0.676)
Epoch: [10][80/200]	Time 0.542 (0.687)	Data 0.002 (0.082)	Loss 0.678 (0.743)
Epoch: [10][100/200]	Time 0.667 (0.679)	Data 0.001 (0.079)	Loss 0.725 (0.777)
Epoch: [10][120/200]	Time 0.571 (0.675)	Data 0.001 (0.077)	Loss 1.073 (0.800)
Epoch: [10][140/200]	Time 0.544 (0.672)	Data 0.001 (0.077)	Loss 1.555 (0.819)
Epoch: [10][160/200]	Time 0.557 (0.669)	Data 0.001 (0.075)	Loss 0.964 (0.829)
Epoch: [10][180/200]	Time 0.562 (0.667)	Data 0.006 (0.074)	Loss 0.766 (0.830)
Epoch: [10][200/200]	Time 0.640 (0.665)	Data 0.001 (0.074)	Loss 1.227 (0.835)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.202 (0.234)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.832127809524536
==> Statistics for epoch 11: 604 clusters
Epoch: [11][20/200]	Time 0.564 (0.690)	Data 0.001 (0.111)	Loss 0.827 (0.223)
Epoch: [11][40/200]	Time 0.644 (0.671)	Data 0.003 (0.091)	Loss 0.823 (0.528)
Epoch: [11][60/200]	Time 0.572 (0.663)	Data 0.001 (0.085)	Loss 1.175 (0.638)
Epoch: [11][80/200]	Time 0.566 (0.662)	Data 0.001 (0.082)	Loss 0.755 (0.676)
Epoch: [11][100/200]	Time 0.575 (0.663)	Data 0.001 (0.080)	Loss 0.766 (0.708)
Epoch: [11][120/200]	Time 0.565 (0.659)	Data 0.000 (0.077)	Loss 0.812 (0.729)
Epoch: [11][140/200]	Time 0.545 (0.660)	Data 0.000 (0.077)	Loss 0.857 (0.753)
Epoch: [11][160/200]	Time 0.555 (0.659)	Data 0.000 (0.077)	Loss 0.689 (0.766)
Epoch: [11][180/200]	Time 0.648 (0.658)	Data 0.000 (0.077)	Loss 1.075 (0.781)
Epoch: [11][200/200]	Time 0.574 (0.665)	Data 0.001 (0.083)	Loss 0.844 (0.789)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.280 (0.247)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.565672874450684
==> Statistics for epoch 12: 606 clusters
Epoch: [12][20/200]	Time 0.565 (0.711)	Data 0.001 (0.123)	Loss 0.844 (0.233)
Epoch: [12][40/200]	Time 0.541 (0.690)	Data 0.002 (0.103)	Loss 0.869 (0.565)
Epoch: [12][60/200]	Time 0.549 (0.681)	Data 0.001 (0.095)	Loss 0.931 (0.630)
Epoch: [12][80/200]	Time 0.656 (0.678)	Data 0.001 (0.092)	Loss 0.970 (0.699)
Epoch: [12][100/200]	Time 0.565 (0.674)	Data 0.002 (0.088)	Loss 0.920 (0.715)
Epoch: [12][120/200]	Time 0.560 (0.670)	Data 0.000 (0.085)	Loss 0.995 (0.731)
Epoch: [12][140/200]	Time 0.561 (0.668)	Data 0.000 (0.084)	Loss 0.679 (0.752)
Epoch: [12][160/200]	Time 0.543 (0.667)	Data 0.000 (0.083)	Loss 0.604 (0.753)
Epoch: [12][180/200]	Time 0.650 (0.667)	Data 0.000 (0.081)	Loss 1.029 (0.759)
Epoch: [12][200/200]	Time 0.541 (0.671)	Data 0.001 (0.086)	Loss 0.912 (0.768)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.329 (0.245)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.854366540908813
==> Statistics for epoch 13: 608 clusters
Epoch: [13][20/200]	Time 2.062 (0.696)	Data 1.442 (0.106)	Loss 0.647 (0.155)
Epoch: [13][40/200]	Time 0.549 (0.676)	Data 0.001 (0.091)	Loss 0.725 (0.485)
Epoch: [13][60/200]	Time 0.543 (0.670)	Data 0.001 (0.087)	Loss 0.616 (0.581)
Epoch: [13][80/200]	Time 0.671 (0.668)	Data 0.002 (0.084)	Loss 0.708 (0.622)
Epoch: [13][100/200]	Time 0.558 (0.665)	Data 0.001 (0.082)	Loss 0.867 (0.666)
Epoch: [13][120/200]	Time 0.563 (0.661)	Data 0.001 (0.079)	Loss 0.943 (0.698)
Epoch: [13][140/200]	Time 0.547 (0.662)	Data 0.001 (0.079)	Loss 0.996 (0.714)
Epoch: [13][160/200]	Time 0.549 (0.661)	Data 0.001 (0.078)	Loss 0.702 (0.715)
Epoch: [13][180/200]	Time 0.567 (0.661)	Data 0.002 (0.078)	Loss 0.658 (0.725)
Epoch: [13][200/200]	Time 0.542 (0.660)	Data 0.001 (0.078)	Loss 0.992 (0.736)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.301 (0.247)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.5631160736084
==> Statistics for epoch 14: 604 clusters
Epoch: [14][20/200]	Time 0.547 (0.713)	Data 0.001 (0.126)	Loss 0.814 (0.194)
Epoch: [14][40/200]	Time 0.573 (0.689)	Data 0.002 (0.100)	Loss 0.603 (0.438)
Epoch: [14][60/200]	Time 0.546 (0.680)	Data 0.001 (0.093)	Loss 0.646 (0.549)
Epoch: [14][80/200]	Time 0.689 (0.677)	Data 0.001 (0.090)	Loss 0.771 (0.613)
Epoch: [14][100/200]	Time 0.564 (0.675)	Data 0.001 (0.088)	Loss 0.551 (0.647)
Epoch: [14][120/200]	Time 0.561 (0.672)	Data 0.000 (0.086)	Loss 0.718 (0.666)
Epoch: [14][140/200]	Time 0.561 (0.672)	Data 0.000 (0.085)	Loss 0.681 (0.678)
Epoch: [14][160/200]	Time 0.561 (0.670)	Data 0.000 (0.083)	Loss 0.581 (0.685)
Epoch: [14][180/200]	Time 0.566 (0.670)	Data 0.000 (0.083)	Loss 0.648 (0.691)
Epoch: [14][200/200]	Time 0.574 (0.677)	Data 0.001 (0.090)	Loss 0.614 (0.693)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.201 (0.245)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.556851387023926
==> Statistics for epoch 15: 607 clusters
Epoch: [15][20/200]	Time 0.572 (0.694)	Data 0.001 (0.111)	Loss 0.876 (0.200)
Epoch: [15][40/200]	Time 0.564 (0.668)	Data 0.001 (0.088)	Loss 0.776 (0.440)
Epoch: [15][60/200]	Time 0.650 (0.662)	Data 0.001 (0.082)	Loss 0.775 (0.541)
Epoch: [15][80/200]	Time 0.566 (0.657)	Data 0.001 (0.078)	Loss 0.411 (0.593)
