==========
Args:Namespace(dataset='msmt17', batch_size=256, workers=4, height=256, width=128, num_instances=8, eps=0.7, eps_gap=0.02, k1=30, k2=6, arch='resnet_ibn101a', pretrained_path='/home/ma-user/work/Projects/ReIDNet_Finetune/TransReID/log/bs_exp/ResNet101_IBN_Market1501/64bs_lr0.0004_ep120_warm20_seed0/resnet101_ibn_120.pth', features=0, dropout=0, momentum=0.2, drop_path_rate=0.3, hw_ratio=2, self_norm=True, conv_stem=False, lr=0.00035, weight_decay=0.0005, optimizer='SGD', epochs=50, iters=200, step_size=20, seed=1, print_freq=20, eval_step=10, temp=0.05, data_dir='/home/ma-user/work/Projects/ReIDNet_Finetune/CContrast/data', logs_dir='log/market2msmt/resnet101_ibn_cion', pooling_type='gem', use_hard=True)
==========
==> Load unlabeled dataset
=> MSMT17 loaded
Dataset statistics:
  ----------------------------------------
  subset   | # ids | # images | # cameras
  ----------------------------------------
  train    |  1041 |    32621 |        15
  query    |  3060 |    11659 |        15
  gallery  |  3060 |    82161 |        15
  ----------------------------------------
pooling_type: gem
checkpoint loaded!
optimizer: SGD
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.402)	Data 0.000 (0.021)	
Extract Features: [100/128]	Time 0.184 (0.302)	Data 0.001 (0.011)	
Computing jaccard distance...
Jaccard distance computing time cost: 67.42895174026489
Clustering criterion: eps: 0.700
==> Statistics for epoch 0: 815 clusters
Epoch: [0][20/200]	Time 0.543 (1.012)	Data 0.000 (0.044)	Loss 3.111 (2.855)
Epoch: [0][40/200]	Time 0.547 (0.822)	Data 0.001 (0.062)	Loss 2.729 (2.840)
Epoch: [0][60/200]	Time 0.544 (0.757)	Data 0.001 (0.066)	Loss 1.761 (2.710)
Epoch: [0][80/200]	Time 0.547 (0.724)	Data 0.001 (0.069)	Loss 2.246 (2.559)
Epoch: [0][100/200]	Time 0.548 (0.690)	Data 0.000 (0.056)	Loss 1.534 (2.447)
Epoch: [0][120/200]	Time 0.545 (0.680)	Data 0.000 (0.059)	Loss 1.632 (2.353)
Epoch: [0][140/200]	Time 0.547 (0.673)	Data 0.001 (0.062)	Loss 1.782 (2.280)
Epoch: [0][160/200]	Time 0.550 (0.668)	Data 0.001 (0.064)	Loss 1.373 (2.225)
Epoch: [0][180/200]	Time 0.545 (0.663)	Data 0.001 (0.065)	Loss 2.073 (2.176)
Epoch: [0][200/200]	Time 0.545 (0.652)	Data 0.000 (0.059)	Loss 1.683 (2.127)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.183 (0.212)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.182 (0.202)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 65.36257886886597
==> Statistics for epoch 1: 997 clusters
Epoch: [1][20/200]	Time 0.548 (0.600)	Data 0.001 (0.054)	Loss 0.420 (0.419)
Epoch: [1][40/200]	Time 0.669 (0.613)	Data 0.001 (0.063)	Loss 1.539 (0.768)
Epoch: [1][60/200]	Time 0.547 (0.591)	Data 0.000 (0.042)	Loss 1.459 (1.089)
Epoch: [1][80/200]	Time 0.545 (0.599)	Data 0.001 (0.050)	Loss 1.564 (1.285)
Epoch: [1][100/200]	Time 0.686 (0.605)	Data 0.001 (0.055)	Loss 1.744 (1.382)
Epoch: [1][120/200]	Time 0.544 (0.596)	Data 0.000 (0.046)	Loss 1.902 (1.447)
Epoch: [1][140/200]	Time 0.551 (0.600)	Data 0.001 (0.051)	Loss 1.691 (1.492)
Epoch: [1][160/200]	Time 0.552 (0.605)	Data 0.001 (0.055)	Loss 1.294 (1.521)
Epoch: [1][180/200]	Time 0.641 (0.599)	Data 0.000 (0.049)	Loss 2.099 (1.554)
Epoch: [1][200/200]	Time 0.543 (0.602)	Data 0.001 (0.052)	Loss 1.528 (1.568)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.216)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.182 (0.203)	Data 0.000 (0.018)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.48057961463928
==> Statistics for epoch 2: 999 clusters
Epoch: [2][20/200]	Time 0.540 (0.589)	Data 0.001 (0.047)	Loss 0.220 (0.371)
Epoch: [2][40/200]	Time 0.541 (0.611)	Data 0.001 (0.065)	Loss 1.788 (0.660)
Epoch: [2][60/200]	Time 0.542 (0.591)	Data 0.000 (0.044)	Loss 1.715 (1.057)
Epoch: [2][80/200]	Time 0.546 (0.599)	Data 0.001 (0.052)	Loss 1.354 (1.248)
Epoch: [2][100/200]	Time 0.545 (0.607)	Data 0.001 (0.058)	Loss 1.825 (1.354)
Epoch: [2][120/200]	Time 0.550 (0.598)	Data 0.000 (0.048)	Loss 1.962 (1.407)
Epoch: [2][140/200]	Time 0.550 (0.602)	Data 0.001 (0.053)	Loss 1.805 (1.450)
Epoch: [2][160/200]	Time 0.546 (0.607)	Data 0.001 (0.057)	Loss 1.605 (1.494)
Epoch: [2][180/200]	Time 0.544 (0.601)	Data 0.000 (0.051)	Loss 1.738 (1.519)
Epoch: [2][200/200]	Time 0.543 (0.604)	Data 0.001 (0.054)	Loss 1.724 (1.542)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.182 (0.216)	Data 0.001 (0.027)	
Extract Features: [100/128]	Time 0.183 (0.203)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.13111090660095
==> Statistics for epoch 3: 1031 clusters
Epoch: [3][20/200]	Time 0.542 (0.602)	Data 0.001 (0.058)	Loss 0.445 (0.367)
Epoch: [3][40/200]	Time 0.545 (0.622)	Data 0.001 (0.072)	Loss 1.608 (0.617)
Epoch: [3][60/200]	Time 0.544 (0.599)	Data 0.000 (0.048)	Loss 1.735 (0.983)
Epoch: [3][80/200]	Time 0.544 (0.606)	Data 0.001 (0.057)	Loss 1.731 (1.195)
Epoch: [3][100/200]	Time 0.547 (0.613)	Data 0.001 (0.063)	Loss 1.814 (1.311)
Epoch: [3][120/200]	Time 0.548 (0.602)	Data 0.001 (0.053)	Loss 1.788 (1.388)
Epoch: [3][140/200]	Time 0.544 (0.607)	Data 0.001 (0.057)	Loss 1.402 (1.434)
Epoch: [3][160/200]	Time 0.544 (0.601)	Data 0.000 (0.050)	Loss 1.453 (1.471)
Epoch: [3][180/200]	Time 0.549 (0.604)	Data 0.001 (0.053)	Loss 2.418 (1.492)
Epoch: [3][200/200]	Time 0.547 (0.608)	Data 0.001 (0.057)	Loss 1.570 (1.512)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.183 (0.215)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.179 (0.201)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.48267221450806
==> Statistics for epoch 4: 1028 clusters
Epoch: [4][20/200]	Time 0.539 (0.600)	Data 0.001 (0.056)	Loss 0.203 (0.368)
Epoch: [4][40/200]	Time 0.546 (0.617)	Data 0.001 (0.070)	Loss 1.631 (0.604)
Epoch: [4][60/200]	Time 0.543 (0.595)	Data 0.000 (0.047)	Loss 1.914 (0.968)
Epoch: [4][80/200]	Time 0.547 (0.604)	Data 0.001 (0.055)	Loss 1.105 (1.130)
Epoch: [4][100/200]	Time 0.628 (0.611)	Data 0.001 (0.061)	Loss 1.644 (1.249)
Epoch: [4][120/200]	Time 0.544 (0.600)	Data 0.001 (0.051)	Loss 1.736 (1.331)
Epoch: [4][140/200]	Time 0.544 (0.606)	Data 0.001 (0.057)	Loss 1.731 (1.375)
Epoch: [4][160/200]	Time 0.547 (0.599)	Data 0.000 (0.050)	Loss 1.644 (1.406)
Epoch: [4][180/200]	Time 0.548 (0.603)	Data 0.002 (0.054)	Loss 1.835 (1.435)
Epoch: [4][200/200]	Time 0.547 (0.607)	Data 0.001 (0.057)	Loss 1.630 (1.467)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.212)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.193 (0.200)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.364800453186035
==> Statistics for epoch 5: 1024 clusters
Epoch: [5][20/200]	Time 0.545 (0.595)	Data 0.001 (0.049)	Loss 0.399 (0.325)
Epoch: [5][40/200]	Time 0.545 (0.612)	Data 0.001 (0.063)	Loss 1.403 (0.612)
Epoch: [5][60/200]	Time 0.544 (0.592)	Data 0.000 (0.042)	Loss 1.500 (0.944)
Epoch: [5][80/200]	Time 0.547 (0.603)	Data 0.001 (0.051)	Loss 2.050 (1.144)
Epoch: [5][100/200]	Time 0.546 (0.610)	Data 0.001 (0.057)	Loss 1.247 (1.239)
Epoch: [5][120/200]	Time 0.540 (0.600)	Data 0.001 (0.048)	Loss 1.942 (1.306)
