Precision: [tensor(0.8556, device='cuda:0'), tensor(0.8547, device='cuda:0'), tensor(0.8549, device='cuda:0'), tensor(0.8538, device='cuda:0'), tensor(0.8537, device='cuda:0'), tensor(0.8556, device='cuda:0'), tensor(0.8558, device='cuda:0'), tensor(0.8545, device='cuda:0'), tensor(0.8521, device='cuda:0'), tensor(0.8527, device='cuda:0')]

Output distance: [tensor(543.1511, device='cuda:0'), tensor(544.5471, device='cuda:0'), tensor(542.7587, device='cuda:0'), tensor(547.8660, device='cuda:0'), tensor(546.5544, device='cuda:0'), tensor(539.4700, device='cuda:0'), tensor(537.6254, device='cuda:0'), tensor(546.9230, device='cuda:0'), tensor(558.3221, device='cuda:0'), tensor(555.6337, device='cuda:0')]

Prediction loss: [tensor(588.7733, device='cuda:0'), tensor(605.7986, device='cuda:0'), tensor(621.6051, device='cuda:0'), tensor(604.1166, device='cuda:0'), tensor(608.3284, device='cuda:0'), tensor(606.4581, device='cuda:0'), tensor(589.0040, device='cuda:0'), tensor(594.7740, device='cuda:0'), tensor(588.6554, device='cuda:0'), tensor(595.1805, 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': 11, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 11, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 11, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}, {'iter_num': 11, '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': 11, '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': 13, 'num_positive': tensor(18000, device='cuda:0'), 'num_positive_true': tensor(18000, device='cuda:0')}]

Compressed training loss: [tensor(8650110., device='cuda:0'), tensor(8930270., device='cuda:0'), tensor(9079077., device='cuda:0'), tensor(8874128., device='cuda:0'), tensor(8900918., device='cuda:0'), tensor(8898656., device='cuda:0'), tensor(8638856., device='cuda:0'), tensor(8706789., device='cuda:0'), tensor(8680064., device='cuda:0'), tensor(8767551., device='cuda:0')]

Training loss: 8837685.0

Prediction time: [datetime.timedelta(microseconds=953954), datetime.timedelta(microseconds=905161), datetime.timedelta(microseconds=806579), datetime.timedelta(microseconds=811559), datetime.timedelta(microseconds=793631), datetime.timedelta(microseconds=802597), datetime.timedelta(microseconds=889230), datetime.timedelta(microseconds=813549), datetime.timedelta(microseconds=873296), datetime.timedelta(microseconds=892217)]

Phi time: [datetime.timedelta(seconds=1, microseconds=489902), datetime.timedelta(microseconds=916630), datetime.timedelta(microseconds=892478), datetime.timedelta(microseconds=877190), datetime.timedelta(microseconds=876284), datetime.timedelta(microseconds=878574), datetime.timedelta(microseconds=870864), datetime.timedelta(microseconds=879771), datetime.timedelta(microseconds=871869), datetime.timedelta(microseconds=878879)]

