Hello, Would it be easy to reproduce the results for pascal dataset on a single gp

<a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/us

About the reproduction of results on Pascal dataset,about liheyoung/unimatch

Comments (7)

LiheYoung commented on September 24, 2024 1

Our total batch size is 8 (2 imgs / card x 4 cards) for both labeled and unlabeled images. The corresponding learning rate is 0.001. So if you only use a single card and the batch size is 8 / K, the learning rate should also be 0.001 / K. For example, if your batch size is 4 only with a single card, the learning rate then will be 0.0005.

from unimatch.

nysp78 commented on September 24, 2024

I ran the unimatch with batch size 4 and initial lr 0.0005 on a single GPU for the 1/16 partition of pascal and I only took 75.50 mIOU. Τhere is any other trick I can do or perhaps the results can not be reproduced on a single GPU?

from unimatch.

LiheYoung commented on September 24, 2024

Please detail which backbone and cropping size are adopted. And what the reported result is.

from unimatch.

nysp78 commented on September 24, 2024

[2022-11-25 16:02:33,141][ INFO] {'backbone': 'resnet101',
'batch_size': 4,
'conf_thresh': 0.95,
'criterion': {'kwargs': {'ignore_index': 255}, 'name': 'CELoss'},
'crop_size': 321,
'data_root': /path/to/pascal,
'dataset': 'pascal',
'dilations': [6, 12, 18],
'epochs': 80,
'lr': 0.0005,
'lr_multi': 10.0,
'multi_grid': False,
'nclass': 21,
'replace_stride_with_dilation': [False, False, True]}

[2022-11-25 16:02:34,849][ INFO] Total params: 59.5M

[2022-11-25 16:02:37,349][ INFO] ===========> Epoch: 0, LR: 0.0005, Previous best: 0.00
[2022-11-25 16:02:45,618][ INFO] Iters: 0, Total loss: 1.575, Loss x: 3.150, Loss s: 0.000, Loss w_fp: 0.000, Mask: 0.000
[2022-11-25 16:05:12,455][ INFO] Iters: 310, Total loss: 0.556, Loss x: 1.075, Loss s: 0.052, Loss w_fp: 0.021, Mask: 0.222
[2022-11-25 16:07:39,331][ INFO] Iters: 620, Total loss: 0.450, Loss x: 0.855, Loss s: 0.071, Loss w_fp: 0.019, Mask: 0.292
[2022-11-25 16:10:06,300][ INFO] Iters: 930, Total loss: 0.395, Loss x: 0.740, Loss s: 0.082, Loss w_fp: 0.018, Mask: 0.327
[2022-11-25 16:12:33,237][ INFO] Iters: 1240, Total loss: 0.361, Loss x: 0.668, Loss s: 0.090, Loss w_fp: 0.019, Mask: 0.366
[2022-11-25 16:14:58,762][ INFO] Iters: 1550, Total loss: 0.338, Loss x: 0.618, Loss s: 0.096, Loss w_fp: 0.019, Mask: 0.385
[2022-11-25 16:17:24,090][ INFO] Iters: 1860, Total loss: 0.318, Loss x: 0.576, Loss s: 0.102, Loss w_fp: 0.019, Mask: 0.404
[2022-11-25 16:19:49,541][ INFO] Iters: 2170, Total loss: 0.303, Loss x: 0.543, Loss s: 0.107, Loss w_fp: 0.020, Mask: 0.422
[2022-11-25 16:25:08,569][ INFO] ***** Evaluation original ***** >>>> meanIOU: 56.33

[2022-11-25 16:25:09,356][ INFO] ===========> Epoch: 1, LR: 0.0005, Previous best: 56.33
[2022-11-25 16:25:10,444][ INFO] Iters: 0, Total loss: 0.115, Loss x: 0.136, Loss s: 0.165, Loss w_fp: 0.022, Mask: 0.721
[2022-11-25 16:27:37,508][ INFO] Iters: 310, Total loss: 0.182, Loss x: 0.276, Loss s: 0.151, Loss w_fp: 0.024, Mask: 0.551
[2022-11-25 16:30:04,694][ INFO] Iters: 620, Total loss: 0.180, Loss x: 0.274, Loss s: 0.148, Loss w_fp: 0.022, Mask: 0.563
[2022-11-25 16:32:31,996][ INFO] Iters: 930, Total loss: 0.180, Loss x: 0.274, Loss s: 0.150, Loss w_fp: 0.022, Mask: 0.573
[2022-11-25 16:34:59,147][ INFO] Iters: 1240, Total loss: 0.176, Loss x: 0.265, Loss s: 0.151, Loss w_fp: 0.022, Mask: 0.584
[2022-11-25 16:37:26,400][ INFO] Iters: 1550, Total loss: 0.170, Loss x: 0.255, Loss s: 0.149, Loss w_fp: 0.022, Mask: 0.590
[2022-11-25 16:39:53,906][ INFO] Iters: 1860, Total loss: 0.169, Loss x: 0.252, Loss s: 0.150, Loss w_fp: 0.021, Mask: 0.595
[2022-11-25 16:42:21,170][ INFO] Iters: 2170, Total loss: 0.167, Loss x: 0.249, Loss s: 0.150, Loss w_fp: 0.021, Mask: 0.598
[2022-11-25 16:45:43,356][ INFO] ***** Evaluation original ***** >>>> meanIOU: 65.14

[2022-11-25 16:45:43,982][ INFO] ===========> Epoch: 2, LR: 0.0005, Previous best: 65.14
[2022-11-25 16:45:45,041][ INFO] Iters: 0, Total loss: 0.065, Loss x: 0.053, Loss s: 0.147, Loss w_fp: 0.008, Mask: 0.330
[2022-11-25 16:48:12,148][ INFO] Iters: 310, Total loss: 0.148, Loss x: 0.211, Loss s: 0.148, Loss w_fp: 0.020, Mask: 0.639
[2022-11-25 16:50:39,326][ INFO] Iters: 620, Total loss: 0.146, Loss x: 0.206, Loss s: 0.150, Loss w_fp: 0.021, Mask: 0.637
[2022-11-25 16:53:06,565][ INFO] Iters: 930, Total loss: 0.146, Loss x: 0.207, Loss s: 0.148, Loss w_fp: 0.020, Mask: 0.636
[2022-11-25 16:55:33,742][ INFO] Iters: 1240, Total loss: 0.142, Loss x: 0.201, Loss s: 0.148, Loss w_fp: 0.020, Mask: 0.640
[2022-11-25 16:58:00,975][ INFO] Iters: 1550, Total loss: 0.141, Loss x: 0.198, Loss s: 0.148, Loss w_fp: 0.020, Mask: 0.642
[2022-11-25 17:00:28,034][ INFO] Iters: 1860, Total loss: 0.139, Loss x: 0.194, Loss s: 0.146, Loss w_fp: 0.020, Mask: 0.646
[2022-11-25 17:02:55,229][ INFO] Iters: 2170, Total loss: 0.138, Loss x: 0.193, Loss s: 0.145, Loss w_fp: 0.019, Mask: 0.648
[2022-11-25 17:06:12,677][ INFO] ***** Evaluation original ***** >>>> meanIOU: 66.59