Epoch: [5][140/200]	Time 0.549 (0.605)	Data 0.001 (0.053)	Loss 2.097 (1.358)
Epoch: [5][160/200]	Time 0.542 (0.598)	Data 0.000 (0.046)	Loss 1.943 (1.392)
Epoch: [5][180/200]	Time 0.548 (0.601)	Data 0.001 (0.049)	Loss 1.988 (1.415)
Epoch: [5][200/200]	Time 0.549 (0.604)	Data 0.001 (0.052)	Loss 1.835 (1.437)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.208 (0.212)	Data 0.028 (0.024)	
Extract Features: [100/128]	Time 0.185 (0.200)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.277127265930176
==> Statistics for epoch 6: 1002 clusters
Epoch: [6][20/200]	Time 0.550 (0.594)	Data 0.001 (0.049)	Loss 0.260 (0.332)
Epoch: [6][40/200]	Time 0.545 (0.611)	Data 0.001 (0.062)	Loss 1.514 (0.593)
Epoch: [6][60/200]	Time 0.546 (0.591)	Data 0.000 (0.042)	Loss 1.744 (0.915)
Epoch: [6][80/200]	Time 0.551 (0.600)	Data 0.001 (0.050)	Loss 1.343 (1.071)
Epoch: [6][100/200]	Time 0.547 (0.607)	Data 0.001 (0.058)	Loss 1.494 (1.182)
Epoch: [6][120/200]	Time 0.550 (0.597)	Data 0.000 (0.048)	Loss 1.164 (1.246)
Epoch: [6][140/200]	Time 0.551 (0.602)	Data 0.001 (0.052)	Loss 1.693 (1.303)
Epoch: [6][160/200]	Time 0.541 (0.605)	Data 0.001 (0.055)	Loss 1.384 (1.338)
Epoch: [6][180/200]	Time 0.545 (0.599)	Data 0.000 (0.049)	Loss 1.923 (1.360)
Epoch: [6][200/200]	Time 0.545 (0.602)	Data 0.001 (0.053)	Loss 1.246 (1.383)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.213)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.184 (0.199)	Data 0.001 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.185691118240356
==> Statistics for epoch 7: 1002 clusters
Epoch: [7][20/200]	Time 0.546 (0.605)	Data 0.001 (0.054)	Loss 0.223 (0.304)
Epoch: [7][40/200]	Time 0.545 (0.616)	Data 0.001 (0.068)	Loss 1.697 (0.560)
Epoch: [7][60/200]	Time 0.545 (0.592)	Data 0.000 (0.046)	Loss 1.405 (0.885)
Epoch: [7][80/200]	Time 0.548 (0.604)	Data 0.001 (0.056)	Loss 1.392 (1.028)
Epoch: [7][100/200]	Time 0.544 (0.609)	Data 0.001 (0.061)	Loss 1.599 (1.130)
Epoch: [7][120/200]	Time 0.547 (0.599)	Data 0.000 (0.051)	Loss 2.062 (1.196)
Epoch: [7][140/200]	Time 0.549 (0.605)	Data 0.001 (0.056)	Loss 1.608 (1.232)
Epoch: [7][160/200]	Time 0.548 (0.607)	Data 0.001 (0.059)	Loss 1.593 (1.273)
Epoch: [7][180/200]	Time 0.547 (0.601)	Data 0.000 (0.052)	Loss 1.526 (1.302)
Epoch: [7][200/200]	Time 0.548 (0.605)	Data 0.001 (0.055)	Loss 1.733 (1.322)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.191 (0.213)	Data 0.011 (0.027)	
Extract Features: [100/128]	Time 0.182 (0.201)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.92681813240051
==> Statistics for epoch 8: 1004 clusters
Epoch: [8][20/200]	Time 0.550 (0.597)	Data 0.001 (0.053)	Loss 0.180 (0.273)
Epoch: [8][40/200]	Time 0.551 (0.618)	Data 0.001 (0.070)	Loss 1.669 (0.524)
Epoch: [8][60/200]	Time 0.544 (0.597)	Data 0.000 (0.047)	Loss 1.713 (0.838)
Epoch: [8][80/200]	Time 0.542 (0.605)	Data 0.001 (0.056)	Loss 1.672 (0.995)
Epoch: [8][100/200]	Time 0.550 (0.612)	Data 0.001 (0.062)	Loss 1.304 (1.090)
Epoch: [8][120/200]	Time 0.546 (0.601)	Data 0.000 (0.052)	Loss 1.545 (1.155)
Epoch: [8][140/200]	Time 0.548 (0.607)	Data 0.001 (0.057)	Loss 1.640 (1.202)
Epoch: [8][160/200]	Time 0.548 (0.611)	Data 0.001 (0.060)	Loss 1.315 (1.216)
Epoch: [8][180/200]	Time 0.548 (0.604)	Data 0.000 (0.054)	Loss 1.198 (1.240)
Epoch: [8][200/200]	Time 0.545 (0.608)	Data 0.001 (0.057)	Loss 1.171 (1.260)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.219)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.183 (0.205)	Data 0.000 (0.018)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.03591537475586
==> Statistics for epoch 9: 1037 clusters
Epoch: [9][20/200]	Time 0.540 (0.593)	Data 0.001 (0.052)	Loss 0.241 (0.327)
Epoch: [9][40/200]	Time 0.543 (0.609)	Data 0.001 (0.062)	Loss 1.218 (0.510)
Epoch: [9][60/200]	Time 0.546 (0.587)	Data 0.000 (0.042)	Loss 1.568 (0.860)
Epoch: [9][80/200]	Time 0.544 (0.598)	Data 0.001 (0.051)	Loss 1.440 (1.004)
Epoch: [9][100/200]	Time 0.547 (0.605)	Data 0.001 (0.056)	Loss 1.764 (1.097)
Epoch: [9][120/200]	Time 0.548 (0.596)	Data 0.001 (0.047)	Loss 1.096 (1.147)
Epoch: [9][140/200]	Time 0.546 (0.601)	Data 0.001 (0.051)	Loss 1.329 (1.184)
Epoch: [9][160/200]	Time 0.701 (0.595)	Data 0.000 (0.045)	Loss 1.172 (1.216)
Epoch: [9][180/200]	Time 0.557 (0.599)	Data 0.001 (0.049)	Loss 1.136 (1.246)
Epoch: [9][200/200]	Time 0.545 (0.602)	Data 0.001 (0.052)	Loss 1.313 (1.259)
Extract Features: [50/367]	Time 0.182 (0.219)	Data 0.000 (0.029)	
Extract Features: [100/367]	Time 0.185 (0.203)	Data 0.000 (0.015)	
Extract Features: [150/367]	Time 0.180 (0.203)	Data 0.001 (0.010)	
Extract Features: [200/367]	Time 0.186 (0.201)	Data 0.005 (0.008)	
Extract Features: [250/367]	Time 0.550 (0.203)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.186 (0.203)	Data 0.001 (0.005)	
Extract Features: [350/367]	Time 0.187 (0.204)	Data 0.000 (0.005)	
Mean AP: 63.5%

 * Finished epoch   9  model mAP: 63.5%  best: 63.5% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.219)	Data 0.000 (0.035)	
Extract Features: [100/128]	Time 0.183 (0.204)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.740734577178955
==> Statistics for epoch 10: 1041 clusters
Epoch: [10][20/200]	Time 0.553 (0.608)	Data 0.001 (0.057)	Loss 0.178 (0.269)
Epoch: [10][40/200]	Time 0.688 (0.626)	Data 0.001 (0.074)	Loss 1.485 (0.481)
Epoch: [10][60/200]	Time 0.552 (0.599)	Data 0.000 (0.050)	Loss 1.514 (0.774)
Epoch: [10][80/200]	Time 0.547 (0.609)	Data 0.001 (0.061)	Loss 1.402 (0.936)
Epoch: [10][100/200]	Time 0.546 (0.615)	Data 0.001 (0.065)	Loss 1.329 (1.032)
Epoch: [10][120/200]	Time 0.699 (0.605)	Data 0.001 (0.055)	Loss 0.997 (1.088)
Epoch: [10][140/200]	Time 0.543 (0.610)	Data 0.001 (0.060)	Loss 1.671 (1.144)
Epoch: [10][160/200]	Time 0.544 (0.602)	Data 0.000 (0.053)	Loss 1.296 (1.164)
Epoch: [10][180/200]	Time 0.548 (0.607)	Data 0.001 (0.056)	Loss 0.994 (1.189)
Epoch: [10][200/200]	Time 0.544 (0.610)	Data 0.001 (0.060)	Loss 1.186 (1.200)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.337 (0.224)	Data 0.000 (0.035)	
Extract Features: [100/128]	Time 0.184 (0.208)	Data 0.001 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.46794104576111
==> Statistics for epoch 11: 1038 clusters
Epoch: [11][20/200]	Time 0.549 (0.597)	Data 0.001 (0.052)	Loss 0.174 (0.270)
Epoch: [11][40/200]	Time 0.546 (0.618)	Data 0.001 (0.069)	Loss 1.032 (0.453)
Epoch: [11][60/200]	Time 0.545 (0.594)	Data 0.000 (0.046)	Loss 1.447 (0.766)
Epoch: [11][80/200]	Time 0.550 (0.609)	Data 0.001 (0.059)	Loss 1.569 (0.927)
Epoch: [11][100/200]	Time 0.554 (0.615)	Data 0.001 (0.065)	Loss 1.633 (1.009)
Epoch: [11][120/200]	Time 0.547 (0.604)	Data 0.001 (0.054)	Loss 1.233 (1.069)
Epoch: [11][140/200]	Time 0.547 (0.608)	Data 0.001 (0.059)	Loss 1.454 (1.125)
Epoch: [11][160/200]	Time 0.543 (0.600)	Data 0.000 (0.052)	Loss 1.699 (1.156)
Epoch: [11][180/200]	Time 0.551 (0.605)	Data 0.001 (0.056)	Loss 1.294 (1.176)
Epoch: [11][200/200]	Time 0.547 (0.608)	Data 0.001 (0.059)	Loss 1.079 (1.190)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.183 (0.218)	Data 0.001 (0.033)	
