Precision: [tensor(0.8508, device='cuda:0'), tensor(0.8500, device='cuda:0'), tensor(0.8515, device='cuda:0'), tensor(0.8507, device='cuda:0'), tensor(0.8532, device='cuda:0'), tensor(0.8512, device='cuda:0'), tensor(0.8536, device='cuda:0'), tensor(0.8510, device='cuda:0'), tensor(0.8508, device='cuda:0'), tensor(0.8541, device='cuda:0')]

Output distance: [tensor(561.0456, device='cuda:0'), tensor(572.2171, device='cuda:0'), tensor(564.5418, device='cuda:0'), tensor(574.0524, device='cuda:0'), tensor(557.7694, device='cuda:0'), tensor(566.4288, device='cuda:0'), tensor(550.7211, device='cuda:0'), tensor(563.2049, device='cuda:0'), tensor(567.1424, device='cuda:0'), tensor(554.4283, device='cuda:0')]

Prediction loss: [tensor(605.6771, device='cuda:0'), tensor(598.5056, device='cuda:0'), tensor(579.9318, device='cuda:0'), tensor(583.7857, device='cuda:0'), tensor(617.8005, device='cuda:0'), tensor(597.1766, device='cuda:0'), tensor(600.2729, device='cuda:0'), tensor(602.7542, device='cuda:0'), tensor(599.3927, device='cuda:0'), tensor(628.9639, device='cuda:0')]

Others: [{'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 17, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}]

Compressed training loss: [tensor(8825768., device='cuda:0'), tensor(8780471., device='cuda:0'), tensor(8472220., device='cuda:0'), tensor(8523030., device='cuda:0'), tensor(9052894., device='cuda:0'), tensor(8738829., device='cuda:0'), tensor(8719103., device='cuda:0'), tensor(8828791., device='cuda:0'), tensor(8776244., device='cuda:0'), tensor(9107113., device='cuda:0')]

Training loss: 8792844.0

Prediction time: [datetime.timedelta(microseconds=764755), datetime.timedelta(microseconds=795621), datetime.timedelta(microseconds=802599), datetime.timedelta(microseconds=837450), datetime.timedelta(microseconds=787664), datetime.timedelta(microseconds=777701), datetime.timedelta(microseconds=792639), datetime.timedelta(microseconds=766748), datetime.timedelta(microseconds=791641), datetime.timedelta(microseconds=787659)]

Phi time: [datetime.timedelta(seconds=1, microseconds=361337), datetime.timedelta(microseconds=811602), datetime.timedelta(microseconds=740523), datetime.timedelta(microseconds=736175), datetime.timedelta(microseconds=753685), datetime.timedelta(microseconds=739155), datetime.timedelta(microseconds=755940), datetime.timedelta(microseconds=736898), datetime.timedelta(microseconds=741712), datetime.timedelta(microseconds=740654)]