[2022-11-25 17:06:13,541][ INFO] ===========> Epoch: 3, LR: 0.0005, Previous best: 66.59
[2022-11-25 17:06:14,564][ INFO] Iters: 0, Total loss: 0.091, Loss x: 0.057, Loss s: 0.241, Loss w_fp: 0.009, Mask: 0.687
[2022-11-25 17:08:41,982][ INFO] Iters: 310, Total loss: 0.123, Loss x: 0.159, Loss s: 0.155, Loss w_fp: 0.019, Mask: 0.672
[2022-11-25 17:11:09,323][ INFO] Iters: 620, Total loss: 0.125, Loss x: 0.164, Loss s: 0.154, Loss w_fp: 0.018, Mask: 0.672
[2022-11-25 17:13:36,525][ INFO] Iters: 930, Total loss: 0.126, Loss x: 0.167, Loss s: 0.152, Loss w_fp: 0.018, Mask: 0.672
[2022-11-25 17:16:03,706][ INFO] Iters: 1240, Total loss: 0.126, Loss x: 0.169, Loss s: 0.149, Loss w_fp: 0.018, Mask: 0.669
[2022-11-25 17:18:30,888][ INFO] Iters: 1550, Total loss: 0.126, Loss x: 0.170, Loss s: 0.148, Loss w_fp: 0.018, Mask: 0.670
[2022-11-25 17:20:58,025][ INFO] Iters: 1860, Total loss: 0.125, Loss x: 0.168, Loss s: 0.147, Loss w_fp: 0.018, Mask: 0.670
[2022-11-25 17:23:25,015][ INFO] Iters: 2170, Total loss: 0.124, Loss x: 0.167, Loss s: 0.145, Loss w_fp: 0.018, Mask: 0.671
[2022-11-25 17:26:46,643][ INFO] ***** Evaluation original ***** >>>> meanIOU: 68.03

[2022-11-25 17:26:47,472][ INFO] ===========> Epoch: 4, LR: 0.0005, Previous best: 68.03
[2022-11-25 17:26:48,533][ INFO] Iters: 0, Total loss: 0.092, Loss x: 0.106, Loss s: 0.116, Loss w_fp: 0.039, Mask: 0.710
[2022-11-25 17:29:15,607][ INFO] Iters: 310, Total loss: 0.107, Loss x: 0.137, Loss s: 0.138, Loss w_fp: 0.015, Mask: 0.706
[2022-11-25 17:31:42,655][ INFO] Iters: 620, Total loss: 0.109, Loss x: 0.139, Loss s: 0.141, Loss w_fp: 0.016, Mask: 0.702
[2022-11-25 17:34:09,820][ INFO] Iters: 930, Total loss: 0.108, Loss x: 0.138, Loss s: 0.141, Loss w_fp: 0.016, Mask: 0.697
[2022-11-25 17:36:37,012][ INFO] Iters: 1240, Total loss: 0.107, Loss x: 0.137, Loss s: 0.138, Loss w_fp: 0.015, Mask: 0.699
[2022-11-25 17:39:04,242][ INFO] Iters: 1550, Total loss: 0.107, Loss x: 0.136, Loss s: 0.139, Loss w_fp: 0.015, Mask: 0.701
[2022-11-25 17:41:31,489][ INFO] Iters: 1860, Total loss: 0.107, Loss x: 0.137, Loss s: 0.138, Loss w_fp: 0.015, Mask: 0.703
[2022-11-25 17:43:58,650][ INFO] Iters: 2170, Total loss: 0.107, Loss x: 0.137, Loss s: 0.138, Loss w_fp: 0.015, Mask: 0.704
[2022-11-25 17:47:21,482][ INFO] ***** Evaluation original ***** >>>> meanIOU: 69.49

[2022-11-25 17:47:22,141][ INFO] ===========> Epoch: 5, LR: 0.0005, Previous best: 69.49
[2022-11-25 17:47:23,167][ INFO] Iters: 0, Total loss: 0.038, Loss x: 0.039, Loss s: 0.052, Loss w_fp: 0.022, Mask: 0.812
[2022-11-25 17:49:49,580][ INFO] Iters: 310, Total loss: 0.112, Loss x: 0.148, Loss s: 0.140, Loss w_fp: 0.015, Mask: 0.687
[2022-11-25 17:52:15,380][ INFO] Iters: 620, Total loss: 0.106, Loss x: 0.141, Loss s: 0.126, Loss w_fp: 0.014, Mask: 0.697
[2022-11-25 17:54:42,373][ INFO] Iters: 930, Total loss: 0.103, Loss x: 0.137, Loss s: 0.126, Loss w_fp: 0.014, Mask: 0.706
[2022-11-25 17:57:12,858][ INFO] Iters: 1240, Total loss: 0.102, Loss x: 0.134, Loss s: 0.124, Loss w_fp: 0.014, Mask: 0.708
[2022-11-25 17:59:44,193][ INFO] Iters: 1550, Total loss: 0.102, Loss x: 0.134, Loss s: 0.126, Loss w_fp: 0.014, Mask: 0.709
[2022-11-25 18:02:15,723][ INFO] Iters: 1860, Total loss: 0.102, Loss x: 0.134, Loss s: 0.128, Loss w_fp: 0.014, Mask: 0.708
[2022-11-25 18:04:47,196][ INFO] Iters: 2170, Total loss: 0.102, Loss x: 0.132, Loss s: 0.131, Loss w_fp: 0.014, Mask: 0.709
[2022-11-25 18:08:23,400][ INFO] ***** Evaluation original ***** >>>> meanIOU: 69.62