Extract Features: [100/128]	Time 0.181 (0.205)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.540340423583984
==> Statistics for epoch 12: 1029 clusters
Epoch: [12][20/200]	Time 0.542 (0.598)	Data 0.001 (0.051)	Loss 0.199 (0.260)
Epoch: [12][40/200]	Time 0.541 (0.621)	Data 0.001 (0.071)	Loss 1.147 (0.467)
Epoch: [12][60/200]	Time 0.542 (0.597)	Data 0.000 (0.048)	Loss 1.430 (0.748)
Epoch: [12][80/200]	Time 0.544 (0.608)	Data 0.001 (0.060)	Loss 1.284 (0.889)
Epoch: [12][100/200]	Time 0.545 (0.616)	Data 0.001 (0.067)	Loss 0.932 (0.975)
Epoch: [12][120/200]	Time 0.551 (0.604)	Data 0.001 (0.056)	Loss 1.309 (1.037)
Epoch: [12][140/200]	Time 0.543 (0.610)	Data 0.001 (0.061)	Loss 0.963 (1.077)
Epoch: [12][160/200]	Time 0.549 (0.603)	Data 0.000 (0.053)	Loss 1.049 (1.111)
Epoch: [12][180/200]	Time 0.545 (0.607)	Data 0.001 (0.057)	Loss 1.297 (1.129)
Epoch: [12][200/200]	Time 0.545 (0.610)	Data 0.001 (0.061)	Loss 1.308 (1.143)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.216)	Data 0.000 (0.032)	
Extract Features: [100/128]	Time 0.178 (0.204)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.98228335380554
==> Statistics for epoch 13: 1028 clusters
Epoch: [13][20/200]	Time 0.542 (0.601)	Data 0.001 (0.051)	Loss 0.279 (0.255)
Epoch: [13][40/200]	Time 0.544 (0.620)	Data 0.001 (0.069)	Loss 1.408 (0.431)
Epoch: [13][60/200]	Time 0.546 (0.596)	Data 0.000 (0.046)	Loss 1.360 (0.722)
Epoch: [13][80/200]	Time 0.543 (0.608)	Data 0.001 (0.058)	Loss 1.259 (0.861)
Epoch: [13][100/200]	Time 0.546 (0.615)	Data 0.001 (0.065)	Loss 1.452 (0.950)
Epoch: [13][120/200]	Time 0.544 (0.603)	Data 0.001 (0.054)	Loss 1.582 (1.012)
Epoch: [13][140/200]	Time 0.545 (0.608)	Data 0.001 (0.059)	Loss 1.400 (1.059)
Epoch: [13][160/200]	Time 0.545 (0.601)	Data 0.000 (0.052)	Loss 1.070 (1.090)
Epoch: [13][180/200]	Time 0.548 (0.605)	Data 0.001 (0.056)	Loss 1.540 (1.115)
Epoch: [13][200/200]	Time 0.543 (0.608)	Data 0.001 (0.059)	Loss 1.290 (1.138)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.183 (0.218)	Data 0.000 (0.033)	
Extract Features: [100/128]	Time 0.182 (0.205)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.90402913093567
==> Statistics for epoch 14: 1022 clusters
Epoch: [14][20/200]	Time 0.549 (0.605)	Data 0.001 (0.056)	Loss 0.277 (0.252)
Epoch: [14][40/200]	Time 0.542 (0.617)	Data 0.001 (0.068)	Loss 1.156 (0.470)
Epoch: [14][60/200]	Time 0.543 (0.594)	Data 0.000 (0.045)	Loss 1.312 (0.751)
Epoch: [14][80/200]	Time 0.544 (0.605)	Data 0.001 (0.055)	Loss 1.259 (0.874)
Epoch: [14][100/200]	Time 0.545 (0.610)	Data 0.001 (0.062)	Loss 1.307 (0.960)
Epoch: [14][120/200]	Time 0.542 (0.600)	Data 0.000 (0.051)	Loss 1.156 (1.016)
Epoch: [14][140/200]	Time 0.541 (0.604)	Data 0.000 (0.056)	Loss 1.476 (1.056)
Epoch: [14][160/200]	Time 0.545 (0.608)	Data 0.001 (0.060)	Loss 1.394 (1.085)
Epoch: [14][180/200]	Time 0.544 (0.602)	Data 0.000 (0.053)	Loss 1.192 (1.105)
Epoch: [14][200/200]	Time 0.543 (0.605)	Data 0.001 (0.057)	Loss 1.099 (1.117)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.182 (0.217)	Data 0.000 (0.032)	
Extract Features: [100/128]	Time 0.179 (0.204)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.74158072471619
==> Statistics for epoch 15: 1039 clusters
Epoch: [15][20/200]	Time 0.542 (0.605)	Data 0.001 (0.053)	Loss 0.188 (0.232)
Epoch: [15][40/200]	Time 0.547 (0.623)	Data 0.001 (0.071)	Loss 1.475 (0.429)
Epoch: [15][60/200]	Time 0.545 (0.597)	Data 0.000 (0.048)	Loss 1.056 (0.694)
Epoch: [15][80/200]	Time 0.543 (0.609)	Data 0.001 (0.059)	Loss 1.078 (0.823)
Epoch: [15][100/200]	Time 0.550 (0.615)	Data 0.001 (0.065)	Loss 1.209 (0.905)
Epoch: [15][120/200]	Time 0.546 (0.605)	Data 0.001 (0.054)	Loss 0.963 (0.959)
Epoch: [15][140/200]	Time 0.545 (0.610)	Data 0.001 (0.059)	Loss 1.553 (0.997)
Epoch: [15][160/200]	Time 0.544 (0.603)	Data 0.000 (0.052)	Loss 0.996 (1.036)
Epoch: [15][180/200]	Time 0.543 (0.607)	Data 0.001 (0.056)	Loss 1.330 (1.066)
Epoch: [15][200/200]	Time 0.546 (0.610)	Data 0.001 (0.059)	Loss 1.493 (1.080)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.177 (0.214)	Data 0.000 (0.027)	
Extract Features: [100/128]	Time 0.179 (0.200)	Data 0.000 (0.014)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.57578682899475
==> Statistics for epoch 16: 1047 clusters
Epoch: [16][20/200]	Time 0.544 (0.599)	Data 0.001 (0.055)	Loss 0.166 (0.249)
Epoch: [16][40/200]	Time 0.540 (0.619)	Data 0.001 (0.072)	Loss 1.156 (0.432)
Epoch: [16][60/200]	Time 0.542 (0.595)	Data 0.000 (0.049)	Loss 1.210 (0.702)
Epoch: [16][80/200]	Time 0.543 (0.607)	Data 0.001 (0.059)	Loss 1.421 (0.842)
Epoch: [16][100/200]	Time 0.650 (0.614)	Data 0.001 (0.066)	Loss 1.353 (0.930)
Epoch: [16][120/200]	Time 0.547 (0.602)	Data 0.001 (0.055)	Loss 1.090 (0.990)
Epoch: [16][140/200]	Time 0.545 (0.607)	Data 0.001 (0.059)	Loss 1.231 (1.032)
Epoch: [16][160/200]	Time 0.543 (0.599)	Data 0.000 (0.052)	Loss 1.216 (1.059)
Epoch: [16][180/200]	Time 0.540 (0.604)	Data 0.001 (0.056)	Loss 1.354 (1.079)
Epoch: [16][200/200]	Time 0.549 (0.607)	Data 0.001 (0.059)	Loss 1.534 (1.092)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.219)	Data 0.000 (0.033)	
Extract Features: [100/128]	Time 0.180 (0.205)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.152377128601074
==> Statistics for epoch 17: 1033 clusters
Epoch: [17][20/200]	Time 0.541 (0.598)	Data 0.001 (0.054)	Loss 0.226 (0.245)
Epoch: [17][40/200]	Time 0.542 (0.621)	Data 0.001 (0.074)	Loss 1.221 (0.420)
Epoch: [17][60/200]	Time 0.537 (0.598)	Data 0.000 (0.050)	Loss 1.327 (0.685)
Epoch: [17][80/200]	Time 0.547 (0.610)	Data 0.001 (0.061)	Loss 1.137 (0.813)
Epoch: [17][100/200]	Time 0.543 (0.617)	Data 0.001 (0.068)	Loss 1.050 (0.881)
Epoch: [17][120/200]	Time 0.546 (0.606)	Data 0.001 (0.057)	Loss 1.312 (0.949)
Epoch: [17][140/200]	Time 0.546 (0.611)	Data 0.001 (0.061)	Loss 1.211 (0.994)
Epoch: [17][160/200]	Time 0.543 (0.603)	Data 0.000 (0.053)	Loss 1.126 (1.019)
Epoch: [17][180/200]	Time 0.547 (0.607)	Data 0.000 (0.057)	Loss 1.459 (1.052)
Epoch: [17][200/200]	Time 0.544 (0.610)	Data 0.001 (0.060)	Loss 0.720 (1.066)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.226)	Data 0.000 (0.039)	
Extract Features: [100/128]	Time 0.181 (0.209)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.80254030227661
==> Statistics for epoch 18: 1057 clusters
Epoch: [18][20/200]	Time 0.544 (0.596)	Data 0.001 (0.053)	Loss 0.256 (0.232)
Epoch: [18][40/200]	Time 0.544 (0.619)	Data 0.001 (0.072)	Loss 1.026 (0.349)
Epoch: [18][60/200]	Time 0.548 (0.596)	Data 0.000 (0.048)	Loss 1.284 (0.655)
Epoch: [18][80/200]	Time 0.543 (0.606)	Data 0.001 (0.058)	Loss 1.114 (0.793)
Epoch: [18][100/200]	Time 2.436 (0.614)	Data 1.864 (0.065)	Loss 1.708 (0.866)
Epoch: [18][120/200]	Time 0.544 (0.604)	Data 0.001 (0.054)	Loss 1.055 (0.928)
Epoch: [18][140/200]	Time 0.547 (0.610)	Data 0.001 (0.060)	Loss 0.780 (0.963)
Epoch: [18][160/200]	Time 0.550 (0.602)	Data 0.000 (0.053)	Loss 1.132 (0.986)