Epoch: [15][100/200]	Time 0.540 (0.655)	Data 0.001 (0.076)	Loss 0.500 (0.622)
Epoch: [15][120/200]	Time 0.559 (0.656)	Data 0.000 (0.076)	Loss 0.749 (0.641)
Epoch: [15][140/200]	Time 0.557 (0.654)	Data 0.000 (0.074)	Loss 0.685 (0.658)
Epoch: [15][160/200]	Time 0.665 (0.654)	Data 0.000 (0.073)	Loss 0.749 (0.666)
Epoch: [15][180/200]	Time 0.557 (0.652)	Data 0.000 (0.072)	Loss 0.814 (0.673)
Epoch: [15][200/200]	Time 0.555 (0.662)	Data 0.001 (0.081)	Loss 0.749 (0.678)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.199 (0.241)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.849841117858887
==> Statistics for epoch 16: 604 clusters
Epoch: [16][20/200]	Time 0.562 (0.687)	Data 0.001 (0.102)	Loss 0.722 (0.191)
Epoch: [16][40/200]	Time 0.543 (0.668)	Data 0.001 (0.085)	Loss 0.769 (0.428)
Epoch: [16][60/200]	Time 0.567 (0.665)	Data 0.001 (0.082)	Loss 0.597 (0.527)
Epoch: [16][80/200]	Time 0.553 (0.663)	Data 0.001 (0.081)	Loss 0.723 (0.559)
Epoch: [16][100/200]	Time 0.641 (0.660)	Data 0.001 (0.077)	Loss 0.409 (0.597)
Epoch: [16][120/200]	Time 0.551 (0.658)	Data 0.000 (0.075)	Loss 1.063 (0.626)
Epoch: [16][140/200]	Time 0.690 (0.658)	Data 0.000 (0.074)	Loss 0.619 (0.640)
Epoch: [16][160/200]	Time 0.565 (0.657)	Data 0.000 (0.074)	Loss 0.623 (0.641)
Epoch: [16][180/200]	Time 0.565 (0.657)	Data 0.000 (0.074)	Loss 0.606 (0.645)
Epoch: [16][200/200]	Time 0.574 (0.666)	Data 0.001 (0.081)	Loss 0.796 (0.657)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.201 (0.241)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.759397506713867
==> Statistics for epoch 17: 606 clusters
Epoch: [17][20/200]	Time 0.578 (0.703)	Data 0.002 (0.115)	Loss 0.500 (0.159)
Epoch: [17][40/200]	Time 0.686 (0.688)	Data 0.002 (0.099)	Loss 0.588 (0.408)
Epoch: [17][60/200]	Time 0.565 (0.676)	Data 0.002 (0.091)	Loss 0.707 (0.499)
Epoch: [17][80/200]	Time 0.602 (0.673)	Data 0.002 (0.087)	Loss 0.664 (0.553)
Epoch: [17][100/200]	Time 0.561 (0.676)	Data 0.001 (0.089)	Loss 0.656 (0.578)
Epoch: [17][120/200]	Time 0.563 (0.673)	Data 0.000 (0.087)	Loss 0.609 (0.593)
Epoch: [17][140/200]	Time 0.668 (0.671)	Data 0.000 (0.086)	Loss 0.611 (0.608)
Epoch: [17][160/200]	Time 0.565 (0.670)	Data 0.000 (0.085)	Loss 0.721 (0.615)
Epoch: [17][180/200]	Time 0.640 (0.669)	Data 0.000 (0.084)	Loss 0.764 (0.621)
Epoch: [17][200/200]	Time 0.538 (0.676)	Data 0.001 (0.091)	Loss 0.552 (0.627)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.295 (0.242)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.89576268196106
==> Statistics for epoch 18: 602 clusters
Epoch: [18][20/200]	Time 0.674 (0.687)	Data 0.001 (0.106)	Loss 0.740 (0.173)
Epoch: [18][40/200]	Time 0.537 (0.664)	Data 0.002 (0.086)	Loss 0.909 (0.437)
Epoch: [18][60/200]	Time 0.538 (0.657)	Data 0.001 (0.082)	Loss 0.596 (0.508)
Epoch: [18][80/200]	Time 0.569 (0.654)	Data 0.001 (0.079)	Loss 0.814 (0.544)
Epoch: [18][100/200]	Time 0.541 (0.651)	Data 0.001 (0.077)	Loss 0.493 (0.577)
Epoch: [18][120/200]	Time 0.559 (0.652)	Data 0.000 (0.077)	Loss 0.414 (0.592)
Epoch: [18][140/200]	Time 0.567 (0.650)	Data 0.000 (0.075)	Loss 0.529 (0.604)
Epoch: [18][160/200]	Time 0.655 (0.650)	Data 0.000 (0.074)	Loss 0.637 (0.607)
Epoch: [18][180/200]	Time 0.545 (0.649)	Data 0.000 (0.073)	Loss 0.839 (0.615)
Epoch: [18][200/200]	Time 0.559 (0.655)	Data 0.001 (0.079)	Loss 0.631 (0.621)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.199 (0.248)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.57071089744568
==> Statistics for epoch 19: 604 clusters
Epoch: [19][20/200]	Time 0.565 (0.725)	Data 0.001 (0.126)	Loss 0.558 (0.178)
Epoch: [19][40/200]	Time 0.567 (0.698)	Data 0.002 (0.105)	Loss 0.773 (0.408)
Epoch: [19][60/200]	Time 0.660 (0.688)	Data 0.001 (0.096)	Loss 0.602 (0.469)
Epoch: [19][80/200]	Time 0.562 (0.683)	Data 0.001 (0.093)	Loss 0.999 (0.521)
Epoch: [19][100/200]	Time 0.639 (0.680)	Data 0.001 (0.090)	Loss 0.690 (0.542)
Epoch: [19][120/200]	Time 0.564 (0.676)	Data 0.000 (0.088)	Loss 0.604 (0.567)
Epoch: [19][140/200]	Time 0.553 (0.675)	Data 0.000 (0.087)	Loss 0.451 (0.575)
Epoch: [19][160/200]	Time 0.555 (0.673)	Data 0.000 (0.085)	Loss 0.537 (0.585)
Epoch: [19][180/200]	Time 0.581 (0.672)	Data 0.000 (0.084)	Loss 0.727 (0.593)
Epoch: [19][200/200]	Time 0.578 (0.680)	Data 0.001 (0.091)	Loss 0.481 (0.604)
Extract Features: [50/76]	Time 0.189 (0.247)	Data 0.000 (0.025)	
Mean AP: 93.7%

 * Finished epoch  19  model mAP: 93.7%  best: 93.7% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.341 (0.246)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.008641242980957
==> Statistics for epoch 20: 605 clusters
Epoch: [20][20/200]	Time 0.550 (0.706)	Data 0.001 (0.119)	Loss 0.871 (0.179)
Epoch: [20][40/200]	Time 0.569 (0.682)	Data 0.002 (0.095)	Loss 0.605 (0.383)
Epoch: [20][60/200]	Time 0.570 (0.672)	Data 0.001 (0.088)	Loss 0.678 (0.461)
