Precision: [tensor(0.8204, device='cuda:0'), tensor(0.8213, device='cuda:0'), tensor(0.8238, device='cuda:0'), tensor(0.8213, device='cuda:0'), tensor(0.8189, device='cuda:0'), tensor(0.8222, device='cuda:0'), tensor(0.8189, device='cuda:0'), tensor(0.8219, device='cuda:0'), tensor(0.8224, device='cuda:0'), tensor(0.8222, device='cuda:0')]

Output distance: [tensor(14125.0127, device='cuda:0'), tensor(14027.0615, device='cuda:0'), tensor(13839.5947, device='cuda:0'), tensor(14023.3496, device='cuda:0'), tensor(14200.1328, device='cuda:0'), tensor(13996.6973, device='cuda:0'), tensor(14217.2080, device='cuda:0'), tensor(14039.5176, device='cuda:0'), tensor(13963.2891, device='cuda:0'), tensor(14003.3721, device='cuda:0')]

Prediction loss: [tensor(10234.0762, device='cuda:0'), tensor(10044.5928, device='cuda:0'), tensor(10586.8203, device='cuda:0'), tensor(10429.5967, device='cuda:0'), tensor(10049.4609, device='cuda:0'), tensor(10515.6191, device='cuda:0'), tensor(10570.4590, device='cuda:0'), tensor(10308.3057, device='cuda:0'), tensor(10319.9639, device='cuda:0'), tensor(10698.4141, 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': 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')}]

Compressed training loss: [tensor(1.8820e+08, device='cuda:0'), tensor(1.8429e+08, device='cuda:0'), tensor(1.9419e+08, device='cuda:0'), tensor(1.9111e+08, device='cuda:0'), tensor(1.8497e+08, device='cuda:0'), tensor(1.9197e+08, device='cuda:0'), tensor(1.9422e+08, device='cuda:0'), tensor(1.8955e+08, device='cuda:0'), tensor(1.8844e+08, device='cuda:0'), tensor(1.9650e+08, device='cuda:0')]

Training loss: 191104928.0

Prediction time: [datetime.timedelta(microseconds=798613), datetime.timedelta(microseconds=820518), datetime.timedelta(microseconds=818529), datetime.timedelta(microseconds=806576), datetime.timedelta(microseconds=888236), datetime.timedelta(microseconds=811559), datetime.timedelta(microseconds=816537), datetime.timedelta(microseconds=812555), datetime.timedelta(microseconds=811560), datetime.timedelta(microseconds=816535)]

Phi time: [datetime.timedelta(seconds=1, microseconds=370179), datetime.timedelta(microseconds=818198), datetime.timedelta(microseconds=743080), datetime.timedelta(microseconds=742278), datetime.timedelta(microseconds=744440), datetime.timedelta(microseconds=736698), datetime.timedelta(microseconds=744026), datetime.timedelta(microseconds=741784), datetime.timedelta(microseconds=745872), datetime.timedelta(microseconds=740784)]