Epoch: [18][180/200]	Time 0.546 (0.608)	Data 0.001 (0.058)	Loss 1.009 (1.008)
Epoch: [18][200/200]	Time 0.549 (0.612)	Data 0.001 (0.062)	Loss 1.230 (1.032)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.226)	Data 0.000 (0.041)	
Extract Features: [100/128]	Time 0.182 (0.209)	Data 0.000 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.77721405029297
==> Statistics for epoch 19: 1049 clusters
Epoch: [19][20/200]	Time 0.546 (0.610)	Data 0.001 (0.058)	Loss 0.272 (0.216)
Epoch: [19][40/200]	Time 0.544 (0.622)	Data 0.001 (0.071)	Loss 1.800 (0.413)
Epoch: [19][60/200]	Time 0.543 (0.598)	Data 0.000 (0.047)	Loss 1.237 (0.655)
Epoch: [19][80/200]	Time 0.545 (0.609)	Data 0.001 (0.059)	Loss 1.320 (0.788)
Epoch: [19][100/200]	Time 0.548 (0.616)	Data 0.001 (0.065)	Loss 1.204 (0.881)
Epoch: [19][120/200]	Time 0.551 (0.605)	Data 0.001 (0.054)	Loss 1.228 (0.935)
Epoch: [19][140/200]	Time 0.549 (0.610)	Data 0.001 (0.059)	Loss 1.517 (0.978)
Epoch: [19][160/200]	Time 0.545 (0.602)	Data 0.000 (0.051)	Loss 1.082 (1.006)
Epoch: [19][180/200]	Time 0.552 (0.606)	Data 0.001 (0.055)	Loss 1.418 (1.036)
Epoch: [19][200/200]	Time 0.544 (0.609)	Data 0.000 (0.059)	Loss 1.141 (1.057)
Extract Features: [50/367]	Time 0.182 (0.217)	Data 0.000 (0.030)	
Extract Features: [100/367]	Time 0.186 (0.202)	Data 0.000 (0.015)	
Extract Features: [150/367]	Time 0.181 (0.198)	Data 0.000 (0.010)	
Extract Features: [200/367]	Time 0.185 (0.195)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.185 (0.193)	Data 0.000 (0.006)	
Extract Features: [300/367]	Time 0.189 (0.192)	Data 0.000 (0.005)	
Extract Features: [350/367]	Time 0.186 (0.191)	Data 0.000 (0.005)	
Mean AP: 67.4%

 * Finished epoch  19  model mAP: 67.4%  best: 67.4% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.212)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.181 (0.201)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.219696283340454
==> Statistics for epoch 20: 1057 clusters
Epoch: [20][20/200]	Time 0.544 (0.605)	Data 0.000 (0.056)	Loss 0.191 (0.212)
Epoch: [20][40/200]	Time 0.545 (0.619)	Data 0.001 (0.068)	Loss 1.328 (0.360)
Epoch: [20][60/200]	Time 0.545 (0.594)	Data 0.000 (0.046)	Loss 1.096 (0.642)
Epoch: [20][80/200]	Time 0.543 (0.606)	Data 0.001 (0.057)	Loss 1.266 (0.761)
Epoch: [20][100/200]	Time 2.516 (0.615)	Data 1.956 (0.065)	Loss 1.491 (0.840)
Epoch: [20][120/200]	Time 0.545 (0.603)	Data 0.001 (0.054)	Loss 1.577 (0.897)
Epoch: [20][140/200]	Time 0.544 (0.608)	Data 0.000 (0.058)	Loss 1.089 (0.934)
Epoch: [20][160/200]	Time 0.546 (0.601)	Data 0.000 (0.051)	Loss 1.171 (0.959)
Epoch: [20][180/200]	Time 0.543 (0.606)	Data 0.001 (0.056)	Loss 1.358 (0.982)
Epoch: [20][200/200]	Time 0.551 (0.610)	Data 0.000 (0.059)	Loss 0.818 (0.992)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.220)	Data 0.000 (0.034)	
Extract Features: [100/128]	Time 0.181 (0.207)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.933188915252686
==> Statistics for epoch 21: 1074 clusters
Epoch: [21][20/200]	Time 0.546 (0.603)	Data 0.001 (0.050)	Loss 0.171 (0.196)
Epoch: [21][40/200]	Time 0.545 (0.621)	Data 0.001 (0.072)	Loss 1.147 (0.374)
Epoch: [21][60/200]	Time 0.543 (0.597)	Data 0.000 (0.048)	Loss 1.037 (0.614)
Epoch: [21][80/200]	Time 0.544 (0.607)	Data 0.001 (0.059)	Loss 1.324 (0.744)
Epoch: [21][100/200]	Time 2.394 (0.614)	Data 1.821 (0.066)	Loss 0.926 (0.817)
Epoch: [21][120/200]	Time 0.545 (0.603)	Data 0.000 (0.055)	Loss 1.326 (0.886)
Epoch: [21][140/200]	Time 0.547 (0.608)	Data 0.000 (0.058)	Loss 1.562 (0.933)
Epoch: [21][160/200]	Time 0.547 (0.600)	Data 0.000 (0.051)	Loss 1.010 (0.955)
Epoch: [21][180/200]	Time 0.547 (0.605)	Data 0.001 (0.055)	Loss 1.170 (0.985)
Epoch: [21][200/200]	Time 0.554 (0.609)	Data 0.001 (0.060)	Loss 1.306 (1.014)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.183 (0.220)	Data 0.000 (0.032)	
Extract Features: [100/128]	Time 0.185 (0.206)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.32213425636292
==> Statistics for epoch 22: 1075 clusters
Epoch: [22][20/200]	Time 0.545 (0.603)	Data 0.001 (0.051)	Loss 0.135 (0.206)
Epoch: [22][40/200]	Time 0.544 (0.621)	Data 0.001 (0.073)	Loss 1.426 (0.352)
Epoch: [22][60/200]	Time 0.544 (0.598)	Data 0.000 (0.049)	Loss 1.021 (0.619)
Epoch: [22][80/200]	Time 0.544 (0.612)	Data 0.001 (0.061)	Loss 0.899 (0.747)
Epoch: [22][100/200]	Time 2.389 (0.617)	Data 1.815 (0.067)	Loss 1.346 (0.830)
Epoch: [22][120/200]	Time 0.546 (0.607)	Data 0.001 (0.056)	Loss 1.738 (0.892)
Epoch: [22][140/200]	Time 0.548 (0.612)	Data 0.001 (0.062)	Loss 1.029 (0.933)
Epoch: [22][160/200]	Time 0.544 (0.604)	Data 0.000 (0.054)	Loss 1.057 (0.966)
Epoch: [22][180/200]	Time 0.547 (0.608)	Data 0.001 (0.058)	Loss 1.356 (0.992)
Epoch: [22][200/200]	Time 0.551 (0.613)	Data 0.001 (0.062)	Loss 0.995 (1.010)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.182 (0.222)	Data 0.000 (0.038)	
Extract Features: [100/128]	Time 0.187 (0.208)	Data 0.001 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.171600580215454
==> Statistics for epoch 23: 1080 clusters
Epoch: [23][20/200]	Time 0.543 (0.609)	Data 0.001 (0.056)	Loss 0.141 (0.198)
Epoch: [23][40/200]	Time 0.547 (0.624)	Data 0.001 (0.074)	Loss 1.186 (0.358)
Epoch: [23][60/200]	Time 0.541 (0.600)	Data 0.000 (0.050)	Loss 1.220 (0.595)
Epoch: [23][80/200]	Time 0.547 (0.611)	Data 0.001 (0.060)	Loss 0.960 (0.718)
Epoch: [23][100/200]	Time 2.490 (0.617)	Data 1.779 (0.066)	Loss 1.070 (0.798)
Epoch: [23][120/200]	Time 0.549 (0.606)	Data 0.001 (0.055)	Loss 1.099 (0.866)
Epoch: [23][140/200]	Time 0.548 (0.611)	Data 0.001 (0.060)	Loss 0.981 (0.902)
Epoch: [23][160/200]	Time 0.549 (0.603)	Data 0.000 (0.052)	Loss 1.087 (0.934)
Epoch: [23][180/200]	Time 0.545 (0.608)	Data 0.000 (0.056)	Loss 1.133 (0.952)
Epoch: [23][200/200]	Time 0.553 (0.612)	Data 0.001 (0.060)	Loss 1.148 (0.973)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.186 (0.215)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.181 (0.205)	Data 0.000 (0.018)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.25883436203003
==> Statistics for epoch 24: 1061 clusters
Epoch: [24][20/200]	Time 0.539 (0.609)	Data 0.001 (0.059)	Loss 0.276 (0.184)
Epoch: [24][40/200]	Time 0.547 (0.624)	Data 0.001 (0.077)	Loss 1.031 (0.333)
Epoch: [24][60/200]	Time 0.543 (0.599)	Data 0.000 (0.051)	Loss 1.240 (0.606)
Epoch: [24][80/200]	Time 0.545 (0.609)	Data 0.001 (0.062)	Loss 1.152 (0.726)
Epoch: [24][100/200]	Time 2.515 (0.616)	Data 1.952 (0.069)	Loss 1.391 (0.820)
Epoch: [24][120/200]	Time 0.545 (0.606)	Data 0.001 (0.058)	Loss 1.486 (0.869)
Epoch: [24][140/200]	Time 0.549 (0.612)	Data 0.001 (0.063)	Loss 0.926 (0.909)
Epoch: [24][160/200]	Time 0.548 (0.604)	Data 0.000 (0.055)	Loss 1.216 (0.937)
Epoch: [24][180/200]	Time 0.544 (0.609)	Data 0.001 (0.060)	Loss 1.233 (0.960)
Epoch: [24][200/200]	Time 0.545 (0.611)	Data 0.001 (0.062)	Loss 1.132 (0.983)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.215)	Data 0.000 (0.026)	
Extract Features: [100/128]	Time 0.183 (0.200)	Data 0.000 (0.013)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.82673931121826