Epoch: [20][80/200]	Time 0.551 (0.668)	Data 0.001 (0.085)	Loss 0.694 (0.492)
Epoch: [20][100/200]	Time 0.553 (0.666)	Data 0.003 (0.083)	Loss 0.660 (0.520)
Epoch: [20][120/200]	Time 0.651 (0.666)	Data 0.000 (0.082)	Loss 0.539 (0.533)
Epoch: [20][140/200]	Time 0.566 (0.665)	Data 0.000 (0.082)	Loss 0.531 (0.549)
Epoch: [20][160/200]	Time 0.648 (0.665)	Data 0.000 (0.081)	Loss 0.549 (0.557)
Epoch: [20][180/200]	Time 0.562 (0.664)	Data 0.000 (0.080)	Loss 0.493 (0.558)
Epoch: [20][200/200]	Time 0.554 (0.672)	Data 0.001 (0.087)	Loss 0.629 (0.561)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.200 (0.249)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.347121477127075
==> Statistics for epoch 21: 602 clusters
Epoch: [21][20/200]	Time 0.562 (0.709)	Data 0.001 (0.111)	Loss 0.533 (0.136)
Epoch: [21][40/200]	Time 0.565 (0.675)	Data 0.001 (0.088)	Loss 0.476 (0.337)
Epoch: [21][60/200]	Time 0.570 (0.669)	Data 0.001 (0.084)	Loss 0.551 (0.428)
Epoch: [21][80/200]	Time 0.544 (0.663)	Data 0.001 (0.081)	Loss 0.489 (0.466)
Epoch: [21][100/200]	Time 0.646 (0.662)	Data 0.001 (0.080)	Loss 0.509 (0.486)
Epoch: [21][120/200]	Time 0.560 (0.660)	Data 0.000 (0.079)	Loss 0.513 (0.505)
Epoch: [21][140/200]	Time 0.640 (0.659)	Data 0.000 (0.078)	Loss 0.535 (0.521)
Epoch: [21][160/200]	Time 0.544 (0.659)	Data 0.000 (0.077)	Loss 0.604 (0.527)
Epoch: [21][180/200]	Time 0.556 (0.659)	Data 0.000 (0.077)	Loss 0.510 (0.536)
Epoch: [21][200/200]	Time 0.571 (0.667)	Data 0.001 (0.083)	Loss 0.551 (0.544)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.203 (0.237)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.92853832244873
==> Statistics for epoch 22: 601 clusters
Epoch: [22][20/200]	Time 0.574 (0.699)	Data 0.002 (0.122)	Loss 0.563 (0.142)
Epoch: [22][40/200]	Time 0.659 (0.679)	Data 0.002 (0.096)	Loss 0.357 (0.351)
Epoch: [22][60/200]	Time 0.548 (0.671)	Data 0.002 (0.089)	Loss 0.610 (0.436)
Epoch: [22][80/200]	Time 0.562 (0.666)	Data 0.002 (0.086)	Loss 0.824 (0.483)
Epoch: [22][100/200]	Time 0.558 (0.662)	Data 0.002 (0.081)	Loss 0.590 (0.511)
Epoch: [22][120/200]	Time 0.559 (0.660)	Data 0.000 (0.080)	Loss 0.659 (0.522)
Epoch: [22][140/200]	Time 0.541 (0.659)	Data 0.000 (0.078)	Loss 0.728 (0.531)
Epoch: [22][160/200]	Time 0.541 (0.659)	Data 0.000 (0.078)	Loss 0.514 (0.539)
Epoch: [22][180/200]	Time 0.560 (0.659)	Data 0.000 (0.076)	Loss 0.593 (0.547)
Epoch: [22][200/200]	Time 0.559 (0.665)	Data 0.001 (0.082)	Loss 0.868 (0.550)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.203 (0.247)	Data 0.000 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.590953826904297
==> Statistics for epoch 23: 603 clusters
Epoch: [23][20/200]	Time 0.677 (0.703)	Data 0.001 (0.113)	Loss 0.715 (0.147)
Epoch: [23][40/200]	Time 0.558 (0.680)	Data 0.001 (0.097)	Loss 0.607 (0.328)
Epoch: [23][60/200]	Time 0.700 (0.676)	Data 0.001 (0.090)	Loss 0.627 (0.412)
Epoch: [23][80/200]	Time 0.561 (0.673)	Data 0.001 (0.086)	Loss 0.447 (0.437)
Epoch: [23][100/200]	Time 0.567 (0.671)	Data 0.001 (0.085)	Loss 0.719 (0.468)
Epoch: [23][120/200]	Time 0.559 (0.669)	Data 0.000 (0.083)	Loss 0.676 (0.489)
Epoch: [23][140/200]	Time 0.550 (0.668)	Data 0.000 (0.082)	Loss 0.814 (0.503)
Epoch: [23][160/200]	Time 0.559 (0.668)	Data 0.000 (0.082)	Loss 0.297 (0.508)
Epoch: [23][180/200]	Time 0.560 (0.667)	Data 0.000 (0.081)	Loss 0.434 (0.514)
Epoch: [23][200/200]	Time 0.708 (0.674)	Data 0.001 (0.087)	Loss 0.570 (0.517)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.244)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.58131718635559
==> Statistics for epoch 24: 602 clusters
Epoch: [24][20/200]	Time 0.546 (0.688)	Data 0.001 (0.111)	Loss 0.652 (0.136)
Epoch: [24][40/200]	Time 0.562 (0.678)	Data 0.002 (0.098)	Loss 0.895 (0.345)
Epoch: [24][60/200]	Time 0.557 (0.670)	Data 0.001 (0.088)	Loss 0.654 (0.412)
Epoch: [24][80/200]	Time 0.562 (0.661)	Data 0.001 (0.083)	Loss 0.444 (0.455)
Epoch: [24][100/200]	Time 0.563 (0.660)	Data 0.001 (0.080)	Loss 0.557 (0.482)
Epoch: [24][120/200]	Time 0.551 (0.657)	Data 0.000 (0.077)	Loss 0.519 (0.488)
Epoch: [24][140/200]	Time 0.562 (0.656)	Data 0.000 (0.077)	Loss 0.612 (0.495)
Epoch: [24][160/200]	Time 0.563 (0.655)	Data 0.000 (0.076)	Loss 0.723 (0.504)
Epoch: [24][180/200]	Time 0.642 (0.658)	Data 0.000 (0.077)	Loss 0.654 (0.509)
Epoch: [24][200/200]	Time 0.575 (0.663)	Data 0.001 (0.082)	Loss 0.472 (0.514)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.272 (0.242)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.261285066604614
==> Statistics for epoch 25: 602 clusters
Epoch: [25][20/200]	Time 0.556 (0.698)	Data 0.001 (0.108)	Loss 0.698 (0.158)
Epoch: [25][40/200]	Time 0.568 (0.676)	Data 0.007 (0.090)	Loss 0.732 (0.353)
Epoch: [25][60/200]	Time 0.556 (0.665)	Data 0.006 (0.082)	Loss 0.337 (0.433)
Epoch: [25][80/200]	Time 0.566 (0.666)	Data 0.001 (0.081)	Loss 0.819 (0.477)
