Precision: [tensor(0.8691, device='cuda:0'), tensor(0.8672, device='cuda:0'), tensor(0.8707, device='cuda:0'), tensor(0.8671, device='cuda:0'), tensor(0.8706, device='cuda:0'), tensor(0.8680, device='cuda:0'), tensor(0.8698, device='cuda:0'), tensor(0.8677, device='cuda:0'), tensor(0.8669, device='cuda:0'), tensor(0.8674, device='cuda:0')]

Output distance: [tensor(13829.1074, device='cuda:0'), tensor(13827.7373, device='cuda:0'), tensor(13657.1719, device='cuda:0'), tensor(14027.6221, device='cuda:0'), tensor(13657.6758, device='cuda:0'), tensor(13960.4229, device='cuda:0'), tensor(13851.8066, device='cuda:0'), tensor(13862.1025, device='cuda:0'), tensor(14040.6064, device='cuda:0'), tensor(13896.6240, device='cuda:0')]

Prediction loss: [tensor(10792.8096, device='cuda:0'), tensor(11348.0303, device='cuda:0'), tensor(11100.7148, device='cuda:0'), tensor(11288.1650, device='cuda:0'), tensor(10900.8340, device='cuda:0'), tensor(10571.7402, device='cuda:0'), tensor(10495.6943, device='cuda:0'), tensor(10591.0889, device='cuda:0'), tensor(10831.6445, device='cuda:0'), tensor(10985.4971, device='cuda:0')]

Others: [{'num_positive': tensor(16872, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16930, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16907, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16862, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16911, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16846, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16853, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16889, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16855, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'num_positive': tensor(16900, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}]

Compressed training loss: [tensor(1.9131e+08, device='cuda:0'), tensor(2.0053e+08, device='cuda:0'), tensor(1.9721e+08, device='cuda:0'), tensor(1.9909e+08, device='cuda:0'), tensor(1.9446e+08, device='cuda:0'), tensor(1.8782e+08, device='cuda:0'), tensor(1.8749e+08, device='cuda:0'), tensor(1.8799e+08, device='cuda:0'), tensor(1.9202e+08, device='cuda:0'), tensor(1.9519e+08, device='cuda:0')]

Training loss: 192410960.0

Prediction time: [datetime.timedelta(seconds=99, microseconds=891375), datetime.timedelta(seconds=100, microseconds=86184), datetime.timedelta(seconds=99, microseconds=366792), datetime.timedelta(seconds=99, microseconds=992644), datetime.timedelta(seconds=100, microseconds=199332), datetime.timedelta(seconds=99, microseconds=755431), datetime.timedelta(seconds=100, microseconds=429972), datetime.timedelta(seconds=100, microseconds=392505), datetime.timedelta(seconds=99, microseconds=396443), datetime.timedelta(seconds=100, microseconds=115200)]

Phi time: [datetime.timedelta(seconds=1, microseconds=188779), datetime.timedelta(microseconds=636627), datetime.timedelta(microseconds=640233), datetime.timedelta(microseconds=643721), datetime.timedelta(microseconds=648669), datetime.timedelta(microseconds=638201), datetime.timedelta(microseconds=642857), datetime.timedelta(microseconds=639778), datetime.timedelta(microseconds=640942), datetime.timedelta(microseconds=639468)]