==> Statistics for epoch 25: 1068 clusters
Epoch: [25][20/200]	Time 0.543 (0.599)	Data 0.001 (0.054)	Loss 0.157 (0.186)
Epoch: [25][40/200]	Time 0.550 (0.618)	Data 0.001 (0.070)	Loss 1.455 (0.360)
Epoch: [25][60/200]	Time 0.543 (0.593)	Data 0.000 (0.047)	Loss 1.313 (0.612)
Epoch: [25][80/200]	Time 0.543 (0.603)	Data 0.001 (0.056)	Loss 0.769 (0.741)
Epoch: [25][100/200]	Time 2.329 (0.610)	Data 1.769 (0.063)	Loss 1.317 (0.811)
Epoch: [25][120/200]	Time 0.545 (0.600)	Data 0.001 (0.052)	Loss 0.904 (0.863)
Epoch: [25][140/200]	Time 0.545 (0.607)	Data 0.001 (0.058)	Loss 1.192 (0.910)
Epoch: [25][160/200]	Time 0.670 (0.600)	Data 0.000 (0.051)	Loss 1.139 (0.941)
Epoch: [25][180/200]	Time 0.544 (0.604)	Data 0.001 (0.055)	Loss 1.330 (0.960)
Epoch: [25][200/200]	Time 0.546 (0.608)	Data 0.001 (0.059)	Loss 1.294 (0.979)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.190 (0.219)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.182 (0.204)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.841158866882324
==> Statistics for epoch 26: 1070 clusters
Epoch: [26][20/200]	Time 0.539 (0.597)	Data 0.001 (0.054)	Loss 0.120 (0.184)
Epoch: [26][40/200]	Time 0.544 (0.613)	Data 0.001 (0.069)	Loss 1.283 (0.340)
Epoch: [26][60/200]	Time 0.550 (0.592)	Data 0.001 (0.046)	Loss 0.743 (0.594)
Epoch: [26][80/200]	Time 0.547 (0.605)	Data 0.001 (0.057)	Loss 0.876 (0.718)
Epoch: [26][100/200]	Time 2.402 (0.611)	Data 1.836 (0.064)	Loss 1.292 (0.796)
Epoch: [26][120/200]	Time 0.544 (0.602)	Data 0.001 (0.053)	Loss 1.312 (0.860)
Epoch: [26][140/200]	Time 0.544 (0.608)	Data 0.000 (0.058)	Loss 1.136 (0.889)
Epoch: [26][160/200]	Time 0.545 (0.600)	Data 0.000 (0.051)	Loss 0.941 (0.918)
Epoch: [26][180/200]	Time 0.550 (0.605)	Data 0.001 (0.055)	Loss 1.035 (0.942)
Epoch: [26][200/200]	Time 0.546 (0.608)	Data 0.000 (0.058)	Loss 1.025 (0.959)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.191 (0.219)	Data 0.000 (0.034)	
Extract Features: [100/128]	Time 0.182 (0.207)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.61585831642151
==> Statistics for epoch 27: 1075 clusters
Epoch: [27][20/200]	Time 0.543 (0.607)	Data 0.000 (0.055)	Loss 0.179 (0.166)
Epoch: [27][40/200]	Time 0.546 (0.618)	Data 0.001 (0.068)	Loss 1.369 (0.338)
Epoch: [27][60/200]	Time 0.542 (0.594)	Data 0.000 (0.046)	Loss 0.986 (0.605)
Epoch: [27][80/200]	Time 0.541 (0.607)	Data 0.001 (0.059)	Loss 1.363 (0.726)
Epoch: [27][100/200]	Time 2.254 (0.613)	Data 1.662 (0.064)	Loss 1.016 (0.798)
Epoch: [27][120/200]	Time 0.549 (0.602)	Data 0.001 (0.053)	Loss 1.100 (0.869)
Epoch: [27][140/200]	Time 0.545 (0.608)	Data 0.001 (0.059)	Loss 1.130 (0.901)
Epoch: [27][160/200]	Time 0.546 (0.601)	Data 0.000 (0.051)	Loss 1.069 (0.932)
Epoch: [27][180/200]	Time 0.543 (0.605)	Data 0.000 (0.055)	Loss 1.025 (0.942)
Epoch: [27][200/200]	Time 0.547 (0.608)	Data 0.000 (0.058)	Loss 1.073 (0.961)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.218)	Data 0.000 (0.034)	
Extract Features: [100/128]	Time 0.187 (0.204)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.29239416122437
==> Statistics for epoch 28: 1076 clusters
Epoch: [28][20/200]	Time 0.548 (0.607)	Data 0.001 (0.054)	Loss 0.145 (0.200)
Epoch: [28][40/200]	Time 0.543 (0.620)	Data 0.001 (0.071)	Loss 0.896 (0.341)
Epoch: [28][60/200]	Time 0.546 (0.598)	Data 0.000 (0.048)	Loss 1.206 (0.586)
Epoch: [28][80/200]	Time 0.544 (0.606)	Data 0.001 (0.056)	Loss 1.020 (0.738)
Epoch: [28][100/200]	Time 2.239 (0.611)	Data 1.668 (0.062)	Loss 1.366 (0.821)
Epoch: [28][120/200]	Time 0.544 (0.601)	Data 0.000 (0.051)	Loss 1.301 (0.879)
Epoch: [28][140/200]	Time 0.548 (0.606)	Data 0.001 (0.056)	Loss 1.384 (0.911)
Epoch: [28][160/200]	Time 0.544 (0.599)	Data 0.000 (0.049)	Loss 0.995 (0.939)
Epoch: [28][180/200]	Time 0.549 (0.603)	Data 0.001 (0.054)	Loss 0.650 (0.963)
Epoch: [28][200/200]	Time 0.551 (0.608)	Data 0.001 (0.058)	Loss 0.945 (0.980)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.215)	Data 0.001 (0.029)	
Extract Features: [100/128]	Time 0.188 (0.202)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.07378816604614
==> Statistics for epoch 29: 1067 clusters
Epoch: [29][20/200]	Time 0.660 (0.604)	Data 0.001 (0.052)	Loss 0.205 (0.185)
Epoch: [29][40/200]	Time 0.546 (0.616)	Data 0.001 (0.069)	Loss 1.067 (0.346)
Epoch: [29][60/200]	Time 0.549 (0.595)	Data 0.000 (0.046)	Loss 1.084 (0.605)
Epoch: [29][80/200]	Time 0.548 (0.608)	Data 0.001 (0.057)	Loss 1.101 (0.738)
Epoch: [29][100/200]	Time 2.258 (0.613)	Data 1.683 (0.063)	Loss 1.264 (0.812)
Epoch: [29][120/200]	Time 0.548 (0.602)	Data 0.001 (0.053)	Loss 1.044 (0.858)
Epoch: [29][140/200]	Time 0.545 (0.606)	Data 0.001 (0.057)	Loss 1.323 (0.895)
Epoch: [29][160/200]	Time 0.550 (0.598)	Data 0.000 (0.050)	Loss 1.254 (0.929)
Epoch: [29][180/200]	Time 0.552 (0.604)	Data 0.001 (0.054)	Loss 1.120 (0.955)
Epoch: [29][200/200]	Time 0.551 (0.608)	Data 0.001 (0.058)	Loss 1.262 (0.971)
Extract Features: [50/367]	Time 0.180 (0.214)	Data 0.000 (0.029)	
Extract Features: [100/367]	Time 0.186 (0.204)	Data 0.000 (0.017)	
Extract Features: [150/367]	Time 0.188 (0.199)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.183 (0.196)	Data 0.000 (0.009)	
Extract Features: [250/367]	Time 0.181 (0.194)	Data 0.001 (0.007)	
Extract Features: [300/367]	Time 0.185 (0.193)	Data 0.000 (0.006)	
Extract Features: [350/367]	Time 0.185 (0.192)	Data 0.000 (0.005)	
Mean AP: 70.1%

 * Finished epoch  29  model mAP: 70.1%  best: 70.1% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.219)	Data 0.000 (0.038)	
Extract Features: [100/128]	Time 0.349 (0.206)	Data 0.001 (0.023)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.5770161151886
==> Statistics for epoch 30: 1073 clusters
Epoch: [30][20/200]	Time 0.539 (0.606)	Data 0.001 (0.055)	Loss 0.130 (0.201)
Epoch: [30][40/200]	Time 0.540 (0.628)	Data 0.001 (0.077)	Loss 1.280 (0.368)
Epoch: [30][60/200]	Time 0.539 (0.599)	Data 0.000 (0.052)	Loss 1.079 (0.609)
Epoch: [30][80/200]	Time 0.547 (0.611)	Data 0.001 (0.063)	Loss 1.091 (0.738)
Epoch: [30][100/200]	Time 2.348 (0.616)	Data 1.776 (0.068)	Loss 1.271 (0.821)
Epoch: [30][120/200]	Time 0.544 (0.605)	Data 0.001 (0.057)	Loss 1.178 (0.882)
Epoch: [30][140/200]	Time 0.544 (0.610)	Data 0.001 (0.062)	Loss 1.154 (0.915)
Epoch: [30][160/200]	Time 0.543 (0.603)	Data 0.000 (0.054)	Loss 1.678 (0.940)
Epoch: [30][180/200]	Time 0.545 (0.606)	Data 0.001 (0.058)	Loss 1.115 (0.965)
Epoch: [30][200/200]	Time 0.560 (0.610)	Data 0.001 (0.062)	Loss 1.112 (0.981)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.221)	Data 0.001 (0.032)	
Extract Features: [100/128]	Time 0.181 (0.204)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 62.386916637420654
==> Statistics for epoch 31: 1058 clusters
Epoch: [31][20/200]	Time 0.566 (0.597)	Data 0.001 (0.053)	Loss 0.193 (0.170)
Epoch: [31][40/200]	Time 0.542 (0.614)	Data 0.001 (0.067)	Loss 1.341 (0.350)
Epoch: [31][60/200]	Time 0.540 (0.590)	Data 0.000 (0.045)	Loss 1.463 (0.592)
