The performance in Table about person-reid-triplet-loss-baseline HOT 2 CLOSED

`=========> Test on dataset: market1501 <=========

Extracting feature...
1000/1000 batches done, +1.59s, total 84.73s
Done, 85.02s
Computing distance...
Done, 1.29s
Computing scores...
User Warning: Version 0.18.1 is required for package scikit-learn, your current version is 0.19.1. As a result, the mAP score may not be totally correct. You can try pip uninstall scikit-learn and then pip install scikit-learn==0.18.1
Done, 14.85s
Single Query: [mAP: 58.49%], [cmc1: 77.58%], [cmc5: 90.74%], [cmc10: 94.18%]
Multi Query, Computing distance...
Done, 1.71s
Multi Query, Computing scores...
User Warning: Version 0.18.1 is required for package scikit-learn, your current version is 0.19.1. As a result, the mAP score may not be totally correct. You can try pip uninstall scikit-learn and then pip install scikit-learn==0.18.1
Done, 14.56s
Multi Query: [mAP: 67.08%], [cmc1: 84.12%], [cmc5: 94.45%], [cmc10: 96.59%]
Re-ranking distance...
Done, 95.28s
Computing scores for re-ranked distance...
User Warning: Version 0.18.1 is required for package scikit-learn, your current version is 0.19.1. As a result, the mAP score may not be totally correct. You can try pip uninstall scikit-learn and then pip install scikit-learn==0.18.1
Done, 15.77s
Re-ranked Single Query: [mAP: 74.54%], [cmc1: 81.32%], [cmc5: 89.49%], [cmc10: 91.89%]
Multi Query, Re-ranking distance...
Done, 90.14s
Multi Query, Computing scores for re-ranked distance...
User Warning: Version 0.18.1 is required for package scikit-learn, your current version is 0.19.1. As a result, the mAP score may not be totally correct. You can try pip uninstall scikit-learn and then pip install scikit-learn==0.18.1
Done, 15.77s
Re-ranked Multi Query: [mAP: 81.72%], [cmc1: 87.44%], [cmc5: 93.50%], [cmc10: 95.28%]
`

from person-reid-triplet-loss-baseline.

Thanks for attention!
I train and test on single dataset, without combining datasets. The scores in the table takes 300 epochs to train. You may set the argument exp_decay_at_epoch to 150, total_epochs to 300 and re-try.

from person-reid-triplet-loss-baseline.