[2022-11-25 18:08:25,530][ INFO] ===========> Epoch: 6, LR: 0.0005, Previous best: 69.62
[2022-11-25 18:08:26,844][ INFO] Iters: 0, Total loss: 0.070, Loss x: 0.128, Loss s: 0.021, Loss w_fp: 0.001, Mask: 0.816
[2022-11-25 18:10:58,762][ INFO] Iters: 310, Total loss: 0.097, Loss x: 0.120, Loss s: 0.133, Loss w_fp: 0.013, Mask: 0.727
[2022-11-25 18:13:30,480][ INFO] Iters: 620, Total loss: 0.097, Loss x: 0.121, Loss s: 0.133, Loss w_fp: 0.013, Mask: 0.728
[2022-11-25 18:16:01,912][ INFO] Iters: 930, Total loss: 0.095, Loss x: 0.118, Loss s: 0.130, Loss w_fp: 0.014, Mask: 0.730
[2022-11-25 18:18:33,570][ INFO] Iters: 1240, Total loss: 0.094, Loss x: 0.117, Loss s: 0.131, Loss w_fp: 0.014, Mask: 0.730
[2022-11-25 18:21:05,225][ INFO] Iters: 1550, Total loss: 0.095, Loss x: 0.116, Loss s: 0.132, Loss w_fp: 0.014, Mask: 0.731
[2022-11-25 18:23:36,861][ INFO] Iters: 1860, Total loss: 0.095, Loss x: 0.116, Loss s: 0.134, Loss w_fp: 0.014, Mask: 0.733
[2022-11-25 18:26:08,254][ INFO] Iters: 2170, Total loss: 0.094, Loss x: 0.114, Loss s: 0.133, Loss w_fp: 0.014, Mask: 0.734
[2022-11-25 18:29:43,723][ INFO] ***** Evaluation original ***** >>>> meanIOU: 70.17

[2022-11-25 18:29:44,586][ INFO] ===========> Epoch: 7, LR: 0.0005, Previous best: 70.17
[2022-11-25 18:29:45,906][ INFO] Iters: 0, Total loss: 0.073, Loss x: 0.131, Loss s: 0.027, Loss w_fp: 0.005, Mask: 0.635
[2022-11-25 18:32:17,722][ INFO] Iters: 310, Total loss: 0.091, Loss x: 0.114, Loss s: 0.122, Loss w_fp: 0.013, Mask: 0.734
[2022-11-25 18:34:50,059][ INFO] Iters: 620, Total loss: 0.094, Loss x: 0.118, Loss s: 0.128, Loss w_fp: 0.013, Mask: 0.731
[2022-11-25 18:37:21,780][ INFO] Iters: 930, Total loss: 0.093, Loss x: 0.114, Loss s: 0.130, Loss w_fp: 0.013, Mask: 0.732
[2022-11-25 18:39:53,319][ INFO] Iters: 1240, Total loss: 0.092, Loss x: 0.114, Loss s: 0.128, Loss w_fp: 0.013, Mask: 0.734
[2022-11-25 18:42:25,440][ INFO] Iters: 1550, Total loss: 0.093, Loss x: 0.113, Loss s: 0.131, Loss w_fp: 0.013, Mask: 0.733
[2022-11-25 18:44:56,969][ INFO] Iters: 1860, Total loss: 0.092, Loss x: 0.112, Loss s: 0.130, Loss w_fp: 0.013, Mask: 0.733
[2022-11-25 18:47:28,275][ INFO] Iters: 2170, Total loss: 0.091, Loss x: 0.111, Loss s: 0.130, Loss w_fp: 0.013, Mask: 0.735
[2022-11-25 18:51:04,887][ INFO] ***** Evaluation original ***** >>>> meanIOU: 69.25

[2022-11-25 18:51:04,888][ INFO] ===========> Epoch: 8, LR: 0.0005, Previous best: 70.17
[2022-11-25 18:51:06,098][ INFO] Iters: 0, Total loss: 0.045, Loss x: 0.071, Loss s: 0.033, Loss w_fp: 0.006, Mask: 0.870
[2022-11-25 18:53:37,571][ INFO] Iters: 310, Total loss: 0.083, Loss x: 0.099, Loss s: 0.119, Loss w_fp: 0.014, Mask: 0.752
[2022-11-25 18:56:09,372][ INFO] Iters: 620, Total loss: 0.085, Loss x: 0.100, Loss s: 0.124, Loss w_fp: 0.014, Mask: 0.749
[2022-11-25 18:58:40,992][ INFO] Iters: 930, Total loss: 0.085, Loss x: 0.101, Loss s: 0.123, Loss w_fp: 0.014, Mask: 0.751
[2022-11-25 19:01:12,512][ INFO] Iters: 1240, Total loss: 0.085, Loss x: 0.102, Loss s: 0.123, Loss w_fp: 0.013, Mask: 0.749
[2022-11-25 19:03:43,991][ INFO] Iters: 1550, Total loss: 0.084, Loss x: 0.101, Loss s: 0.122, Loss w_fp: 0.013, Mask: 0.750
[2022-11-25 19:06:15,637][ INFO] Iters: 1860, Total loss: 0.083, Loss x: 0.100, Loss s: 0.120, Loss w_fp: 0.013, Mask: 0.751
[2022-11-25 19:08:46,753][ INFO] Iters: 2170, Total loss: 0.083, Loss x: 0.099, Loss s: 0.120, Loss w_fp: 0.013, Mask: 0.752
[2022-11-25 19:12:21,983][ INFO] ***** Evaluation original ***** >>>> meanIOU: 70.08

[2022-11-25 19:12:21,983][ INFO] ===========> Epoch: 9, LR: 0.0004, Previous best: 70.17
[2022-11-25 19:12:23,223][ INFO] Iters: 0, Total loss: 0.042, Loss x: 0.060, Loss s: 0.041, Loss w_fp: 0.006, Mask: 0.712
[2022-11-25 19:14:55,485][ INFO] Iters: 310, Total loss: 0.081, Loss x: 0.089, Loss s: 0.133, Loss w_fp: 0.014, Mask: 0.761
[2022-11-25 19:17:28,156][ INFO] Iters: 620, Total loss: 0.080, Loss x: 0.092, Loss s: 0.124, Loss w_fp: 0.013, Mask: 0.763
[2022-11-25 19:19:59,950][ INFO] Iters: 930, Total loss: 0.080, Loss x: 0.094, Loss s: 0.120, Loss w_fp: 0.013, Mask: 0.765
[2022-11-25 19:22:30,970][ INFO] Iters: 1240, Total loss: 0.080, Loss x: 0.094, Loss s: 0.119, Loss w_fp: 0.013, Mask: 0.764
[2022-11-25 19:25:03,080][ INFO] Iters: 1550, Total loss: 0.079, Loss x: 0.093, Loss s: 0.119, Loss w_fp: 0.012, Mask: 0.765
[2022-11-25 19:27:34,553][ INFO] Iters: 1860, Total loss: 0.079, Loss x: 0.093, Loss s: 0.118, Loss w_fp: 0.012, Mask: 0.763
[2022-11-25 19:30:05,906][ INFO] Iters: 2170, Total loss: 0.079, Loss x: 0.093, Loss s: 0.118, Loss w_fp: 0.012, Mask: 0.764
[2022-11-25 19:33:40,794][ INFO] ***** Evaluation original ***** >>>> meanIOU: 71.09