Epoch: [31][80/200]	Time 0.542 (0.601)	Data 0.001 (0.054)	Loss 1.060 (0.718)
Epoch: [31][100/200]	Time 2.521 (0.610)	Data 1.955 (0.063)	Loss 1.133 (0.784)
Epoch: [31][120/200]	Time 0.543 (0.601)	Data 0.001 (0.053)	Loss 1.014 (0.830)
Epoch: [31][140/200]	Time 0.544 (0.606)	Data 0.001 (0.058)	Loss 1.667 (0.878)
Epoch: [31][160/200]	Time 0.541 (0.599)	Data 0.000 (0.050)	Loss 1.177 (0.902)
Epoch: [31][180/200]	Time 0.548 (0.602)	Data 0.001 (0.054)	Loss 1.311 (0.921)
Epoch: [31][200/200]	Time 0.553 (0.605)	Data 0.001 (0.057)	Loss 1.675 (0.944)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.225)	Data 0.000 (0.038)	
Extract Features: [100/128]	Time 0.180 (0.207)	Data 0.000 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 64.30483388900757
==> Statistics for epoch 32: 1063 clusters
Epoch: [32][20/200]	Time 0.541 (0.592)	Data 0.001 (0.050)	Loss 0.191 (0.207)
Epoch: [32][40/200]	Time 0.545 (0.618)	Data 0.001 (0.072)	Loss 1.344 (0.356)
Epoch: [32][60/200]	Time 0.544 (0.595)	Data 0.000 (0.049)	Loss 1.177 (0.608)
Epoch: [32][80/200]	Time 0.540 (0.606)	Data 0.001 (0.058)	Loss 1.012 (0.723)
Epoch: [32][100/200]	Time 2.317 (0.612)	Data 1.745 (0.064)	Loss 1.223 (0.811)
Epoch: [32][120/200]	Time 0.544 (0.601)	Data 0.001 (0.054)	Loss 1.108 (0.865)
Epoch: [32][140/200]	Time 0.543 (0.607)	Data 0.001 (0.059)	Loss 0.665 (0.896)
Epoch: [32][160/200]	Time 0.542 (0.599)	Data 0.000 (0.051)	Loss 1.406 (0.926)
Epoch: [32][180/200]	Time 0.547 (0.604)	Data 0.001 (0.056)	Loss 1.113 (0.950)
Epoch: [32][200/200]	Time 0.544 (0.607)	Data 0.001 (0.059)	Loss 1.353 (0.972)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.231 (0.217)	Data 0.047 (0.029)	
Extract Features: [100/128]	Time 0.181 (0.204)	Data 0.000 (0.018)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.973477840423584
==> Statistics for epoch 33: 1063 clusters
Epoch: [33][20/200]	Time 0.539 (0.595)	Data 0.001 (0.052)	Loss 0.165 (0.174)
Epoch: [33][40/200]	Time 0.541 (0.617)	Data 0.001 (0.071)	Loss 1.223 (0.337)
Epoch: [33][60/200]	Time 0.541 (0.593)	Data 0.000 (0.047)	Loss 1.167 (0.583)
Epoch: [33][80/200]	Time 0.544 (0.606)	Data 0.001 (0.060)	Loss 0.824 (0.714)
Epoch: [33][100/200]	Time 2.413 (0.614)	Data 1.835 (0.066)	Loss 0.998 (0.788)
Epoch: [33][120/200]	Time 0.544 (0.603)	Data 0.001 (0.055)	Loss 1.192 (0.839)
Epoch: [33][140/200]	Time 0.550 (0.608)	Data 0.001 (0.060)	Loss 0.955 (0.872)
Epoch: [33][160/200]	Time 0.545 (0.601)	Data 0.000 (0.052)	Loss 1.255 (0.895)
Epoch: [33][180/200]	Time 0.547 (0.604)	Data 0.001 (0.056)	Loss 1.327 (0.921)
Epoch: [33][200/200]	Time 0.553 (0.610)	Data 0.001 (0.060)	Loss 1.874 (0.944)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.189 (0.222)	Data 0.000 (0.037)	
Extract Features: [100/128]	Time 0.186 (0.208)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.40938591957092
==> Statistics for epoch 34: 1064 clusters
Epoch: [34][20/200]	Time 0.542 (0.607)	Data 0.001 (0.056)	Loss 0.235 (0.181)
Epoch: [34][40/200]	Time 0.547 (0.617)	Data 0.001 (0.069)	Loss 0.836 (0.328)
Epoch: [34][60/200]	Time 0.543 (0.593)	Data 0.000 (0.047)	Loss 1.175 (0.577)
Epoch: [34][80/200]	Time 0.544 (0.607)	Data 0.001 (0.059)	Loss 0.976 (0.701)
Epoch: [34][100/200]	Time 2.448 (0.615)	Data 1.867 (0.066)	Loss 1.114 (0.784)
Epoch: [34][120/200]	Time 0.544 (0.603)	Data 0.001 (0.055)	Loss 1.074 (0.849)
Epoch: [34][140/200]	Time 0.544 (0.610)	Data 0.001 (0.061)	Loss 1.136 (0.877)
Epoch: [34][160/200]	Time 0.546 (0.602)	Data 0.000 (0.054)	Loss 1.398 (0.914)
Epoch: [34][180/200]	Time 0.544 (0.607)	Data 0.001 (0.058)	Loss 1.305 (0.933)
Epoch: [34][200/200]	Time 0.548 (0.611)	Data 0.001 (0.062)	Loss 1.092 (0.951)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.223)	Data 0.000 (0.037)	
Extract Features: [100/128]	Time 0.182 (0.208)	Data 0.000 (0.021)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.920719146728516
==> Statistics for epoch 35: 1063 clusters
Epoch: [35][20/200]	Time 0.542 (0.596)	Data 0.001 (0.054)	Loss 0.148 (0.186)
Epoch: [35][40/200]	Time 0.643 (0.614)	Data 0.001 (0.069)	Loss 0.909 (0.331)
Epoch: [35][60/200]	Time 0.546 (0.591)	Data 0.000 (0.047)	Loss 1.163 (0.574)
Epoch: [35][80/200]	Time 0.549 (0.605)	Data 0.002 (0.058)	Loss 1.451 (0.702)
Epoch: [35][100/200]	Time 2.197 (0.609)	Data 1.630 (0.063)	Loss 1.041 (0.781)
Epoch: [35][120/200]	Time 0.546 (0.599)	Data 0.001 (0.053)	Loss 1.113 (0.843)
Epoch: [35][140/200]	Time 0.544 (0.604)	Data 0.001 (0.058)	Loss 1.402 (0.878)
Epoch: [35][160/200]	Time 0.546 (0.597)	Data 0.000 (0.051)	Loss 1.156 (0.906)
Epoch: [35][180/200]	Time 0.553 (0.602)	Data 0.001 (0.055)	Loss 0.722 (0.929)
Epoch: [35][200/200]	Time 0.547 (0.605)	Data 0.001 (0.058)	Loss 1.206 (0.942)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.220)	Data 0.000 (0.035)	
Extract Features: [100/128]	Time 0.188 (0.206)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.911330461502075
==> Statistics for epoch 36: 1057 clusters
Epoch: [36][20/200]	Time 0.543 (0.608)	Data 0.001 (0.056)	Loss 0.104 (0.180)
Epoch: [36][40/200]	Time 0.545 (0.625)	Data 0.001 (0.072)	Loss 1.184 (0.342)
Epoch: [36][60/200]	Time 0.542 (0.600)	Data 0.000 (0.048)	Loss 0.975 (0.580)
Epoch: [36][80/200]	Time 0.544 (0.608)	Data 0.001 (0.058)	Loss 1.130 (0.699)
Epoch: [36][100/200]	Time 2.232 (0.613)	Data 1.654 (0.063)	Loss 1.069 (0.776)
Epoch: [36][120/200]	Time 0.547 (0.602)	Data 0.001 (0.052)	Loss 0.838 (0.825)
Epoch: [36][140/200]	Time 0.544 (0.608)	Data 0.001 (0.058)	Loss 1.037 (0.875)
Epoch: [36][160/200]	Time 0.546 (0.601)	Data 0.000 (0.051)	Loss 1.331 (0.906)
Epoch: [36][180/200]	Time 0.547 (0.606)	Data 0.001 (0.056)	Loss 1.249 (0.934)
Epoch: [36][200/200]	Time 0.553 (0.610)	Data 0.001 (0.059)	Loss 1.264 (0.950)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.225)	Data 0.000 (0.039)	
Extract Features: [100/128]	Time 0.185 (0.208)	Data 0.000 (0.022)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.16236615180969
==> Statistics for epoch 37: 1064 clusters
Epoch: [37][20/200]	Time 0.541 (0.598)	Data 0.001 (0.054)	Loss 0.117 (0.199)
Epoch: [37][40/200]	Time 0.542 (0.616)	Data 0.001 (0.071)	Loss 1.434 (0.349)
Epoch: [37][60/200]	Time 0.542 (0.594)	Data 0.000 (0.048)	Loss 0.946 (0.582)
Epoch: [37][80/200]	Time 0.546 (0.604)	Data 0.001 (0.057)	Loss 1.124 (0.710)
Epoch: [37][100/200]	Time 2.520 (0.612)	Data 1.953 (0.065)	Loss 0.967 (0.781)
Epoch: [37][120/200]	Time 0.543 (0.602)	Data 0.001 (0.054)	Loss 1.036 (0.836)
Epoch: [37][140/200]	Time 0.545 (0.607)	Data 0.001 (0.059)	Loss 1.137 (0.891)
Epoch: [37][160/200]	Time 0.547 (0.599)	Data 0.000 (0.052)	Loss 0.747 (0.911)
Epoch: [37][180/200]	Time 0.544 (0.603)	Data 0.001 (0.055)	Loss 0.860 (0.926)
Epoch: [37][200/200]	Time 0.556 (0.607)	Data 0.001 (0.058)	Loss 1.504 (0.945)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.217)	Data 0.000 (0.030)	
Extract Features: [100/128]	Time 0.181 (0.202)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.5588915348053
==> Statistics for epoch 38: 1064 clusters
