Special test with strict convergence condition

Precision: [tensor(0.6221, device='cuda:0'), tensor(0.6240, device='cuda:0'), tensor(0.6250, device='cuda:0'), tensor(0.6244, device='cuda:0'), tensor(0.6287, device='cuda:0'), tensor(0.6297, device='cuda:0'), tensor(0.6284, device='cuda:0'), tensor(0.6218, device='cuda:0'), tensor(0.6287, device='cuda:0'), tensor(0.6233, device='cuda:0')]

Output distance: [tensor(4.9748, device='cuda:0'), tensor(4.9640, device='cuda:0'), tensor(4.9659, device='cuda:0'), tensor(4.9648, device='cuda:0'), tensor(4.9533, device='cuda:0'), tensor(4.9514, device='cuda:0'), tensor(4.9514, device='cuda:0'), tensor(4.9748, device='cuda:0'), tensor(4.9509, device='cuda:0'), tensor(4.9709, device='cuda:0')]

Prediction loss: [tensor(18854718., device='cuda:0'), tensor(19717912., device='cuda:0'), tensor(18247722., device='cuda:0'), tensor(18543246., device='cuda:0'), tensor(18010714., device='cuda:0'), tensor(19926280., device='cuda:0'), tensor(18008200., device='cuda:0'), tensor(18321272., device='cuda:0'), tensor(18719828., device='cuda:0'), tensor(18354980., device='cuda:0')]

Others: [{'iter_num': 5, 'num_positive': tensor(5170, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5253, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5186, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5226, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5222, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5209, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5261, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5180, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5257, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 5, 'num_positive': tensor(5179, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}]

Compressed training loss: [tensor(40843.8711, device='cuda:0'), tensor(40919.6875, device='cuda:0'), tensor(40854.8359, device='cuda:0'), tensor(40880.1914, device='cuda:0'), tensor(40926.5078, device='cuda:0'), tensor(40990.0859, device='cuda:0'), tensor(41058.2500, device='cuda:0'), tensor(40816.2695, device='cuda:0'), tensor(40855.8711, device='cuda:0'), tensor(40765.2344, device='cuda:0')]

Training loss: 0

Prediction time: [datetime.timedelta(seconds=6, microseconds=261445), datetime.timedelta(seconds=6, microseconds=153900), datetime.timedelta(seconds=6, microseconds=227589), datetime.timedelta(seconds=6, microseconds=292314), datetime.timedelta(seconds=6, microseconds=35403), datetime.timedelta(seconds=6, microseconds=239537), datetime.timedelta(seconds=6, microseconds=62290), datetime.timedelta(seconds=6, microseconds=81209), datetime.timedelta(seconds=6, microseconds=48348), datetime.timedelta(seconds=6, microseconds=252482)]

Phi time: [datetime.timedelta(microseconds=398310), datetime.timedelta(microseconds=380386), datetime.timedelta(microseconds=374412), datetime.timedelta(microseconds=407272), datetime.timedelta(microseconds=418227), datetime.timedelta(microseconds=423206), datetime.timedelta(microseconds=401297), datetime.timedelta(microseconds=432167), datetime.timedelta(microseconds=456066), datetime.timedelta(microseconds=338564)]