[2022-11-25 19:33:41,548][ INFO] ===========> Epoch: 10, LR: 0.0004, Previous best: 71.09
[2022-11-25 19:33:42,738][ INFO] Iters: 0, Total loss: 0.207, Loss x: 0.108, Loss s: 0.588, Loss w_fp: 0.026, Mask: 0.719
[2022-11-25 19:36:14,250][ INFO] Iters: 310, Total loss: 0.079, Loss x: 0.093, Loss s: 0.120, Loss w_fp: 0.011, Mask: 0.760
[2022-11-25 19:38:45,860][ INFO] Iters: 620, Total loss: 0.080, Loss x: 0.096, Loss s: 0.118, Loss w_fp: 0.011, Mask: 0.758
[2022-11-25 19:41:17,631][ INFO] Iters: 930, Total loss: 0.080, Loss x: 0.095, Loss s: 0.118, Loss w_fp: 0.011, Mask: 0.762
[2022-11-25 19:43:49,302][ INFO] Iters: 1240, Total loss: 0.080, Loss x: 0.096, Loss s: 0.116, Loss w_fp: 0.012, Mask: 0.761
[2022-11-25 19:46:21,250][ INFO] Iters: 1550, Total loss: 0.080, Loss x: 0.096, Loss s: 0.118, Loss w_fp: 0.012, Mask: 0.762
[2022-11-25 19:48:53,459][ INFO] Iters: 1860, Total loss: 0.080, Loss x: 0.095, Loss s: 0.118, Loss w_fp: 0.012, Mask: 0.760
[2022-11-25 19:51:25,058][ INFO] Iters: 2170, Total loss: 0.080, Loss x: 0.096, Loss s: 0.117, Loss w_fp: 0.012, Mask: 0.761
[2022-11-25 19:55:00,769][ INFO] ***** Evaluation original ***** >>>> meanIOU: 72.05

[2022-11-25 19:55:01,511][ INFO] ===========> Epoch: 11, LR: 0.0004, Previous best: 72.05
[2022-11-25 19:55:02,744][ INFO] Iters: 0, Total loss: 0.098, Loss x: 0.028, Loss s: 0.335, Loss w_fp: 0.003, Mask: 0.270
[2022-11-25 19:57:34,080][ INFO] Iters: 310, Total loss: 0.083, Loss x: 0.105, Loss s: 0.112, Loss w_fp: 0.011, Mask: 0.768
[2022-11-25 20:00:05,952][ INFO] Iters: 620, Total loss: 0.078, Loss x: 0.097, Loss s: 0.108, Loss w_fp: 0.011, Mask: 0.771
[2022-11-25 20:02:37,223][ INFO] Iters: 930, Total loss: 0.078, Loss x: 0.095, Loss s: 0.110, Loss w_fp: 0.011, Mask: 0.775
[2022-11-25 20:05:09,282][ INFO] Iters: 1240, Total loss: 0.077, Loss x: 0.093, Loss s: 0.112, Loss w_fp: 0.011, Mask: 0.773
[2022-11-25 20:07:41,667][ INFO] Iters: 1550, Total loss: 0.076, Loss x: 0.091, Loss s: 0.112, Loss w_fp: 0.011, Mask: 0.774
[2022-11-25 20:10:14,422][ INFO] Iters: 1860, Total loss: 0.077, Loss x: 0.090, Loss s: 0.116, Loss w_fp: 0.011, Mask: 0.774
[2022-11-25 20:12:46,237][ INFO] Iters: 2170, Total loss: 0.076, Loss x: 0.090, Loss s: 0.115, Loss w_fp: 0.011, Mask: 0.776
[2022-11-25 20:16:23,248][ INFO] ***** Evaluation original ***** >>>> meanIOU: 71.68

[2022-11-25 20:16:23,248][ INFO] ===========> Epoch: 12, LR: 0.0004, Previous best: 72.05
[2022-11-25 20:16:24,392][ INFO] Iters: 0, Total loss: 0.051, Loss x: 0.050, Loss s: 0.071, Loss w_fp: 0.034, Mask: 0.730
[2022-11-25 20:18:56,379][ INFO] Iters: 310, Total loss: 0.077, Loss x: 0.088, Loss s: 0.122, Loss w_fp: 0.011, Mask: 0.779
[2022-11-25 20:21:28,478][ INFO] Iters: 620, Total loss: 0.075, Loss x: 0.085, Loss s: 0.122, Loss w_fp: 0.011, Mask: 0.782
[2022-11-25 20:24:00,077][ INFO] Iters: 930, Total loss: 0.073, Loss x: 0.084, Loss s: 0.115, Loss w_fp: 0.011, Mask: 0.783
[2022-11-25 20:26:31,537][ INFO] Iters: 1240, Total loss: 0.074, Loss x: 0.084, Loss s: 0.117, Loss w_fp: 0.011, Mask: 0.782
[2022-11-25 20:29:03,227][ INFO] Iters: 1550, Total loss: 0.074, Loss x: 0.085, Loss s: 0.116, Loss w_fp: 0.011, Mask: 0.779
[2022-11-25 20:31:35,276][ INFO] Iters: 1860, Total loss: 0.074, Loss x: 0.085, Loss s: 0.115, Loss w_fp: 0.011, Mask: 0.779
[2022-11-25 20:34:06,577][ INFO] Iters: 2170, Total loss: 0.074, Loss x: 0.084, Loss s: 0.115, Loss w_fp: 0.011, Mask: 0.781
[2022-11-25 20:37:42,721][ INFO] ***** Evaluation original ***** >>>> meanIOU: 71.79

[2022-11-25 20:37:42,721][ INFO] ===========> Epoch: 13, LR: 0.0004, Previous best: 72.05
[2022-11-25 20:37:44,065][ INFO] Iters: 0, Total loss: 0.057, Loss x: 0.095, Loss s: 0.035, Loss w_fp: 0.004, Mask: 0.880
[2022-11-25 20:40:15,848][ INFO] Iters: 310, Total loss: 0.073, Loss x: 0.085, Loss s: 0.112, Loss w_fp: 0.009, Mask: 0.781
[2022-11-25 20:42:47,639][ INFO] Iters: 620, Total loss: 0.072, Loss x: 0.084, Loss s: 0.113, Loss w_fp: 0.010, Mask: 0.784
[2022-11-25 20:45:18,999][ INFO] Iters: 930, Total loss: 0.072, Loss x: 0.082, Loss s: 0.111, Loss w_fp: 0.010, Mask: 0.785
[2022-11-25 20:47:51,388][ INFO] Iters: 1240, Total loss: 0.072, Loss x: 0.083, Loss s: 0.112, Loss w_fp: 0.010, Mask: 0.785
[2022-11-25 20:50:23,184][ INFO] Iters: 1550, Total loss: 0.071, Loss x: 0.082, Loss s: 0.111, Loss w_fp: 0.011, Mask: 0.787
[2022-11-25 20:52:54,934][ INFO] Iters: 1860, Total loss: 0.071, Loss x: 0.083, Loss s: 0.109, Loss w_fp: 0.011, Mask: 0.789
[2022-11-25 20:55:26,701][ INFO] Iters: 2170, Total loss: 0.072, Loss x: 0.083, Loss s: 0.111, Loss w_fp: 0.010, Mask: 0.789
[2022-11-25 20:59:03,060][ INFO] ***** Evaluation original ***** >>>> meanIOU: 72.21