Epoch: [38][20/200]	Time 0.542 (0.598)	Data 0.001 (0.051)	Loss 0.216 (0.172)
Epoch: [38][40/200]	Time 0.542 (0.618)	Data 0.001 (0.070)	Loss 0.839 (0.325)
Epoch: [38][60/200]	Time 0.640 (0.595)	Data 0.000 (0.047)	Loss 0.574 (0.564)
Epoch: [38][80/200]	Time 0.541 (0.603)	Data 0.001 (0.056)	Loss 1.140 (0.688)
Epoch: [38][100/200]	Time 2.262 (0.608)	Data 1.687 (0.061)	Loss 1.162 (0.761)
Epoch: [38][120/200]	Time 0.545 (0.599)	Data 0.001 (0.051)	Loss 0.999 (0.812)
Epoch: [38][140/200]	Time 0.544 (0.606)	Data 0.001 (0.057)	Loss 0.969 (0.848)
Epoch: [38][160/200]	Time 0.542 (0.599)	Data 0.000 (0.050)	Loss 0.943 (0.873)
Epoch: [38][180/200]	Time 0.691 (0.604)	Data 0.002 (0.055)	Loss 0.734 (0.897)
Epoch: [38][200/200]	Time 0.547 (0.607)	Data 0.001 (0.058)	Loss 1.063 (0.918)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.185 (0.215)	Data 0.001 (0.027)	
Extract Features: [100/128]	Time 0.185 (0.202)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.878056049346924
==> Statistics for epoch 39: 1065 clusters
Epoch: [39][20/200]	Time 0.545 (0.600)	Data 0.001 (0.057)	Loss 0.143 (0.184)
Epoch: [39][40/200]	Time 0.540 (0.624)	Data 0.001 (0.078)	Loss 0.769 (0.329)
Epoch: [39][60/200]	Time 0.545 (0.598)	Data 0.000 (0.052)	Loss 1.154 (0.599)
Epoch: [39][80/200]	Time 0.544 (0.607)	Data 0.001 (0.060)	Loss 1.330 (0.735)
Epoch: [39][100/200]	Time 2.305 (0.613)	Data 1.733 (0.065)	Loss 1.191 (0.804)
Epoch: [39][120/200]	Time 0.542 (0.602)	Data 0.000 (0.055)	Loss 1.093 (0.869)
Epoch: [39][140/200]	Time 0.555 (0.608)	Data 0.001 (0.061)	Loss 1.142 (0.899)
Epoch: [39][160/200]	Time 0.543 (0.601)	Data 0.000 (0.053)	Loss 1.043 (0.924)
Epoch: [39][180/200]	Time 0.544 (0.605)	Data 0.001 (0.057)	Loss 1.083 (0.940)
Epoch: [39][200/200]	Time 0.546 (0.608)	Data 0.001 (0.060)	Loss 1.083 (0.956)
Extract Features: [50/367]	Time 0.186 (0.228)	Data 0.001 (0.041)	
Extract Features: [100/367]	Time 0.184 (0.207)	Data 0.000 (0.021)	
Extract Features: [150/367]	Time 0.186 (0.200)	Data 0.000 (0.015)	
Extract Features: [200/367]	Time 0.180 (0.198)	Data 0.000 (0.011)	
Extract Features: [250/367]	Time 0.180 (0.196)	Data 0.000 (0.009)	
Extract Features: [300/367]	Time 0.188 (0.194)	Data 0.001 (0.008)	
Extract Features: [350/367]	Time 0.185 (0.193)	Data 0.000 (0.006)	
Mean AP: 70.8%

 * Finished epoch  39  model mAP: 70.8%  best: 70.8% *

==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.218)	Data 0.000 (0.033)	
Extract Features: [100/128]	Time 0.181 (0.204)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.74872708320618
==> Statistics for epoch 40: 1064 clusters
Epoch: [40][20/200]	Time 0.543 (0.600)	Data 0.001 (0.049)	Loss 0.171 (0.184)
Epoch: [40][40/200]	Time 0.538 (0.617)	Data 0.000 (0.070)	Loss 1.189 (0.312)
Epoch: [40][60/200]	Time 0.541 (0.594)	Data 0.000 (0.047)	Loss 0.796 (0.565)
Epoch: [40][80/200]	Time 0.543 (0.603)	Data 0.001 (0.057)	Loss 0.959 (0.694)
Epoch: [40][100/200]	Time 2.387 (0.610)	Data 1.824 (0.064)	Loss 1.139 (0.771)
Epoch: [40][120/200]	Time 0.545 (0.599)	Data 0.001 (0.053)	Loss 1.082 (0.833)
Epoch: [40][140/200]	Time 0.545 (0.604)	Data 0.001 (0.058)	Loss 1.193 (0.870)
Epoch: [40][160/200]	Time 0.543 (0.597)	Data 0.000 (0.051)	Loss 1.130 (0.896)
Epoch: [40][180/200]	Time 0.544 (0.602)	Data 0.001 (0.055)	Loss 1.070 (0.918)
Epoch: [40][200/200]	Time 0.554 (0.605)	Data 0.001 (0.058)	Loss 1.348 (0.940)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.216)	Data 0.000 (0.031)	
Extract Features: [100/128]	Time 0.190 (0.205)	Data 0.000 (0.017)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.599557638168335
==> Statistics for epoch 41: 1062 clusters
Epoch: [41][20/200]	Time 0.545 (0.605)	Data 0.001 (0.054)	Loss 0.157 (0.175)
Epoch: [41][40/200]	Time 0.548 (0.620)	Data 0.001 (0.071)	Loss 0.982 (0.336)
Epoch: [41][60/200]	Time 0.544 (0.597)	Data 0.000 (0.048)	Loss 0.885 (0.559)
Epoch: [41][80/200]	Time 0.547 (0.611)	Data 0.001 (0.060)	Loss 1.089 (0.687)
Epoch: [41][100/200]	Time 2.460 (0.619)	Data 1.896 (0.067)	Loss 1.122 (0.768)
Epoch: [41][120/200]	Time 0.546 (0.607)	Data 0.001 (0.056)	Loss 0.824 (0.829)
Epoch: [41][140/200]	Time 0.541 (0.611)	Data 0.000 (0.060)	Loss 0.994 (0.869)
Epoch: [41][160/200]	Time 0.542 (0.603)	Data 0.000 (0.053)	Loss 1.246 (0.894)
Epoch: [41][180/200]	Time 0.544 (0.607)	Data 0.001 (0.057)	Loss 1.196 (0.917)
Epoch: [41][200/200]	Time 0.550 (0.610)	Data 0.001 (0.060)	Loss 1.277 (0.933)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.178 (0.217)	Data 0.000 (0.034)	
Extract Features: [100/128]	Time 0.183 (0.204)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.60489344596863
==> Statistics for epoch 42: 1058 clusters
Epoch: [42][20/200]	Time 0.548 (0.609)	Data 0.001 (0.058)	Loss 0.111 (0.177)
Epoch: [42][40/200]	Time 0.545 (0.624)	Data 0.001 (0.075)	Loss 1.442 (0.332)
Epoch: [42][60/200]	Time 0.549 (0.600)	Data 0.000 (0.050)	Loss 1.066 (0.585)
Epoch: [42][80/200]	Time 0.542 (0.608)	Data 0.001 (0.060)	Loss 1.278 (0.700)
Epoch: [42][100/200]	Time 2.444 (0.615)	Data 1.836 (0.067)	Loss 1.182 (0.778)
Epoch: [42][120/200]	Time 0.553 (0.604)	Data 0.001 (0.056)	Loss 0.891 (0.830)
Epoch: [42][140/200]	Time 0.546 (0.609)	Data 0.001 (0.060)	Loss 1.172 (0.862)
Epoch: [42][160/200]	Time 0.543 (0.601)	Data 0.000 (0.053)	Loss 1.078 (0.889)
Epoch: [42][180/200]	Time 0.545 (0.605)	Data 0.001 (0.056)	Loss 0.874 (0.919)
Epoch: [42][200/200]	Time 0.560 (0.609)	Data 0.001 (0.059)	Loss 1.454 (0.936)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.182 (0.217)	Data 0.000 (0.032)	
Extract Features: [100/128]	Time 0.187 (0.205)	Data 0.001 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.74478483200073
==> Statistics for epoch 43: 1062 clusters
Epoch: [43][20/200]	Time 0.541 (0.605)	Data 0.001 (0.055)	Loss 0.163 (0.183)
Epoch: [43][40/200]	Time 0.542 (0.622)	Data 0.001 (0.073)	Loss 1.026 (0.323)
Epoch: [43][60/200]	Time 0.539 (0.596)	Data 0.000 (0.049)	Loss 1.114 (0.582)
Epoch: [43][80/200]	Time 0.543 (0.607)	Data 0.001 (0.060)	Loss 0.916 (0.718)
Epoch: [43][100/200]	Time 2.333 (0.614)	Data 1.758 (0.065)	Loss 0.804 (0.777)
Epoch: [43][120/200]	Time 0.544 (0.604)	Data 0.001 (0.055)	Loss 1.761 (0.844)
Epoch: [43][140/200]	Time 0.543 (0.609)	Data 0.001 (0.060)	Loss 0.836 (0.886)
Epoch: [43][160/200]	Time 0.542 (0.602)	Data 0.000 (0.052)	Loss 1.209 (0.920)
Epoch: [43][180/200]	Time 0.547 (0.605)	Data 0.001 (0.056)	Loss 0.971 (0.940)
Epoch: [43][200/200]	Time 0.549 (0.609)	Data 0.001 (0.060)	Loss 0.781 (0.958)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.180 (0.222)	Data 0.000 (0.036)	
Extract Features: [100/128]	Time 0.192 (0.207)	Data 0.000 (0.020)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.15301871299744
==> Statistics for epoch 44: 1044 clusters
Epoch: [44][20/200]	Time 0.543 (0.599)	Data 0.001 (0.055)	Loss 0.210 (0.184)
Epoch: [44][40/200]	Time 0.548 (0.621)	Data 0.001 (0.074)	Loss 0.987 (0.369)
Epoch: [44][60/200]	Time 0.544 (0.598)	Data 0.000 (0.050)	Loss 0.982 (0.605)
Epoch: [44][80/200]	Time 0.544 (0.608)	Data 0.001 (0.060)	Loss 1.433 (0.705)
