Precision: [tensor(0.5153, device='cuda:0'), tensor(0.4869, device='cuda:0'), tensor(0.5432, device='cuda:0'), tensor(0.5167, device='cuda:0'), tensor(0.5385, device='cuda:0'), tensor(0.5172, device='cuda:0'), tensor(0.5243, device='cuda:0'), tensor(0.5434, device='cuda:0'), tensor(0.5385, device='cuda:0'), tensor(0.5192, device='cuda:0')]

Output distance: [tensor(257352.0625, device='cuda:0'), tensor(271459.4062, device='cuda:0'), tensor(274433.8125, device='cuda:0'), tensor(227487.1094, device='cuda:0'), tensor(232141.0938, device='cuda:0'), tensor(267104.5625, device='cuda:0'), tensor(247528.7656, device='cuda:0'), tensor(224949.0156, device='cuda:0'), tensor(231888.3125, device='cuda:0'), tensor(253549.7188, device='cuda:0')]

Prediction loss: [tensor(275740.6875, device='cuda:0'), tensor(254121.7969, device='cuda:0'), tensor(339037.9375, device='cuda:0'), tensor(208795.7969, device='cuda:0'), tensor(246022.2656, device='cuda:0'), tensor(241529.3594, device='cuda:0'), tensor(233209.7969, device='cuda:0'), tensor(246626.1094, device='cuda:0'), tensor(237017.8906, device='cuda:0'), tensor(304669.7500, device='cuda:0')]

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

Compressed training loss: [tensor(1.8599e+08, device='cuda:0'), tensor(1.8123e+08, device='cuda:0'), tensor(1.9501e+08, device='cuda:0'), tensor(1.8091e+08, device='cuda:0'), tensor(1.9058e+08, device='cuda:0'), tensor(1.9055e+08, device='cuda:0'), tensor(1.8880e+08, device='cuda:0'), tensor(1.9491e+08, device='cuda:0'), tensor(1.9267e+08, device='cuda:0'), tensor(1.8769e+08, device='cuda:0')]

Training loss: 191779248.0

Prediction time: [datetime.timedelta(microseconds=28878), datetime.timedelta(microseconds=23901), datetime.timedelta(microseconds=24895), datetime.timedelta(microseconds=24895), datetime.timedelta(microseconds=23899), datetime.timedelta(microseconds=25851), datetime.timedelta(microseconds=24895), datetime.timedelta(microseconds=29872), datetime.timedelta(microseconds=24894), datetime.timedelta(microseconds=28875)]

Phi time: [datetime.timedelta(seconds=1, microseconds=306697), datetime.timedelta(microseconds=670894), datetime.timedelta(microseconds=679347), datetime.timedelta(microseconds=667388), datetime.timedelta(microseconds=670788), datetime.timedelta(microseconds=666083), datetime.timedelta(microseconds=668106), datetime.timedelta(microseconds=667980), datetime.timedelta(microseconds=667296), datetime.timedelta(microseconds=670906)]