[2022-11-25 20:59:03,741][ INFO] ===========> Epoch: 14, LR: 0.0004, Previous best: 72.21
[2022-11-25 20:59:04,979][ INFO] Iters: 0, Total loss: 0.110, Loss x: 0.145, Loss s: 0.132, Loss w_fp: 0.016, Mask: 0.884
[2022-11-25 21:01:36,628][ INFO] Iters: 310, Total loss: 0.069, Loss x: 0.080, Loss s: 0.106, Loss w_fp: 0.011, Mask: 0.795
[2022-11-25 21:04:08,967][ INFO] Iters: 620, Total loss: 0.069, Loss x: 0.080, Loss s: 0.108, Loss w_fp: 0.011, Mask: 0.796
[2022-11-25 21:06:40,441][ INFO] Iters: 930, Total loss: 0.068, Loss x: 0.078, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.796
[2022-11-25 21:09:12,309][ INFO] Iters: 1240, Total loss: 0.069, Loss x: 0.080, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.794
[2022-11-25 21:11:44,643][ INFO] Iters: 1550, Total loss: 0.068, Loss x: 0.079, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.794
[2022-11-25 21:14:16,856][ INFO] Iters: 1860, Total loss: 0.068, Loss x: 0.078, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.793
[2022-11-25 21:16:48,650][ INFO] Iters: 2170, Total loss: 0.069, Loss x: 0.079, Loss s: 0.108, Loss w_fp: 0.010, Mask: 0.791
[2022-11-25 21:20:24,370][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.29

[2022-11-25 21:20:25,107][ INFO] ===========> Epoch: 15, LR: 0.0004, Previous best: 73.29
[2022-11-25 21:20:26,304][ INFO] Iters: 0, Total loss: 0.046, Loss x: 0.058, Loss s: 0.051, Loss w_fp: 0.018, Mask: 0.899
[2022-11-25 21:22:57,513][ INFO] Iters: 310, Total loss: 0.067, Loss x: 0.077, Loss s: 0.103, Loss w_fp: 0.009, Mask: 0.804
[2022-11-25 21:25:29,872][ INFO] Iters: 620, Total loss: 0.067, Loss x: 0.076, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.797
[2022-11-25 21:28:01,570][ INFO] Iters: 930, Total loss: 0.069, Loss x: 0.077, Loss s: 0.112, Loss w_fp: 0.010, Mask: 0.796
[2022-11-25 21:30:33,722][ INFO] Iters: 1240, Total loss: 0.069, Loss x: 0.078, Loss s: 0.112, Loss w_fp: 0.010, Mask: 0.793
[2022-11-25 21:33:05,826][ INFO] Iters: 1550, Total loss: 0.068, Loss x: 0.077, Loss s: 0.110, Loss w_fp: 0.010, Mask: 0.795
[2022-11-25 21:35:37,127][ INFO] Iters: 1860, Total loss: 0.068, Loss x: 0.077, Loss s: 0.111, Loss w_fp: 0.010, Mask: 0.795
[2022-11-25 21:38:09,178][ INFO] Iters: 2170, Total loss: 0.069, Loss x: 0.077, Loss s: 0.110, Loss w_fp: 0.009, Mask: 0.793
[2022-11-25 21:41:44,043][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.38

[2022-11-25 21:41:44,735][ INFO] ===========> Epoch: 16, LR: 0.0004, Previous best: 73.38
[2022-11-25 21:41:45,856][ INFO] Iters: 0, Total loss: 0.032, Loss x: 0.036, Loss s: 0.054, Loss w_fp: 0.004, Mask: 0.877
[2022-11-25 21:44:17,663][ INFO] Iters: 310, Total loss: 0.070, Loss x: 0.080, Loss s: 0.112, Loss w_fp: 0.009, Mask: 0.796
[2022-11-25 21:46:49,893][ INFO] Iters: 620, Total loss: 0.067, Loss x: 0.077, Loss s: 0.105, Loss w_fp: 0.011, Mask: 0.798
[2022-11-25 21:49:21,569][ INFO] Iters: 930, Total loss: 0.067, Loss x: 0.076, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.799
[2022-11-25 21:51:53,145][ INFO] Iters: 1240, Total loss: 0.067, Loss x: 0.076, Loss s: 0.105, Loss w_fp: 0.010, Mask: 0.798
[2022-11-25 21:54:24,692][ INFO] Iters: 1550, Total loss: 0.067, Loss x: 0.076, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.801
[2022-11-25 21:56:56,705][ INFO] Iters: 1860, Total loss: 0.066, Loss x: 0.075, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.801
[2022-11-25 21:59:28,836][ INFO] Iters: 2170, Total loss: 0.066, Loss x: 0.075, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.799
[2022-11-25 22:03:05,959][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.11

[2022-11-25 22:03:05,959][ INFO] ===========> Epoch: 17, LR: 0.0004, Previous best: 73.38
[2022-11-25 22:03:07,126][ INFO] Iters: 0, Total loss: 0.060, Loss x: 0.083, Loss s: 0.064, Loss w_fp: 0.012, Mask: 0.715
[2022-11-25 22:05:38,886][ INFO] Iters: 310, Total loss: 0.064, Loss x: 0.075, Loss s: 0.097, Loss w_fp: 0.010, Mask: 0.810
[2022-11-25 22:08:10,586][ INFO] Iters: 620, Total loss: 0.065, Loss x: 0.074, Loss s: 0.103, Loss w_fp: 0.010, Mask: 0.807
[2022-11-25 22:10:42,156][ INFO] Iters: 930, Total loss: 0.067, Loss x: 0.075, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.805
[2022-11-25 22:13:13,982][ INFO] Iters: 1240, Total loss: 0.066, Loss x: 0.074, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.805
[2022-11-25 22:15:46,311][ INFO] Iters: 1550, Total loss: 0.066, Loss x: 0.073, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.807
[2022-11-25 22:18:18,647][ INFO] Iters: 1860, Total loss: 0.066, Loss x: 0.073, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.807
[2022-11-25 22:20:50,422][ INFO] Iters: 2170, Total loss: 0.066, Loss x: 0.073, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.805
[2022-11-25 22:24:26,194][ INFO] ***** Evaluation original ***** >>>> meanIOU: 72.51

