Precision: [tensor(0.8549, device='cuda:0'), tensor(0.8095, device='cuda:0'), tensor(0.8511, device='cuda:0'), tensor(0.8203, device='cuda:0'), tensor(0.8114, device='cuda:0'), tensor(0.7998, device='cuda:0'), tensor(0.8485, device='cuda:0'), tensor(0.8441, device='cuda:0'), tensor(0.8423, device='cuda:0'), tensor(0.8355, device='cuda:0')]

Output distance: [tensor(16327.7236, device='cuda:0'), tensor(44305416., device='cuda:0'), tensor(190041.8594, device='cuda:0'), tensor(3.1623e+08, device='cuda:0'), tensor(7.7619e+08, device='cuda:0'), tensor(38657148., device='cuda:0'), tensor(213605.7812, device='cuda:0'), tensor(8060061., device='cuda:0'), tensor(21579782., device='cuda:0'), tensor(12755220., device='cuda:0')]

Prediction loss: [tensor(21658.4043, device='cuda:0'), tensor(56398260., device='cuda:0'), tensor(250872., device='cuda:0'), tensor(4.0777e+08, device='cuda:0'), tensor(1.0243e+09, device='cuda:0'), tensor(50225664., device='cuda:0'), tensor(288163.8438, device='cuda:0'), tensor(10488658., device='cuda:0'), tensor(30563600., device='cuda:0'), tensor(15865924., device='cuda:0')]

Others: [{'iter_num': 30, 'num_positive': tensor(17996, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17976, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17991, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17958, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17973, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17994, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17992, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17992, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17998, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 30, 'num_positive': tensor(17983, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}]

Compressed training loss: [tensor(8599333., device='cuda:0'), tensor(9018116., device='cuda:0'), tensor(8605614., device='cuda:0'), tensor(9059728., device='cuda:0'), tensor(8899799., device='cuda:0'), tensor(9116024., device='cuda:0'), tensor(8374837.5000, device='cuda:0'), tensor(8991477., device='cuda:0'), tensor(8819026., device='cuda:0'), tensor(9204768., device='cuda:0')]

Training loss: 8875484.0

Prediction time: [datetime.timedelta(seconds=1, microseconds=534492), datetime.timedelta(seconds=1, microseconds=558389), datetime.timedelta(seconds=1, microseconds=537477), datetime.timedelta(seconds=1, microseconds=565361), datetime.timedelta(seconds=1, microseconds=553410), datetime.timedelta(seconds=1, microseconds=532501), datetime.timedelta(seconds=1, microseconds=562374), datetime.timedelta(seconds=1, microseconds=532502), datetime.timedelta(seconds=1, microseconds=566350), datetime.timedelta(seconds=1, microseconds=559387)]

Phi time: [datetime.timedelta(seconds=1, microseconds=373867), datetime.timedelta(microseconds=813928), datetime.timedelta(microseconds=739584), datetime.timedelta(microseconds=743744), datetime.timedelta(microseconds=740356), datetime.timedelta(microseconds=742896), datetime.timedelta(microseconds=745934), datetime.timedelta(microseconds=743114), datetime.timedelta(microseconds=742668), datetime.timedelta(microseconds=748511)]