Epoch: [25][100/200]	Time 0.542 (0.663)	Data 0.001 (0.078)	Loss 0.495 (0.494)
Epoch: [25][120/200]	Time 0.558 (0.662)	Data 0.000 (0.077)	Loss 0.630 (0.511)
Epoch: [25][140/200]	Time 0.553 (0.660)	Data 0.000 (0.076)	Loss 0.595 (0.517)
Epoch: [25][160/200]	Time 0.648 (0.659)	Data 0.000 (0.074)	Loss 0.712 (0.523)
Epoch: [25][180/200]	Time 0.547 (0.657)	Data 0.000 (0.074)	Loss 0.342 (0.533)
Epoch: [25][200/200]	Time 0.760 (0.664)	Data 0.001 (0.080)	Loss 0.870 (0.541)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.194 (0.240)	Data 0.000 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.361112594604492
==> Statistics for epoch 26: 602 clusters
Epoch: [26][20/200]	Time 0.568 (0.685)	Data 0.001 (0.110)	Loss 0.565 (0.125)
Epoch: [26][40/200]	Time 0.541 (0.666)	Data 0.002 (0.091)	Loss 0.454 (0.327)
Epoch: [26][60/200]	Time 0.564 (0.660)	Data 0.001 (0.083)	Loss 0.479 (0.391)
Epoch: [26][80/200]	Time 0.570 (0.656)	Data 0.001 (0.079)	Loss 0.650 (0.438)
Epoch: [26][100/200]	Time 0.662 (0.653)	Data 0.001 (0.076)	Loss 0.473 (0.463)
Epoch: [26][120/200]	Time 0.554 (0.651)	Data 0.000 (0.075)	Loss 0.737 (0.484)
Epoch: [26][140/200]	Time 0.667 (0.651)	Data 0.000 (0.075)	Loss 0.563 (0.498)
Epoch: [26][160/200]	Time 0.570 (0.651)	Data 0.000 (0.073)	Loss 0.626 (0.510)
Epoch: [26][180/200]	Time 0.561 (0.654)	Data 0.000 (0.075)	Loss 0.651 (0.517)
Epoch: [26][200/200]	Time 0.562 (0.662)	Data 0.001 (0.082)	Loss 0.702 (0.524)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.241)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.308910131454468
==> Statistics for epoch 27: 600 clusters
Epoch: [27][20/200]	Time 0.569 (0.699)	Data 0.001 (0.118)	Loss 0.478 (0.123)
Epoch: [27][40/200]	Time 0.553 (0.683)	Data 0.001 (0.098)	Loss 0.809 (0.344)
Epoch: [27][60/200]	Time 0.564 (0.675)	Data 0.001 (0.092)	Loss 0.551 (0.413)
Epoch: [27][80/200]	Time 0.565 (0.671)	Data 0.001 (0.088)	Loss 0.911 (0.447)
Epoch: [27][100/200]	Time 0.542 (0.671)	Data 0.001 (0.088)	Loss 0.549 (0.465)
Epoch: [27][120/200]	Time 0.546 (0.670)	Data 0.000 (0.087)	Loss 0.563 (0.481)
Epoch: [27][140/200]	Time 0.569 (0.669)	Data 0.000 (0.085)	Loss 0.498 (0.485)
Epoch: [27][160/200]	Time 0.560 (0.667)	Data 0.000 (0.084)	Loss 0.526 (0.493)
Epoch: [27][180/200]	Time 0.546 (0.666)	Data 0.000 (0.083)	Loss 0.540 (0.502)
Epoch: [27][200/200]	Time 0.568 (0.676)	Data 0.002 (0.091)	Loss 0.467 (0.507)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.192 (0.251)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.5325767993927
==> Statistics for epoch 28: 601 clusters
Epoch: [28][20/200]	Time 0.663 (0.698)	Data 0.001 (0.113)	Loss 0.683 (0.144)
Epoch: [28][40/200]	Time 0.559 (0.677)	Data 0.002 (0.094)	Loss 0.500 (0.340)
Epoch: [28][60/200]	Time 0.549 (0.672)	Data 0.005 (0.091)	Loss 0.698 (0.422)
Epoch: [28][80/200]	Time 0.557 (0.669)	Data 0.001 (0.086)	Loss 0.711 (0.466)
Epoch: [28][100/200]	Time 0.569 (0.665)	Data 0.001 (0.082)	Loss 0.708 (0.484)
Epoch: [28][120/200]	Time 0.669 (0.663)	Data 0.000 (0.080)	Loss 0.640 (0.497)
Epoch: [28][140/200]	Time 0.546 (0.661)	Data 0.000 (0.079)	Loss 0.597 (0.507)
Epoch: [28][160/200]	Time 0.557 (0.660)	Data 0.000 (0.078)	Loss 0.558 (0.512)
Epoch: [28][180/200]	Time 0.537 (0.658)	Data 0.000 (0.077)	Loss 0.335 (0.517)
Epoch: [28][200/200]	Time 0.558 (0.664)	Data 0.001 (0.083)	Loss 0.594 (0.519)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.244)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.5473473072052
==> Statistics for epoch 29: 603 clusters
Epoch: [29][20/200]	Time 0.547 (0.705)	Data 0.001 (0.114)	Loss 0.391 (0.131)
Epoch: [29][40/200]	Time 0.563 (0.673)	Data 0.001 (0.094)	Loss 0.423 (0.343)
Epoch: [29][60/200]	Time 0.566 (0.667)	Data 0.001 (0.088)	Loss 0.625 (0.411)
Epoch: [29][80/200]	Time 0.579 (0.664)	Data 0.001 (0.085)	Loss 0.511 (0.450)
Epoch: [29][100/200]	Time 0.647 (0.660)	Data 0.001 (0.081)	Loss 0.496 (0.469)
Epoch: [29][120/200]	Time 0.563 (0.660)	Data 0.000 (0.082)	Loss 0.592 (0.482)
Epoch: [29][140/200]	Time 0.644 (0.660)	Data 0.000 (0.082)	Loss 0.414 (0.490)
Epoch: [29][160/200]	Time 0.539 (0.657)	Data 0.000 (0.080)	Loss 0.713 (0.502)
Epoch: [29][180/200]	Time 0.541 (0.657)	Data 0.000 (0.080)	Loss 0.383 (0.512)
Epoch: [29][200/200]	Time 0.548 (0.665)	Data 0.001 (0.087)	Loss 0.478 (0.514)
Extract Features: [50/76]	Time 0.188 (0.233)	Data 0.000 (0.023)	
Mean AP: 94.2%

 * Finished epoch  29  model mAP: 94.2%  best: 94.2% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.239)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.417512893676758
==> Statistics for epoch 30: 599 clusters
Epoch: [30][20/200]	Time 0.571 (0.688)	Data 0.001 (0.097)	Loss 0.535 (0.131)
Epoch: [30][40/200]	Time 0.687 (0.670)	Data 0.001 (0.081)	Loss 0.609 (0.350)
Epoch: [30][60/200]	Time 0.566 (0.660)	Data 0.002 (0.075)	Loss 0.424 (0.410)