[2022-11-25 22:24:26,195][ INFO] ===========> Epoch: 18, LR: 0.0004, Previous best: 73.38
[2022-11-25 22:24:27,416][ INFO] Iters: 0, Total loss: 0.075, Loss x: 0.096, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.850
[2022-11-25 22:26:59,062][ INFO] Iters: 310, Total loss: 0.063, Loss x: 0.069, Loss s: 0.104, Loss w_fp: 0.010, Mask: 0.802
[2022-11-25 22:29:31,213][ INFO] Iters: 620, Total loss: 0.063, Loss x: 0.070, Loss s: 0.101, Loss w_fp: 0.010, Mask: 0.805
[2022-11-25 22:32:03,120][ INFO] Iters: 930, Total loss: 0.065, Loss x: 0.072, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.800
[2022-11-25 22:34:34,807][ INFO] Iters: 1240, Total loss: 0.065, Loss x: 0.073, Loss s: 0.106, Loss w_fp: 0.010, Mask: 0.802
[2022-11-25 22:37:06,460][ INFO] Iters: 1550, Total loss: 0.066, Loss x: 0.073, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.803
[2022-11-25 22:39:38,286][ INFO] Iters: 1860, Total loss: 0.066, Loss x: 0.075, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.802
[2022-11-25 22:42:10,421][ INFO] Iters: 2170, Total loss: 0.067, Loss x: 0.074, Loss s: 0.108, Loss w_fp: 0.010, Mask: 0.802
[2022-11-25 22:45:47,389][ INFO] ***** Evaluation original ***** >>>> meanIOU: 72.03

[2022-11-25 22:45:47,390][ INFO] ===========> Epoch: 19, LR: 0.0004, Previous best: 73.38
[2022-11-25 22:45:48,542][ INFO] Iters: 0, Total loss: 0.044, Loss x: 0.023, Loss s: 0.126, Loss w_fp: 0.006, Mask: 0.831
[2022-11-25 22:48:20,031][ INFO] Iters: 310, Total loss: 0.061, Loss x: 0.067, Loss s: 0.103, Loss w_fp: 0.009, Mask: 0.807
[2022-11-25 22:50:51,460][ INFO] Iters: 620, Total loss: 0.062, Loss x: 0.067, Loss s: 0.105, Loss w_fp: 0.009, Mask: 0.813
[2022-11-25 22:53:23,441][ INFO] Iters: 930, Total loss: 0.063, Loss x: 0.068, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.810
[2022-11-25 22:55:55,332][ INFO] Iters: 1240, Total loss: 0.063, Loss x: 0.068, Loss s: 0.105, Loss w_fp: 0.009, Mask: 0.810
[2022-11-25 22:58:27,054][ INFO] Iters: 1550, Total loss: 0.063, Loss x: 0.068, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.810
[2022-11-25 23:00:59,185][ INFO] Iters: 1860, Total loss: 0.063, Loss x: 0.069, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.810
[2022-11-25 23:03:30,841][ INFO] Iters: 2170, Total loss: 0.063, Loss x: 0.069, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.811
[2022-11-25 23:07:07,834][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.40

[2022-11-25 23:07:08,760][ INFO] ===========> Epoch: 20, LR: 0.0004, Previous best: 73.40
[2022-11-25 23:07:10,052][ INFO] Iters: 0, Total loss: 0.109, Loss x: 0.097, Loss s: 0.241, Loss w_fp: 0.002, Mask: 0.909
[2022-11-25 23:09:41,545][ INFO] Iters: 310, Total loss: 0.064, Loss x: 0.072, Loss s: 0.102, Loss w_fp: 0.010, Mask: 0.804
[2022-11-25 23:12:13,575][ INFO] Iters: 620, Total loss: 0.063, Loss x: 0.071, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.803
[2022-11-25 23:14:45,947][ INFO] Iters: 930, Total loss: 0.062, Loss x: 0.070, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.807
[2022-11-25 23:17:18,124][ INFO] Iters: 1240, Total loss: 0.062, Loss x: 0.070, Loss s: 0.099, Loss w_fp: 0.009, Mask: 0.808
[2022-11-25 23:19:50,623][ INFO] Iters: 1550, Total loss: 0.062, Loss x: 0.071, Loss s: 0.099, Loss w_fp: 0.009, Mask: 0.811
[2022-11-25 23:22:23,298][ INFO] Iters: 1860, Total loss: 0.063, Loss x: 0.071, Loss s: 0.099, Loss w_fp: 0.009, Mask: 0.812
[2022-11-25 23:24:55,918][ INFO] Iters: 2170, Total loss: 0.063, Loss x: 0.071, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.811
[2022-11-25 23:28:33,052][ INFO] ***** Evaluation original ***** >>>> meanIOU: 72.97

[2022-11-25 23:28:33,053][ INFO] ===========> Epoch: 21, LR: 0.0004, Previous best: 73.40
[2022-11-25 23:28:34,409][ INFO] Iters: 0, Total loss: 0.164, Loss x: 0.097, Loss s: 0.416, Loss w_fp: 0.048, Mask: 0.562
[2022-11-25 23:31:06,463][ INFO] Iters: 310, Total loss: 0.065, Loss x: 0.072, Loss s: 0.106, Loss w_fp: 0.008, Mask: 0.819
[2022-11-25 23:33:38,695][ INFO] Iters: 620, Total loss: 0.065, Loss x: 0.071, Loss s: 0.108, Loss w_fp: 0.009, Mask: 0.820
[2022-11-25 23:36:10,812][ INFO] Iters: 930, Total loss: 0.063, Loss x: 0.070, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.819
[2022-11-25 23:38:43,080][ INFO] Iters: 1240, Total loss: 0.063, Loss x: 0.070, Loss s: 0.103, Loss w_fp: 0.009, Mask: 0.817
[2022-11-25 23:41:14,832][ INFO] Iters: 1550, Total loss: 0.063, Loss x: 0.069, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.816
[2022-11-25 23:43:46,867][ INFO] Iters: 1860, Total loss: 0.062, Loss x: 0.069, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.817
[2022-11-25 23:46:18,692][ INFO] Iters: 2170, Total loss: 0.062, Loss x: 0.068, Loss s: 0.102, Loss w_fp: 0.009, Mask: 0.817
[2022-11-25 23:49:56,159][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.91

