Special test with strict convergence condition

Precision: [tensor(0.7416, device='cuda:0'), tensor(0.7400, device='cuda:0'), tensor(0.7382, device='cuda:0'), tensor(0.7292, device='cuda:0'), tensor(0.7338, device='cuda:0'), tensor(0.7329, device='cuda:0'), tensor(0.7310, device='cuda:0'), tensor(0.7343, device='cuda:0'), tensor(0.7403, device='cuda:0'), tensor(0.7430, device='cuda:0')]

Output distance: [tensor(4.9992, device='cuda:0'), tensor(5.0047, device='cuda:0'), tensor(5.0060, device='cuda:0'), tensor(5.0155, device='cuda:0'), tensor(5.0123, device='cuda:0'), tensor(5.0150, device='cuda:0'), tensor(5.0144, device='cuda:0'), tensor(5.0097, device='cuda:0'), tensor(5.0029, device='cuda:0'), tensor(5.0013, device='cuda:0')]

Prediction loss: [tensor(18547362., device='cuda:0'), tensor(18817328., device='cuda:0'), tensor(18875404., device='cuda:0'), tensor(18088096., device='cuda:0'), tensor(19339732., device='cuda:0'), tensor(18906744., device='cuda:0'), tensor(17902896., device='cuda:0'), tensor(19280452., device='cuda:0'), tensor(19390836., device='cuda:0'), tensor(17736064., device='cuda:0')]

Others: [{'iter_num': 7, 'num_positive': tensor(2419, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(2392, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(2399, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(2415, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 11, 'num_positive': tensor(2393, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(2381, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(2405, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 7, 'num_positive': tensor(2409, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 9, 'num_positive': tensor(2403, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}, {'iter_num': 7, 'num_positive': tensor(2389, device='cuda:0'), 'num_positive_true': tensor(20211, device='cuda:0')}]

Compressed training loss: [tensor(40855.7500, device='cuda:0'), tensor(40754.3086, device='cuda:0'), tensor(40723.8438, device='cuda:0'), tensor(40909.2422, device='cuda:0'), tensor(40634.3203, device='cuda:0'), tensor(40911.9961, device='cuda:0'), tensor(40797.4531, device='cuda:0'), tensor(40876.3281, device='cuda:0'), tensor(40817.2539, device='cuda:0'), tensor(40763.7617, device='cuda:0')]

Training loss: 0

Prediction time: [datetime.timedelta(seconds=1, microseconds=44570), datetime.timedelta(seconds=1, microseconds=53531), datetime.timedelta(seconds=1, microseconds=12705), datetime.timedelta(seconds=1, microseconds=15692), datetime.timedelta(seconds=1, microseconds=22664), datetime.timedelta(seconds=1, microseconds=31624), datetime.timedelta(seconds=1, microseconds=13701), datetime.timedelta(seconds=1, microseconds=27641), datetime.timedelta(seconds=1, microseconds=41582), datetime.timedelta(seconds=1, microseconds=26645)]

Phi time: [datetime.timedelta(microseconds=238987), datetime.timedelta(microseconds=259898), datetime.timedelta(microseconds=247949), datetime.timedelta(microseconds=248945), datetime.timedelta(microseconds=233012), datetime.timedelta(microseconds=235999), datetime.timedelta(microseconds=255915), datetime.timedelta(microseconds=237991), datetime.timedelta(microseconds=252928), datetime.timedelta(microseconds=256911)]

