Precision: [tensor(0.8221, device='cuda:0'), tensor(0.8236, device='cuda:0'), tensor(0.8228, device='cuda:0'), tensor(0.8218, device='cuda:0'), tensor(0.8191, device='cuda:0'), tensor(0.8231, device='cuda:0'), tensor(0.8214, device='cuda:0'), tensor(0.8223, device='cuda:0'), tensor(0.8213, device='cuda:0'), tensor(0.8225, device='cuda:0')]

Output distance: [tensor(14021.3594, device='cuda:0'), tensor(13854.8008, device='cuda:0'), tensor(13917.7695, device='cuda:0'), tensor(13980.8730, device='cuda:0'), tensor(14199.7568, device='cuda:0'), tensor(13846.0879, device='cuda:0'), tensor(13987.2422, device='cuda:0'), tensor(13908.1533, device='cuda:0'), tensor(14066.1641, device='cuda:0'), tensor(13948.7080, device='cuda:0')]

Prediction loss: [tensor(10576.0156, device='cuda:0'), tensor(10711.5488, device='cuda:0'), tensor(10584.8750, device='cuda:0'), tensor(9694.5312, device='cuda:0'), tensor(9807.9141, device='cuda:0'), tensor(10501.8320, device='cuda:0'), tensor(10400.8115, device='cuda:0'), tensor(10489.3047, device='cuda:0'), tensor(10417.2637, device='cuda:0'), tensor(10317.0303, device='cuda:0')]

Others: [{'iter_num': 15, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 13, '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': 13, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}]

Compressed training loss: [tensor(1.9520e+08, device='cuda:0'), tensor(1.9755e+08, device='cuda:0'), tensor(1.9560e+08, device='cuda:0'), tensor(1.8010e+08, device='cuda:0'), tensor(1.8218e+08, device='cuda:0'), tensor(1.9363e+08, device='cuda:0'), tensor(1.9263e+08, device='cuda:0'), tensor(1.9298e+08, device='cuda:0'), tensor(1.9395e+08, device='cuda:0'), tensor(1.9131e+08, device='cuda:0')]

Training loss: 191636768.0

Prediction time: [datetime.timedelta(microseconds=725922), datetime.timedelta(microseconds=693061), datetime.timedelta(microseconds=745835), datetime.timedelta(microseconds=804587), datetime.timedelta(microseconds=738866), datetime.timedelta(microseconds=746829), datetime.timedelta(microseconds=740858), datetime.timedelta(microseconds=734884), datetime.timedelta(microseconds=747829), datetime.timedelta(microseconds=683102)]

Phi time: [datetime.timedelta(seconds=1, microseconds=188434), datetime.timedelta(microseconds=714682), datetime.timedelta(microseconds=647796), datetime.timedelta(microseconds=653863), datetime.timedelta(microseconds=647750), datetime.timedelta(microseconds=656210), datetime.timedelta(microseconds=652915), datetime.timedelta(microseconds=645692), datetime.timedelta(microseconds=649125), datetime.timedelta(microseconds=649415)]