Epoch: [30][80/200]	Time 0.645 (0.658)	Data 0.001 (0.074)	Loss 0.684 (0.442)
Epoch: [30][100/200]	Time 0.556 (0.654)	Data 0.001 (0.072)	Loss 0.608 (0.465)
Epoch: [30][120/200]	Time 0.561 (0.653)	Data 0.000 (0.072)	Loss 0.445 (0.471)
Epoch: [30][140/200]	Time 0.559 (0.654)	Data 0.000 (0.072)	Loss 0.614 (0.486)
Epoch: [30][160/200]	Time 0.545 (0.654)	Data 0.000 (0.072)	Loss 0.622 (0.496)
Epoch: [30][180/200]	Time 0.556 (0.654)	Data 0.000 (0.072)	Loss 0.675 (0.504)
Epoch: [30][200/200]	Time 0.569 (0.663)	Data 0.001 (0.080)	Loss 0.701 (0.508)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.189 (0.249)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.737436056137085
==> Statistics for epoch 31: 603 clusters
Epoch: [31][20/200]	Time 0.578 (0.697)	Data 0.001 (0.110)	Loss 0.410 (0.128)
Epoch: [31][40/200]	Time 0.558 (0.670)	Data 0.002 (0.091)	Loss 0.604 (0.332)
Epoch: [31][60/200]	Time 0.648 (0.664)	Data 0.001 (0.084)	Loss 0.526 (0.399)
Epoch: [31][80/200]	Time 0.569 (0.660)	Data 0.001 (0.081)	Loss 0.949 (0.452)
Epoch: [31][100/200]	Time 0.561 (0.656)	Data 0.001 (0.078)	Loss 0.279 (0.475)
Epoch: [31][120/200]	Time 0.574 (0.654)	Data 0.000 (0.076)	Loss 0.528 (0.480)
Epoch: [31][140/200]	Time 0.550 (0.658)	Data 0.000 (0.079)	Loss 0.495 (0.488)
Epoch: [31][160/200]	Time 0.627 (0.658)	Data 0.000 (0.079)	Loss 0.451 (0.495)
Epoch: [31][180/200]	Time 0.563 (0.658)	Data 0.000 (0.078)	Loss 0.790 (0.506)
Epoch: [31][200/200]	Time 0.581 (0.665)	Data 0.002 (0.085)	Loss 0.417 (0.507)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.205 (0.250)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.552035570144653
==> Statistics for epoch 32: 601 clusters
Epoch: [32][20/200]	Time 0.565 (0.714)	Data 0.001 (0.113)	Loss 0.614 (0.136)
Epoch: [32][40/200]	Time 0.550 (0.685)	Data 0.002 (0.095)	Loss 0.400 (0.340)
Epoch: [32][60/200]	Time 0.706 (0.678)	Data 0.007 (0.087)	Loss 0.759 (0.413)
Epoch: [32][80/200]	Time 0.560 (0.675)	Data 0.001 (0.085)	Loss 0.420 (0.443)
Epoch: [32][100/200]	Time 0.547 (0.671)	Data 0.001 (0.082)	Loss 0.322 (0.466)
Epoch: [32][120/200]	Time 0.568 (0.670)	Data 0.000 (0.082)	Loss 0.410 (0.484)
Epoch: [32][140/200]	Time 0.574 (0.668)	Data 0.000 (0.079)	Loss 0.706 (0.491)
Epoch: [32][160/200]	Time 0.564 (0.667)	Data 0.000 (0.079)	Loss 0.618 (0.503)
Epoch: [32][180/200]	Time 0.563 (0.667)	Data 0.000 (0.080)	Loss 0.442 (0.507)
Epoch: [32][200/200]	Time 0.670 (0.675)	Data 0.001 (0.087)	Loss 0.900 (0.511)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.188 (0.247)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.179251432418823
==> Statistics for epoch 33: 601 clusters
Epoch: [33][20/200]	Time 0.546 (0.686)	Data 0.002 (0.103)	Loss 0.533 (0.126)
Epoch: [33][40/200]	Time 0.566 (0.673)	Data 0.002 (0.091)	Loss 0.432 (0.303)
Epoch: [33][60/200]	Time 0.567 (0.670)	Data 0.001 (0.087)	Loss 0.419 (0.369)
Epoch: [33][80/200]	Time 0.564 (0.667)	Data 0.001 (0.086)	Loss 0.430 (0.418)
Epoch: [33][100/200]	Time 0.563 (0.668)	Data 0.001 (0.084)	Loss 0.743 (0.436)
Epoch: [33][120/200]	Time 0.560 (0.666)	Data 0.000 (0.083)	Loss 0.705 (0.459)
Epoch: [33][140/200]	Time 0.560 (0.663)	Data 0.000 (0.080)	Loss 0.462 (0.472)
Epoch: [33][160/200]	Time 0.557 (0.663)	Data 0.000 (0.080)	Loss 0.412 (0.481)
Epoch: [33][180/200]	Time 0.641 (0.662)	Data 0.000 (0.079)	Loss 0.473 (0.483)
Epoch: [33][200/200]	Time 0.575 (0.669)	Data 0.001 (0.085)	Loss 0.554 (0.488)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.249)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.99234700202942
==> Statistics for epoch 34: 599 clusters
Epoch: [34][20/200]	Time 0.697 (0.720)	Data 0.000 (0.119)	Loss 0.562 (0.123)
Epoch: [34][40/200]	Time 0.567 (0.690)	Data 0.002 (0.095)	Loss 0.382 (0.316)
Epoch: [34][60/200]	Time 0.553 (0.681)	Data 0.001 (0.090)	Loss 0.568 (0.400)
Epoch: [34][80/200]	Time 0.543 (0.677)	Data 0.001 (0.087)	Loss 0.401 (0.435)
Epoch: [34][100/200]	Time 0.559 (0.674)	Data 0.001 (0.085)	Loss 0.524 (0.451)
Epoch: [34][120/200]	Time 0.640 (0.671)	Data 0.000 (0.083)	Loss 0.499 (0.463)
Epoch: [34][140/200]	Time 0.562 (0.669)	Data 0.000 (0.082)	Loss 0.399 (0.469)
Epoch: [34][160/200]	Time 0.566 (0.669)	Data 0.000 (0.082)	Loss 0.558 (0.482)
Epoch: [34][180/200]	Time 0.567 (0.667)	Data 0.000 (0.081)	Loss 0.704 (0.491)
Epoch: [34][200/200]	Time 0.551 (0.675)	Data 0.002 (0.088)	Loss 0.642 (0.493)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.185 (0.241)	Data 0.000 (0.026)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.374438285827637
==> Statistics for epoch 35: 601 clusters
Epoch: [35][20/200]	Time 0.547 (0.698)	Data 0.001 (0.119)	Loss 0.584 (0.128)
Epoch: [35][40/200]	Time 0.565 (0.682)	Data 0.001 (0.096)	Loss 0.503 (0.324)
Epoch: [35][60/200]	Time 0.565 (0.676)	Data 0.001 (0.090)	Loss 0.457 (0.392)
Epoch: [35][80/200]	Time 0.667 (0.673)	Data 0.001 (0.086)	Loss 0.580 (0.420)