[2022-11-25 23:49:56,920][ INFO] ===========> Epoch: 22, LR: 0.0004, Previous best: 73.91
[2022-11-25 23:49:58,238][ INFO] Iters: 0, Total loss: 0.044, Loss x: 0.050, Loss s: 0.072, Loss w_fp: 0.005, Mask: 0.837
[2022-11-25 23:52:30,334][ INFO] Iters: 310, Total loss: 0.062, Loss x: 0.065, Loss s: 0.107, Loss w_fp: 0.010, Mask: 0.814
[2022-11-25 23:55:02,457][ INFO] Iters: 620, Total loss: 0.060, Loss x: 0.064, Loss s: 0.103, Loss w_fp: 0.009, Mask: 0.817
[2022-11-25 23:57:34,409][ INFO] Iters: 930, Total loss: 0.061, Loss x: 0.065, Loss s: 0.105, Loss w_fp: 0.009, Mask: 0.819
[2022-11-26 00:00:06,086][ INFO] Iters: 1240, Total loss: 0.061, Loss x: 0.065, Loss s: 0.105, Loss w_fp: 0.009, Mask: 0.817
[2022-11-26 00:02:37,744][ INFO] Iters: 1550, Total loss: 0.062, Loss x: 0.066, Loss s: 0.105, Loss w_fp: 0.009, Mask: 0.818
[2022-11-26 00:05:09,704][ INFO] Iters: 1860, Total loss: 0.061, Loss x: 0.066, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.817
[2022-11-26 00:07:42,003][ INFO] Iters: 2170, Total loss: 0.061, Loss x: 0.066, Loss s: 0.103, Loss w_fp: 0.009, Mask: 0.816
[2022-11-26 00:11:18,139][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.62

[2022-11-26 00:11:18,139][ INFO] ===========> Epoch: 23, LR: 0.0004, Previous best: 73.91
[2022-11-26 00:11:19,442][ INFO] Iters: 0, Total loss: 0.109, Loss x: 0.082, Loss s: 0.261, Loss w_fp: 0.013, Mask: 0.752
[2022-11-26 00:13:51,170][ INFO] Iters: 310, Total loss: 0.058, Loss x: 0.066, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.813
[2022-11-26 00:16:23,112][ INFO] Iters: 620, Total loss: 0.057, Loss x: 0.064, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.820
[2022-11-26 00:18:55,130][ INFO] Iters: 930, Total loss: 0.058, Loss x: 0.063, Loss s: 0.095, Loss w_fp: 0.009, Mask: 0.818
[2022-11-26 00:21:26,712][ INFO] Iters: 1240, Total loss: 0.058, Loss x: 0.064, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.820
[2022-11-26 00:23:58,878][ INFO] Iters: 1550, Total loss: 0.058, Loss x: 0.065, Loss s: 0.096, Loss w_fp: 0.009, Mask: 0.820
[2022-11-26 00:26:26,518][ INFO] Iters: 1860, Total loss: 0.059, Loss x: 0.065, Loss s: 0.097, Loss w_fp: 0.009, Mask: 0.821
[2022-11-26 00:28:53,696][ INFO] Iters: 2170, Total loss: 0.059, Loss x: 0.065, Loss s: 0.097, Loss w_fp: 0.009, Mask: 0.822
[2022-11-26 00:32:15,014][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.14

[2022-11-26 00:32:15,015][ INFO] ===========> Epoch: 24, LR: 0.0004, Previous best: 73.91
[2022-11-26 00:32:16,136][ INFO] Iters: 0, Total loss: 0.022, Loss x: 0.030, Loss s: 0.024, Loss w_fp: 0.003, Mask: 0.900
[2022-11-26 00:34:43,617][ INFO] Iters: 310, Total loss: 0.056, Loss x: 0.064, Loss s: 0.087, Loss w_fp: 0.008, Mask: 0.829
[2022-11-26 00:37:15,260][ INFO] Iters: 620, Total loss: 0.056, Loss x: 0.063, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.828
[2022-11-26 00:39:46,748][ INFO] Iters: 930, Total loss: 0.057, Loss x: 0.063, Loss s: 0.094, Loss w_fp: 0.008, Mask: 0.826
[2022-11-26 00:42:18,556][ INFO] Iters: 1240, Total loss: 0.057, Loss x: 0.063, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.825
[2022-11-26 00:44:50,817][ INFO] Iters: 1550, Total loss: 0.058, Loss x: 0.063, Loss s: 0.098, Loss w_fp: 0.008, Mask: 0.825
[2022-11-26 00:47:22,182][ INFO] Iters: 1860, Total loss: 0.058, Loss x: 0.063, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.825
[2022-11-26 00:49:53,805][ INFO] Iters: 2170, Total loss: 0.058, Loss x: 0.063, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.824
[2022-11-26 00:53:30,153][ INFO] ***** Evaluation original ***** >>>> meanIOU: 72.66

[2022-11-26 00:53:30,153][ INFO] ===========> Epoch: 25, LR: 0.0004, Previous best: 73.91
[2022-11-26 00:53:31,353][ INFO] Iters: 0, Total loss: 0.062, Loss x: 0.113, Loss s: 0.018, Loss w_fp: 0.006, Mask: 0.870
[2022-11-26 00:56:02,955][ INFO] Iters: 310, Total loss: 0.059, Loss x: 0.063, Loss s: 0.102, Loss w_fp: 0.009, Mask: 0.825
[2022-11-26 00:58:34,438][ INFO] Iters: 620, Total loss: 0.060, Loss x: 0.063, Loss s: 0.104, Loss w_fp: 0.009, Mask: 0.820
[2022-11-26 01:01:06,570][ INFO] Iters: 930, Total loss: 0.059, Loss x: 0.064, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.822
[2022-11-26 01:03:37,871][ INFO] Iters: 1240, Total loss: 0.059, Loss x: 0.064, Loss s: 0.099, Loss w_fp: 0.009, Mask: 0.823
[2022-11-26 01:06:10,063][ INFO] Iters: 1550, Total loss: 0.058, Loss x: 0.063, Loss s: 0.098, Loss w_fp: 0.008, Mask: 0.824
[2022-11-26 01:08:41,765][ INFO] Iters: 1860, Total loss: 0.058, Loss x: 0.064, Loss s: 0.098, Loss w_fp: 0.009, Mask: 0.824
[2022-11-26 01:11:13,881][ INFO] Iters: 2170, Total loss: 0.059, Loss x: 0.064, Loss s: 0.098, Loss w_fp: 0.009, Mask: 0.824
[2022-11-26 01:14:49,221][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.33

[2022-11-26 01:14:49,940][ INFO] ===========> Epoch: 26, LR: 0.0004, Previous best: 74.33
[2022-11-26 01:14:51,106][ INFO] Iters: 0, Total loss: 0.052, Loss x: 0.058, Loss s: 0.075, Loss w_fp: 0.018, Mask: 0.844
[2022-11-26 01:17:22,892][ INFO] Iters: 310, Total loss: 0.056, Loss x: 0.062, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.832
[2022-11-26 01:19:55,890][ INFO] Iters: 620, Total loss: 0.055, Loss x: 0.061, Loss s: 0.091, Loss w_fp: 0.009, Mask: 0.832
[2022-11-26 01:22:27,307][ INFO] Iters: 930, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.009, Mask: 0.833
[2022-11-26 01:24:59,476][ INFO] Iters: 1240, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.009, Mask: 0.832
[2022-11-26 01:27:31,186][ INFO] Iters: 1550, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.830
[2022-11-26 01:30:03,458][ INFO] Iters: 1860, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.830
[2022-11-26 01:32:35,568][ INFO] Iters: 2170, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.009, Mask: 0.831
[2022-11-26 01:36:12,444][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.00

