Special test with strict convergence condition

Precision: [tensor(0.6299, device='cuda:0'), tensor(0.6275, device='cuda:0'), tensor(0.6308, device='cuda:0'), tensor(0.6278, device='cuda:0'), tensor(0.6258, device='cuda:0'), tensor(0.6254, device='cuda:0'), tensor(0.6255, device='cuda:0'), tensor(0.6300, device='cuda:0'), tensor(0.6255, device='cuda:0'), tensor(0.6253, device='cuda:0')]

Output distance: [tensor(4.9223, device='cuda:0'), tensor(4.9273, device='cuda:0'), tensor(4.9163, device='cuda:0'), tensor(4.9252, device='cuda:0'), tensor(4.9325, device='cuda:0'), tensor(4.9349, device='cuda:0'), tensor(4.9338, device='cuda:0'), tensor(4.9202, device='cuda:0'), tensor(4.9346, device='cuda:0'), tensor(4.9349, device='cuda:0')]

Prediction loss: [tensor(20302376., device='cuda:0'), tensor(18713362., device='cuda:0'), tensor(18446360., device='cuda:0'), tensor(17821910., device='cuda:0'), tensor(17713390., device='cuda:0'), tensor(18486272., device='cuda:0'), tensor(18153424., device='cuda:0'), tensor(18498094., device='cuda:0'), tensor(18466534., device='cuda:0'), tensor(19432366., device='cuda:0')]

Others: [{'iter_num': 9, 'num_positive': tensor(5628, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 11, 'num_positive': tensor(5659, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(5675, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 11, 'num_positive': tensor(5679, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(5655, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(5638, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(5650, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(5652, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 11, 'num_positive': tensor(5639, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 13, 'num_positive': tensor(5642, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}]

Compressed training loss: [tensor(40734.4453, device='cuda:0'), tensor(40788.6094, device='cuda:0'), tensor(40687.0430, device='cuda:0'), tensor(40960.6758, device='cuda:0'), tensor(40762.7578, device='cuda:0'), tensor(40843.3594, device='cuda:0'), tensor(40775.2734, device='cuda:0'), tensor(40742.5820, device='cuda:0'), tensor(40715.3789, device='cuda:0'), tensor(40826.0977, device='cuda:0')]

Training loss: 0

Prediction time: [datetime.timedelta(seconds=1, microseconds=95355), datetime.timedelta(seconds=1, microseconds=81413), datetime.timedelta(seconds=1, microseconds=52537), datetime.timedelta(seconds=1, microseconds=120249), datetime.timedelta(seconds=1, microseconds=69465), datetime.timedelta(seconds=1, microseconds=87388), datetime.timedelta(seconds=1, microseconds=69464), datetime.timedelta(seconds=1, microseconds=48553), datetime.timedelta(seconds=1, microseconds=69464), datetime.timedelta(seconds=1, microseconds=90375)]

Phi time: [datetime.timedelta(microseconds=239982), datetime.timedelta(microseconds=263881), datetime.timedelta(microseconds=236994), datetime.timedelta(microseconds=235999), datetime.timedelta(microseconds=235003), datetime.timedelta(microseconds=235003), datetime.timedelta(microseconds=256910), datetime.timedelta(microseconds=235999), datetime.timedelta(microseconds=241974), datetime.timedelta(microseconds=235998)]