Epoch: [35][100/200]	Time 0.570 (0.671)	Data 0.006 (0.084)	Loss 0.326 (0.450)
Epoch: [35][120/200]	Time 0.546 (0.670)	Data 0.000 (0.083)	Loss 0.621 (0.469)
Epoch: [35][140/200]	Time 0.567 (0.670)	Data 0.000 (0.083)	Loss 0.497 (0.480)
Epoch: [35][160/200]	Time 0.560 (0.668)	Data 0.000 (0.082)	Loss 0.485 (0.485)
Epoch: [35][180/200]	Time 0.563 (0.668)	Data 0.000 (0.082)	Loss 0.619 (0.490)
Epoch: [35][200/200]	Time 0.602 (0.676)	Data 0.001 (0.090)	Loss 0.621 (0.494)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.197 (0.241)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.556522130966187
==> Statistics for epoch 36: 600 clusters
Epoch: [36][20/200]	Time 0.577 (0.700)	Data 0.000 (0.111)	Loss 0.516 (0.113)
Epoch: [36][40/200]	Time 0.546 (0.676)	Data 0.000 (0.091)	Loss 0.673 (0.323)
Epoch: [36][60/200]	Time 0.563 (0.666)	Data 0.000 (0.085)	Loss 0.434 (0.405)
Epoch: [36][80/200]	Time 0.552 (0.660)	Data 0.000 (0.079)	Loss 0.495 (0.434)
Epoch: [36][100/200]	Time 0.564 (0.659)	Data 0.000 (0.080)	Loss 0.788 (0.455)
Epoch: [36][120/200]	Time 0.555 (0.656)	Data 0.000 (0.077)	Loss 0.405 (0.470)
Epoch: [36][140/200]	Time 0.566 (0.654)	Data 0.000 (0.076)	Loss 0.352 (0.482)
Epoch: [36][160/200]	Time 0.577 (0.654)	Data 0.000 (0.075)	Loss 0.553 (0.487)
Epoch: [36][180/200]	Time 0.551 (0.653)	Data 0.000 (0.074)	Loss 0.658 (0.489)
Epoch: [36][200/200]	Time 0.661 (0.660)	Data 0.000 (0.080)	Loss 0.685 (0.496)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.233)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.885538339614868
==> Statistics for epoch 37: 599 clusters
Epoch: [37][20/200]	Time 0.585 (0.727)	Data 0.001 (0.126)	Loss 0.426 (0.139)
Epoch: [37][40/200]	Time 0.571 (0.694)	Data 0.002 (0.103)	Loss 0.647 (0.325)
Epoch: [37][60/200]	Time 0.562 (0.686)	Data 0.001 (0.094)	Loss 0.639 (0.407)
Epoch: [37][80/200]	Time 0.574 (0.681)	Data 0.001 (0.091)	Loss 0.560 (0.441)
Epoch: [37][100/200]	Time 0.633 (0.678)	Data 0.001 (0.088)	Loss 0.515 (0.464)
Epoch: [37][120/200]	Time 0.576 (0.675)	Data 0.000 (0.085)	Loss 0.372 (0.480)
Epoch: [37][140/200]	Time 0.567 (0.672)	Data 0.000 (0.084)	Loss 0.636 (0.494)
Epoch: [37][160/200]	Time 0.545 (0.671)	Data 0.000 (0.082)	Loss 0.484 (0.502)
Epoch: [37][180/200]	Time 0.542 (0.669)	Data 0.000 (0.081)	Loss 0.568 (0.500)
Epoch: [37][200/200]	Time 0.702 (0.677)	Data 0.001 (0.088)	Loss 0.441 (0.507)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.188 (0.243)	Data 0.000 (0.027)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.728610038757324
==> Statistics for epoch 38: 598 clusters
Epoch: [38][20/200]	Time 0.564 (0.694)	Data 0.001 (0.110)	Loss 0.668 (0.129)
Epoch: [38][40/200]	Time 0.561 (0.671)	Data 0.002 (0.090)	Loss 0.661 (0.337)
Epoch: [38][60/200]	Time 0.564 (0.666)	Data 0.001 (0.081)	Loss 0.547 (0.409)
Epoch: [38][80/200]	Time 0.563 (0.662)	Data 0.001 (0.079)	Loss 0.795 (0.436)
Epoch: [38][100/200]	Time 0.553 (0.659)	Data 0.001 (0.076)	Loss 0.486 (0.456)
Epoch: [38][120/200]	Time 0.557 (0.656)	Data 0.000 (0.074)	Loss 0.304 (0.465)
Epoch: [38][140/200]	Time 0.679 (0.657)	Data 0.000 (0.074)	Loss 0.599 (0.475)
Epoch: [38][160/200]	Time 0.560 (0.659)	Data 0.000 (0.076)	Loss 0.622 (0.486)
Epoch: [38][180/200]	Time 0.562 (0.658)	Data 0.000 (0.075)	Loss 0.638 (0.490)
Epoch: [38][200/200]	Time 0.538 (0.665)	Data 0.001 (0.082)	Loss 0.649 (0.493)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.198 (0.233)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.704521656036377
==> Statistics for epoch 39: 601 clusters
Epoch: [39][20/200]	Time 0.539 (0.685)	Data 0.001 (0.103)	Loss 0.414 (0.116)
Epoch: [39][40/200]	Time 0.677 (0.665)	Data 0.001 (0.083)	Loss 0.553 (0.309)
Epoch: [39][60/200]	Time 0.562 (0.658)	Data 0.001 (0.079)	Loss 0.486 (0.405)
Epoch: [39][80/200]	Time 0.563 (0.658)	Data 0.002 (0.080)	Loss 0.462 (0.434)
Epoch: [39][100/200]	Time 0.547 (0.655)	Data 0.001 (0.078)	Loss 0.418 (0.463)
Epoch: [39][120/200]	Time 0.544 (0.653)	Data 0.000 (0.076)	Loss 0.460 (0.474)
Epoch: [39][140/200]	Time 0.559 (0.652)	Data 0.000 (0.075)	Loss 0.365 (0.490)
Epoch: [39][160/200]	Time 0.539 (0.650)	Data 0.000 (0.074)	Loss 0.718 (0.497)
Epoch: [39][180/200]	Time 0.664 (0.650)	Data 0.000 (0.073)	Loss 0.637 (0.501)
Epoch: [39][200/200]	Time 0.564 (0.656)	Data 0.001 (0.079)	Loss 0.494 (0.506)
Extract Features: [50/76]	Time 0.187 (0.240)	Data 0.000 (0.021)	
Mean AP: 94.3%

 * Finished epoch  39  model mAP: 94.3%  best: 94.3% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.201 (0.240)	Data 0.000 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.830167770385742
==> Statistics for epoch 40: 600 clusters
Epoch: [40][20/200]	Time 0.562 (0.682)	Data 0.001 (0.104)	Loss 0.425 (0.115)
Epoch: [40][40/200]	Time 0.560 (0.667)	Data 0.002 (0.086)	Loss 0.675 (0.301)
Epoch: [40][60/200]	Time 0.547 (0.657)	Data 0.001 (0.078)	Loss 0.389 (0.353)
