Precision: [tensor(0.9998, device='cuda:0'), tensor(0.9993, device='cuda:0'), tensor(0.9993, device='cuda:0'), tensor(0.9995, device='cuda:0'), tensor(0.9995, device='cuda:0'), tensor(0.9995, device='cuda:0'), tensor(0.9995, device='cuda:0'), tensor(0.9995, device='cuda:0'), tensor(0.9995, device='cuda:0'), tensor(0.9995, device='cuda:0')]

Output distance: [tensor(148076.7188, device='cuda:0'), tensor(140864.2500, device='cuda:0'), tensor(156966.7188, device='cuda:0'), tensor(146112.0625, device='cuda:0'), tensor(145333.5156, device='cuda:0'), tensor(141409.0625, device='cuda:0'), tensor(146079.8125, device='cuda:0'), tensor(150636.6875, device='cuda:0'), tensor(142080.1094, device='cuda:0'), tensor(143940.7812, device='cuda:0')]

Prediction loss: [tensor(142807.6406, device='cuda:0'), tensor(133179.5469, device='cuda:0'), tensor(165200.8906, device='cuda:0'), tensor(147685.9844, device='cuda:0'), tensor(144918.7656, device='cuda:0'), tensor(142255.5469, device='cuda:0'), tensor(146273.0625, device='cuda:0'), tensor(152961.7812, device='cuda:0'), tensor(140129.0469, device='cuda:0'), tensor(140930.6719, device='cuda:0')]

Others: [{'iter_num': 29, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 29, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 15, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(6000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}]

Compressed training loss: [tensor(1.9266e+08, device='cuda:0'), tensor(1.9162e+08, device='cuda:0'), tensor(1.9683e+08, device='cuda:0'), tensor(1.9587e+08, device='cuda:0'), tensor(1.9387e+08, device='cuda:0'), tensor(1.9351e+08, device='cuda:0'), tensor(1.9523e+08, device='cuda:0'), tensor(1.9424e+08, device='cuda:0'), tensor(1.8925e+08, device='cuda:0'), tensor(1.9145e+08, device='cuda:0')]

Training loss: 192450368.0

Prediction time: [datetime.timedelta(seconds=1, microseconds=645077), datetime.timedelta(seconds=1, microseconds=25650), datetime.timedelta(seconds=1, microseconds=697800), datetime.timedelta(seconds=1, microseconds=638054), datetime.timedelta(seconds=1, microseconds=696803), datetime.timedelta(seconds=1, microseconds=693817), datetime.timedelta(seconds=1, microseconds=697800), datetime.timedelta(seconds=1, microseconds=678880), datetime.timedelta(seconds=1, microseconds=8722), datetime.timedelta(seconds=1, microseconds=712736)]

Phi time: [datetime.timedelta(seconds=1, microseconds=477063), datetime.timedelta(microseconds=908712), datetime.timedelta(microseconds=855866), datetime.timedelta(microseconds=847181), datetime.timedelta(microseconds=849334), datetime.timedelta(microseconds=844629), datetime.timedelta(microseconds=850241), datetime.timedelta(microseconds=847232), datetime.timedelta(microseconds=845879), datetime.timedelta(microseconds=880253)]