Epoch: [44][100/200]	Time 0.652 (0.615)	Data 0.001 (0.066)	Loss 1.241 (0.792)
Epoch: [44][120/200]	Time 0.542 (0.603)	Data 0.001 (0.055)	Loss 0.916 (0.854)
Epoch: [44][140/200]	Time 0.549 (0.608)	Data 0.001 (0.060)	Loss 0.953 (0.882)
Epoch: [44][160/200]	Time 0.547 (0.602)	Data 0.000 (0.053)	Loss 1.352 (0.907)
Epoch: [44][180/200]	Time 0.547 (0.606)	Data 0.001 (0.056)	Loss 1.008 (0.930)
Epoch: [44][200/200]	Time 0.545 (0.609)	Data 0.001 (0.060)	Loss 0.955 (0.940)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.283 (0.221)	Data 0.000 (0.035)	
Extract Features: [100/128]	Time 0.181 (0.204)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.96951723098755
==> Statistics for epoch 45: 1059 clusters
Epoch: [45][20/200]	Time 0.544 (0.598)	Data 0.001 (0.055)	Loss 0.111 (0.180)
Epoch: [45][40/200]	Time 0.545 (0.621)	Data 0.001 (0.074)	Loss 1.136 (0.301)
Epoch: [45][60/200]	Time 0.544 (0.596)	Data 0.000 (0.049)	Loss 0.973 (0.571)
Epoch: [45][80/200]	Time 0.544 (0.610)	Data 0.001 (0.061)	Loss 1.121 (0.699)
Epoch: [45][100/200]	Time 2.395 (0.616)	Data 1.817 (0.067)	Loss 1.110 (0.769)
Epoch: [45][120/200]	Time 0.543 (0.604)	Data 0.001 (0.056)	Loss 0.783 (0.818)
Epoch: [45][140/200]	Time 0.542 (0.610)	Data 0.001 (0.062)	Loss 0.825 (0.854)
Epoch: [45][160/200]	Time 0.546 (0.602)	Data 0.000 (0.054)	Loss 1.449 (0.894)
Epoch: [45][180/200]	Time 0.549 (0.606)	Data 0.001 (0.058)	Loss 0.932 (0.914)
Epoch: [45][200/200]	Time 0.556 (0.610)	Data 0.001 (0.061)	Loss 1.247 (0.936)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.179 (0.214)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.187 (0.202)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 60.554439544677734
==> Statistics for epoch 46: 1062 clusters
Epoch: [46][20/200]	Time 0.540 (0.600)	Data 0.001 (0.055)	Loss 0.094 (0.171)
Epoch: [46][40/200]	Time 0.544 (0.618)	Data 0.001 (0.070)	Loss 0.785 (0.315)
Epoch: [46][60/200]	Time 0.545 (0.595)	Data 0.000 (0.047)	Loss 1.140 (0.564)
Epoch: [46][80/200]	Time 0.541 (0.605)	Data 0.001 (0.057)	Loss 0.876 (0.692)
Epoch: [46][100/200]	Time 2.309 (0.612)	Data 1.706 (0.063)	Loss 1.140 (0.761)
Epoch: [46][120/200]	Time 0.543 (0.601)	Data 0.001 (0.053)	Loss 1.157 (0.814)
Epoch: [46][140/200]	Time 0.544 (0.607)	Data 0.001 (0.058)	Loss 1.326 (0.856)
Epoch: [46][160/200]	Time 0.551 (0.600)	Data 0.000 (0.051)	Loss 1.409 (0.884)
Epoch: [46][180/200]	Time 0.547 (0.605)	Data 0.001 (0.055)	Loss 0.684 (0.906)
Epoch: [46][200/200]	Time 0.545 (0.608)	Data 0.001 (0.059)	Loss 1.349 (0.922)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.183 (0.215)	Data 0.000 (0.029)	
Extract Features: [100/128]	Time 0.181 (0.201)	Data 0.000 (0.016)	
Computing jaccard distance...
Jaccard distance computing time cost: 63.148460388183594
==> Statistics for epoch 47: 1063 clusters
Epoch: [47][20/200]	Time 0.542 (0.600)	Data 0.001 (0.057)	Loss 0.157 (0.169)
Epoch: [47][40/200]	Time 0.549 (0.623)	Data 0.001 (0.075)	Loss 1.503 (0.322)
Epoch: [47][60/200]	Time 0.541 (0.599)	Data 0.000 (0.050)	Loss 0.963 (0.563)
Epoch: [47][80/200]	Time 0.548 (0.610)	Data 0.001 (0.060)	Loss 1.072 (0.709)
Epoch: [47][100/200]	Time 2.489 (0.618)	Data 1.913 (0.067)	Loss 0.725 (0.770)
Epoch: [47][120/200]	Time 0.549 (0.606)	Data 0.001 (0.056)	Loss 1.036 (0.823)
Epoch: [47][140/200]	Time 0.546 (0.612)	Data 0.001 (0.061)	Loss 1.256 (0.868)
Epoch: [47][160/200]	Time 0.543 (0.604)	Data 0.000 (0.054)	Loss 0.809 (0.894)
Epoch: [47][180/200]	Time 0.548 (0.608)	Data 0.001 (0.057)	Loss 0.846 (0.922)
Epoch: [47][200/200]	Time 0.566 (0.612)	Data 0.001 (0.061)	Loss 0.974 (0.934)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.219)	Data 0.001 (0.033)	
Extract Features: [100/128]	Time 0.182 (0.205)	Data 0.000 (0.019)	
Computing jaccard distance...
Jaccard distance computing time cost: 59.90451169013977
==> Statistics for epoch 48: 1058 clusters
Epoch: [48][20/200]	Time 0.544 (0.606)	Data 0.001 (0.054)	Loss 0.117 (0.176)
Epoch: [48][40/200]	Time 0.544 (0.617)	Data 0.001 (0.070)	Loss 0.874 (0.317)
Epoch: [48][60/200]	Time 0.543 (0.595)	Data 0.000 (0.047)	Loss 0.980 (0.576)
Epoch: [48][80/200]	Time 0.541 (0.606)	Data 0.001 (0.058)	Loss 1.051 (0.695)
Epoch: [48][100/200]	Time 2.215 (0.611)	Data 1.642 (0.063)	Loss 1.077 (0.776)
Epoch: [48][120/200]	Time 0.546 (0.600)	Data 0.001 (0.052)	Loss 1.054 (0.823)
Epoch: [48][140/200]	Time 0.664 (0.605)	Data 0.001 (0.057)	Loss 0.997 (0.859)
Epoch: [48][160/200]	Time 0.546 (0.597)	Data 0.000 (0.050)	Loss 1.047 (0.886)
Epoch: [48][180/200]	Time 0.548 (0.603)	Data 0.001 (0.055)	Loss 1.086 (0.910)
Epoch: [48][200/200]	Time 0.554 (0.607)	Data 0.001 (0.058)	Loss 0.955 (0.926)
==> Create pseudo labels for unlabeled data
Extract Features: [50/128]	Time 0.181 (0.215)	Data 0.000 (0.028)	
Extract Features: [100/128]	Time 0.181 (0.202)	Data 0.000 (0.015)	
Computing jaccard distance...
Jaccard distance computing time cost: 61.95853805541992
==> Statistics for epoch 49: 1063 clusters
Epoch: [49][20/200]	Time 0.539 (0.601)	Data 0.001 (0.057)	Loss 0.117 (0.165)
Epoch: [49][40/200]	Time 0.540 (0.616)	Data 0.001 (0.070)	Loss 1.370 (0.341)
Epoch: [49][60/200]	Time 0.540 (0.593)	Data 0.000 (0.047)	Loss 1.115 (0.567)
Epoch: [49][80/200]	Time 0.543 (0.603)	Data 0.001 (0.057)	Loss 1.127 (0.671)
Epoch: [49][100/200]	Time 2.253 (0.609)	Data 1.683 (0.062)	Loss 1.453 (0.761)
Epoch: [49][120/200]	Time 0.547 (0.599)	Data 0.001 (0.052)	Loss 0.646 (0.817)
Epoch: [49][140/200]	Time 0.550 (0.605)	Data 0.001 (0.057)	Loss 1.015 (0.853)
Epoch: [49][160/200]	Time 0.544 (0.599)	Data 0.000 (0.050)	Loss 1.317 (0.879)
Epoch: [49][180/200]	Time 0.552 (0.602)	Data 0.001 (0.053)	Loss 0.812 (0.903)
Epoch: [49][200/200]	Time 0.553 (0.606)	Data 0.001 (0.058)	Loss 0.831 (0.922)
Extract Features: [50/367]	Time 0.183 (0.220)	Data 0.000 (0.032)	
Extract Features: [100/367]	Time 0.183 (0.202)	Data 0.000 (0.016)	
Extract Features: [150/367]	Time 0.183 (0.197)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.187 (0.196)	Data 0.000 (0.008)	
Extract Features: [250/367]	Time 0.179 (0.194)	Data 0.000 (0.007)	
Extract Features: [300/367]	Time 0.182 (0.193)	Data 0.000 (0.006)	
Extract Features: [350/367]	Time 0.190 (0.192)	Data 0.000 (0.005)	
Mean AP: 70.9%

 * Finished epoch  49  model mAP: 70.9%  best: 70.9% *

==> Test with the best model:
=> Loaded checkpoint 'log/market2msmt/resnet101_ibn_cion/model_best.pth.tar'
Extract Features: [50/367]	Time 0.180 (0.214)	Data 0.000 (0.031)	
Extract Features: [100/367]	Time 0.210 (0.200)	Data 0.023 (0.017)	
Extract Features: [150/367]	Time 0.186 (0.196)	Data 0.000 (0.011)	
Extract Features: [200/367]	Time 0.181 (0.193)	Data 0.000 (0.009)	
Extract Features: [250/367]	Time 0.182 (0.192)	Data 0.000 (0.007)	
Extract Features: [300/367]	Time 0.188 (0.192)	Data 0.001 (0.006)	
Extract Features: [350/367]	Time 0.179 (0.191)	Data 0.000 (0.005)	
Mean AP: 70.9%
CMC Scores:
  top-1          88.2%
  top-5          93.7%
  top-10         95.0%
Total running time:  3:31:38.272774