Epoch: [40][80/200]	Time 0.570 (0.658)	Data 0.001 (0.077)	Loss 0.532 (0.394)
Epoch: [40][100/200]	Time 0.563 (0.663)	Data 0.001 (0.081)	Loss 0.498 (0.412)
Epoch: [40][120/200]	Time 0.640 (0.660)	Data 0.000 (0.079)	Loss 0.423 (0.430)
Epoch: [40][140/200]	Time 0.567 (0.661)	Data 0.000 (0.079)	Loss 0.664 (0.448)
Epoch: [40][160/200]	Time 0.573 (0.661)	Data 0.000 (0.078)	Loss 0.475 (0.457)
Epoch: [40][180/200]	Time 0.565 (0.662)	Data 0.000 (0.079)	Loss 0.675 (0.464)
Epoch: [40][200/200]	Time 0.553 (0.670)	Data 0.001 (0.085)	Loss 0.614 (0.467)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.202 (0.255)	Data 0.000 (0.033)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.636826038360596
==> Statistics for epoch 41: 600 clusters
Epoch: [41][20/200]	Time 0.565 (0.701)	Data 0.001 (0.112)	Loss 0.586 (0.126)
Epoch: [41][40/200]	Time 0.549 (0.674)	Data 0.001 (0.091)	Loss 0.646 (0.317)
Epoch: [41][60/200]	Time 0.662 (0.670)	Data 0.001 (0.086)	Loss 0.363 (0.387)
Epoch: [41][80/200]	Time 0.569 (0.667)	Data 0.001 (0.081)	Loss 0.382 (0.414)
Epoch: [41][100/200]	Time 0.564 (0.663)	Data 0.001 (0.079)	Loss 0.457 (0.432)
Epoch: [41][120/200]	Time 0.569 (0.663)	Data 0.000 (0.078)	Loss 0.685 (0.447)
Epoch: [41][140/200]	Time 0.565 (0.662)	Data 0.000 (0.076)	Loss 0.412 (0.452)
Epoch: [41][160/200]	Time 0.560 (0.660)	Data 0.000 (0.075)	Loss 0.627 (0.462)
Epoch: [41][180/200]	Time 0.557 (0.659)	Data 0.000 (0.075)	Loss 0.429 (0.467)
Epoch: [41][200/200]	Time 0.727 (0.667)	Data 0.002 (0.081)	Loss 0.397 (0.474)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.201 (0.240)	Data 0.000 (0.025)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.44990563392639
==> Statistics for epoch 42: 599 clusters
Epoch: [42][20/200]	Time 0.579 (0.705)	Data 0.001 (0.116)	Loss 0.796 (0.140)
Epoch: [42][40/200]	Time 0.570 (0.675)	Data 0.003 (0.096)	Loss 0.357 (0.324)
Epoch: [42][60/200]	Time 0.548 (0.671)	Data 0.001 (0.089)	Loss 0.406 (0.392)
Epoch: [42][80/200]	Time 0.565 (0.667)	Data 0.001 (0.087)	Loss 0.294 (0.429)
Epoch: [42][100/200]	Time 0.572 (0.664)	Data 0.001 (0.083)	Loss 0.574 (0.443)
Epoch: [42][120/200]	Time 0.556 (0.662)	Data 0.000 (0.080)	Loss 0.480 (0.464)
Epoch: [42][140/200]	Time 0.689 (0.661)	Data 0.000 (0.079)	Loss 0.548 (0.473)
Epoch: [42][160/200]	Time 0.556 (0.659)	Data 0.000 (0.077)	Loss 0.601 (0.479)
Epoch: [42][180/200]	Time 0.561 (0.658)	Data 0.000 (0.077)	Loss 0.355 (0.489)
Epoch: [42][200/200]	Time 0.568 (0.664)	Data 0.001 (0.083)	Loss 0.595 (0.494)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.185 (0.240)	Data 0.000 (0.024)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.718705654144287
==> Statistics for epoch 43: 601 clusters
Epoch: [43][20/200]	Time 0.559 (0.697)	Data 0.001 (0.117)	Loss 0.563 (0.141)
Epoch: [43][40/200]	Time 0.562 (0.678)	Data 0.002 (0.094)	Loss 0.556 (0.355)
Epoch: [43][60/200]	Time 0.565 (0.672)	Data 0.002 (0.089)	Loss 0.506 (0.407)
Epoch: [43][80/200]	Time 0.546 (0.667)	Data 0.001 (0.084)	Loss 0.560 (0.427)
Epoch: [43][100/200]	Time 0.557 (0.666)	Data 0.001 (0.083)	Loss 0.746 (0.443)
Epoch: [43][120/200]	Time 0.641 (0.667)	Data 0.000 (0.082)	Loss 0.514 (0.450)
Epoch: [43][140/200]	Time 0.554 (0.664)	Data 0.000 (0.081)	Loss 0.434 (0.458)
Epoch: [43][160/200]	Time 0.631 (0.662)	Data 0.000 (0.080)	Loss 0.551 (0.466)
Epoch: [43][180/200]	Time 0.561 (0.661)	Data 0.000 (0.078)	Loss 0.666 (0.469)
Epoch: [43][200/200]	Time 0.698 (0.668)	Data 0.001 (0.085)	Loss 0.679 (0.471)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.192 (0.243)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.924404621124268
==> Statistics for epoch 44: 601 clusters
Epoch: [44][20/200]	Time 0.569 (0.700)	Data 0.001 (0.112)	Loss 0.490 (0.127)
Epoch: [44][40/200]	Time 0.575 (0.684)	Data 0.003 (0.097)	Loss 0.566 (0.328)
Epoch: [44][60/200]	Time 0.640 (0.679)	Data 0.001 (0.092)	Loss 0.431 (0.384)
Epoch: [44][80/200]	Time 0.567 (0.677)	Data 0.001 (0.091)	Loss 0.562 (0.423)
Epoch: [44][100/200]	Time 0.661 (0.673)	Data 0.007 (0.087)	Loss 0.397 (0.441)
Epoch: [44][120/200]	Time 0.562 (0.670)	Data 0.000 (0.085)	Loss 0.307 (0.453)
Epoch: [44][140/200]	Time 0.669 (0.669)	Data 0.000 (0.083)	Loss 0.436 (0.457)
Epoch: [44][160/200]	Time 0.574 (0.668)	Data 0.000 (0.083)	Loss 0.681 (0.470)
Epoch: [44][180/200]	Time 0.560 (0.668)	Data 0.000 (0.082)	Loss 0.740 (0.477)
Epoch: [44][200/200]	Time 0.561 (0.675)	Data 0.001 (0.089)	Loss 0.582 (0.483)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.296 (0.244)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 20.150581121444702
==> Statistics for epoch 45: 600 clusters
Epoch: [45][20/200]	Time 0.575 (0.697)	Data 0.001 (0.110)	Loss 0.663 (0.137)
Epoch: [45][40/200]	Time 0.541 (0.689)	Data 0.001 (0.105)	Loss 0.459 (0.327)
Epoch: [45][60/200]	Time 0.575 (0.677)	Data 0.002 (0.092)	Loss 0.584 (0.401)
