Special test with strict convergence condition

Precision: [tensor(0.4156, device='cuda:0'), tensor(0.4193, device='cuda:0'), tensor(0.4295, device='cuda:0'), tensor(0.4248, device='cuda:0'), tensor(0.4190, device='cuda:0'), tensor(0.4269, device='cuda:0'), tensor(0.4243, device='cuda:0'), tensor(0.4177, device='cuda:0'), tensor(0.4159, device='cuda:0'), tensor(0.4237, device='cuda:0')]

Output distance: [tensor(5.4749, device='cuda:0'), tensor(5.4676, device='cuda:0'), tensor(5.4471, device='cuda:0'), tensor(5.4566, device='cuda:0'), tensor(5.4681, device='cuda:0'), tensor(5.4523, device='cuda:0'), tensor(5.4576, device='cuda:0'), tensor(5.4707, device='cuda:0'), tensor(5.4744, device='cuda:0'), tensor(5.4587, device='cuda:0')]

Prediction loss: [tensor(17939848., device='cuda:0'), tensor(17834850., device='cuda:0'), tensor(19549576., device='cuda:0'), tensor(19430386., device='cuda:0'), tensor(17792048., device='cuda:0'), tensor(18461166., device='cuda:0'), tensor(17046276., device='cuda:0'), tensor(17911492., device='cuda:0'), tensor(17391086., device='cuda:0'), tensor(19112850., device='cuda:0')]

Others: [{'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 7, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(3809, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}]

Compressed training loss: [tensor(40810.4961, device='cuda:0'), tensor(40907.2617, device='cuda:0'), tensor(40717.7930, device='cuda:0'), tensor(40806.5938, device='cuda:0'), tensor(40910.8750, device='cuda:0'), tensor(40943.7969, device='cuda:0'), tensor(41034.4453, device='cuda:0'), tensor(40723.2305, device='cuda:0'), tensor(40881.1797, device='cuda:0'), tensor(40773.2930, device='cuda:0')]

Training loss: 0

Prediction time: [datetime.timedelta(seconds=1, microseconds=58556), datetime.timedelta(seconds=1, microseconds=42623), datetime.timedelta(seconds=1, microseconds=55568), datetime.timedelta(seconds=1, microseconds=21710), datetime.timedelta(seconds=1, microseconds=62540), datetime.timedelta(seconds=1, microseconds=32664), datetime.timedelta(seconds=1, microseconds=12749), datetime.timedelta(seconds=1, microseconds=49544), datetime.timedelta(seconds=1, microseconds=48599), datetime.timedelta(seconds=1, microseconds=26689)]

Phi time: [datetime.timedelta(microseconds=239992), datetime.timedelta(microseconds=261900), datetime.timedelta(microseconds=233023), datetime.timedelta(microseconds=253935), datetime.timedelta(microseconds=232026), datetime.timedelta(microseconds=249951), datetime.timedelta(microseconds=255925), datetime.timedelta(microseconds=242033), datetime.timedelta(microseconds=253934), datetime.timedelta(microseconds=256922)]