Epoch: [45][80/200]	Time 0.570 (0.671)	Data 0.001 (0.088)	Loss 0.409 (0.432)
Epoch: [45][100/200]	Time 0.566 (0.667)	Data 0.002 (0.083)	Loss 0.367 (0.444)
Epoch: [45][120/200]	Time 0.634 (0.664)	Data 0.000 (0.080)	Loss 0.542 (0.452)
Epoch: [45][140/200]	Time 0.567 (0.663)	Data 0.000 (0.080)	Loss 0.445 (0.462)
Epoch: [45][160/200]	Time 0.554 (0.662)	Data 0.000 (0.080)	Loss 0.405 (0.468)
Epoch: [45][180/200]	Time 0.557 (0.661)	Data 0.000 (0.079)	Loss 0.366 (0.476)
Epoch: [45][200/200]	Time 0.553 (0.665)	Data 0.001 (0.084)	Loss 0.765 (0.486)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.195 (0.253)	Data 0.000 (0.028)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.687437295913696
==> Statistics for epoch 46: 599 clusters
Epoch: [46][20/200]	Time 0.554 (0.708)	Data 0.001 (0.114)	Loss 0.392 (0.131)
Epoch: [46][40/200]	Time 0.562 (0.692)	Data 0.002 (0.097)	Loss 0.577 (0.326)
Epoch: [46][60/200]	Time 0.663 (0.682)	Data 0.001 (0.089)	Loss 0.693 (0.392)
Epoch: [46][80/200]	Time 0.569 (0.680)	Data 0.001 (0.087)	Loss 0.746 (0.429)
Epoch: [46][100/200]	Time 0.668 (0.675)	Data 0.001 (0.083)	Loss 0.463 (0.439)
Epoch: [46][120/200]	Time 0.545 (0.671)	Data 0.000 (0.081)	Loss 0.749 (0.461)
Epoch: [46][140/200]	Time 0.544 (0.670)	Data 0.000 (0.081)	Loss 0.511 (0.470)
Epoch: [46][160/200]	Time 0.570 (0.669)	Data 0.000 (0.079)	Loss 0.762 (0.483)
Epoch: [46][180/200]	Time 0.556 (0.668)	Data 0.000 (0.079)	Loss 0.565 (0.492)
Epoch: [46][200/200]	Time 0.544 (0.674)	Data 0.001 (0.085)	Loss 0.552 (0.502)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.289 (0.240)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.64004135131836
==> Statistics for epoch 47: 601 clusters
Epoch: [47][20/200]	Time 0.548 (0.687)	Data 0.001 (0.102)	Loss 0.511 (0.130)
Epoch: [47][40/200]	Time 0.650 (0.679)	Data 0.002 (0.093)	Loss 0.398 (0.333)
Epoch: [47][60/200]	Time 0.563 (0.667)	Data 0.001 (0.084)	Loss 0.532 (0.398)
Epoch: [47][80/200]	Time 0.558 (0.662)	Data 0.001 (0.081)	Loss 0.717 (0.440)
Epoch: [47][100/200]	Time 0.568 (0.662)	Data 0.001 (0.078)	Loss 0.626 (0.465)
Epoch: [47][120/200]	Time 0.563 (0.659)	Data 0.000 (0.076)	Loss 0.365 (0.472)
Epoch: [47][140/200]	Time 0.641 (0.659)	Data 0.000 (0.075)	Loss 0.781 (0.486)
Epoch: [47][160/200]	Time 0.563 (0.658)	Data 0.000 (0.074)	Loss 0.613 (0.492)
Epoch: [47][180/200]	Time 0.563 (0.658)	Data 0.000 (0.075)	Loss 0.379 (0.500)
Epoch: [47][200/200]	Time 0.549 (0.664)	Data 0.002 (0.081)	Loss 0.458 (0.500)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.202 (0.255)	Data 0.000 (0.039)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.807077169418335
==> Statistics for epoch 48: 601 clusters
Epoch: [48][20/200]	Time 0.550 (0.691)	Data 0.001 (0.105)	Loss 0.425 (0.136)
Epoch: [48][40/200]	Time 0.658 (0.668)	Data 0.002 (0.086)	Loss 0.659 (0.326)
Epoch: [48][60/200]	Time 0.559 (0.667)	Data 0.002 (0.085)	Loss 0.517 (0.401)
Epoch: [48][80/200]	Time 0.557 (0.661)	Data 0.005 (0.080)	Loss 0.451 (0.445)
Epoch: [48][100/200]	Time 0.547 (0.661)	Data 0.001 (0.079)	Loss 0.753 (0.462)
Epoch: [48][120/200]	Time 0.562 (0.659)	Data 0.000 (0.078)	Loss 0.504 (0.473)
Epoch: [48][140/200]	Time 0.681 (0.659)	Data 0.000 (0.077)	Loss 0.506 (0.479)
Epoch: [48][160/200]	Time 0.556 (0.657)	Data 0.000 (0.076)	Loss 0.529 (0.486)
Epoch: [48][180/200]	Time 0.636 (0.657)	Data 0.000 (0.075)	Loss 0.530 (0.492)
Epoch: [48][200/200]	Time 0.539 (0.663)	Data 0.001 (0.081)	Loss 0.647 (0.494)
==> Create pseudo labels for unlabeled data
Extract Features: [50/51]	Time 0.289 (0.240)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 19.838088512420654
==> Statistics for epoch 49: 598 clusters
Epoch: [49][20/200]	Time 0.656 (0.700)	Data 0.001 (0.104)	Loss 0.453 (0.122)
Epoch: [49][40/200]	Time 0.567 (0.680)	Data 0.002 (0.095)	Loss 0.547 (0.325)
Epoch: [49][60/200]	Time 0.552 (0.672)	Data 0.001 (0.086)	Loss 0.537 (0.388)
Epoch: [49][80/200]	Time 0.537 (0.667)	Data 0.001 (0.081)	Loss 0.628 (0.424)
Epoch: [49][100/200]	Time 0.562 (0.661)	Data 0.001 (0.078)	Loss 0.612 (0.451)
Epoch: [49][120/200]	Time 0.651 (0.660)	Data 0.000 (0.077)	Loss 0.485 (0.463)
Epoch: [49][140/200]	Time 0.557 (0.658)	Data 0.000 (0.076)	Loss 0.423 (0.479)
Epoch: [49][160/200]	Time 0.558 (0.656)	Data 0.000 (0.075)	Loss 0.396 (0.486)
Epoch: [49][180/200]	Time 0.635 (0.656)	Data 0.000 (0.074)	Loss 0.388 (0.489)
Epoch: [49][200/200]	Time 0.564 (0.661)	Data 0.001 (0.080)	Loss 0.365 (0.492)
Extract Features: [50/76]	Time 0.201 (0.245)	Data 0.000 (0.023)	
Mean AP: 94.3%

 * Finished epoch  49  model mAP: 94.3%  best: 94.3%

==> Test with the best model:
=> Loaded checkpoint 'log/market/resnet152_ibn_cion/model_best.pth.tar'
Extract Features: [50/76]	Time 0.194 (0.246)	Data 0.000 (0.026)	
Mean AP: 94.3%
CMC Scores:
  top-1          97.1%
  top-5          99.1%
  top-10         99.6%
Total running time:  2:24:09.341342