[2022-11-26 01:36:12,445][ INFO] ===========> Epoch: 27, LR: 0.0003, Previous best: 74.33
[2022-11-26 01:36:13,783][ INFO] Iters: 0, Total loss: 0.070, Loss x: 0.092, Loss s: 0.091, Loss w_fp: 0.006, Mask: 0.827
[2022-11-26 01:38:45,911][ INFO] Iters: 310, Total loss: 0.064, Loss x: 0.065, Loss s: 0.115, Loss w_fp: 0.010, Mask: 0.821
[2022-11-26 01:41:17,162][ INFO] Iters: 620, Total loss: 0.062, Loss x: 0.065, Loss s: 0.108, Loss w_fp: 0.009, Mask: 0.823
[2022-11-26 01:43:49,374][ INFO] Iters: 930, Total loss: 0.061, Loss x: 0.064, Loss s: 0.106, Loss w_fp: 0.009, Mask: 0.823
[2022-11-26 01:46:21,581][ INFO] Iters: 1240, Total loss: 0.060, Loss x: 0.063, Loss s: 0.103, Loss w_fp: 0.009, Mask: 0.824
[2022-11-26 01:48:52,891][ INFO] Iters: 1550, Total loss: 0.059, Loss x: 0.063, Loss s: 0.102, Loss w_fp: 0.009, Mask: 0.825
[2022-11-26 01:51:24,832][ INFO] Iters: 1860, Total loss: 0.059, Loss x: 0.063, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.826
[2022-11-26 01:53:56,355][ INFO] Iters: 2170, Total loss: 0.059, Loss x: 0.063, Loss s: 0.101, Loss w_fp: 0.009, Mask: 0.827
[2022-11-26 01:57:24,997][ INFO] ***** Evaluation original ***** >>>> meanIOU: 75.16

[2022-11-26 01:57:25,717][ INFO] ===========> Epoch: 28, LR: 0.0003, Previous best: 75.16
[2022-11-26 01:57:26,886][ INFO] Iters: 0, Total loss: 0.075, Loss x: 0.039, Loss s: 0.217, Loss w_fp: 0.003, Mask: 0.723
[2022-11-26 01:59:54,536][ INFO] Iters: 310, Total loss: 0.056, Loss x: 0.060, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 02:02:22,276][ INFO] Iters: 620, Total loss: 0.058, Loss x: 0.061, Loss s: 0.100, Loss w_fp: 0.008, Mask: 0.834
[2022-11-26 02:04:53,495][ INFO] Iters: 930, Total loss: 0.058, Loss x: 0.061, Loss s: 0.102, Loss w_fp: 0.008, Mask: 0.832
[2022-11-26 02:07:25,745][ INFO] Iters: 1240, Total loss: 0.057, Loss x: 0.061, Loss s: 0.100, Loss w_fp: 0.008, Mask: 0.831
[2022-11-26 02:09:58,255][ INFO] Iters: 1550, Total loss: 0.057, Loss x: 0.061, Loss s: 0.099, Loss w_fp: 0.008, Mask: 0.831
[2022-11-26 02:12:29,832][ INFO] Iters: 1860, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 02:15:01,580][ INFO] Iters: 2170, Total loss: 0.057, Loss x: 0.061, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.832
[2022-11-26 02:18:38,219][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.21

[2022-11-26 02:18:38,219][ INFO] ===========> Epoch: 29, LR: 0.0003, Previous best: 75.16
[2022-11-26 02:18:39,527][ INFO] Iters: 0, Total loss: 0.039, Loss x: 0.040, Loss s: 0.070, Loss w_fp: 0.008, Mask: 0.750
[2022-11-26 02:21:11,341][ INFO] Iters: 310, Total loss: 0.055, Loss x: 0.061, Loss s: 0.091, Loss w_fp: 0.009, Mask: 0.833
[2022-11-26 02:23:42,798][ INFO] Iters: 620, Total loss: 0.056, Loss x: 0.060, Loss s: 0.094, Loss w_fp: 0.008, Mask: 0.832
[2022-11-26 02:26:14,431][ INFO] Iters: 930, Total loss: 0.055, Loss x: 0.059, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 02:28:46,564][ INFO] Iters: 1240, Total loss: 0.055, Loss x: 0.059, Loss s: 0.092, Loss w_fp: 0.008, Mask: 0.836
[2022-11-26 02:31:18,480][ INFO] Iters: 1550, Total loss: 0.055, Loss x: 0.059, Loss s: 0.094, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 02:33:50,321][ INFO] Iters: 1860, Total loss: 0.054, Loss x: 0.058, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.835
[2022-11-26 02:36:22,296][ INFO] Iters: 2170, Total loss: 0.055, Loss x: 0.059, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.834
[2022-11-26 02:39:58,409][ INFO] ***** Evaluation original ***** >>>> meanIOU: 75.50

[2022-11-26 02:39:59,175][ INFO] ===========> Epoch: 30, LR: 0.0003, Previous best: 75.50
[2022-11-26 02:40:00,395][ INFO] Iters: 0, Total loss: 0.050, Loss x: 0.060, Loss s: 0.073, Loss w_fp: 0.007, Mask: 0.738
[2022-11-26 02:42:32,158][ INFO] Iters: 310, Total loss: 0.056, Loss x: 0.059, Loss s: 0.097, Loss w_fp: 0.007, Mask: 0.836
[2022-11-26 02:45:04,883][ INFO] Iters: 620, Total loss: 0.055, Loss x: 0.058, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.838
[2022-11-26 02:47:37,097][ INFO] Iters: 930, Total loss: 0.054, Loss x: 0.058, Loss s: 0.094, Loss w_fp: 0.007, Mask: 0.838
[2022-11-26 02:50:08,985][ INFO] Iters: 1240, Total loss: 0.056, Loss x: 0.061, Loss s: 0.094, Loss w_fp: 0.007, Mask: 0.834
[2022-11-26 02:52:40,988][ INFO] Iters: 1550, Total loss: 0.057, Loss x: 0.063, Loss s: 0.096, Loss w_fp: 0.008, Mask: 0.831
[2022-11-26 02:55:12,729][ INFO] Iters: 1860, Total loss: 0.058, Loss x: 0.064, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.828
[2022-11-26 02:57:44,575][ INFO] Iters: 2170, Total loss: 0.057, Loss x: 0.063, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.829
[2022-11-26 03:01:19,901][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.61

[2022-11-26 03:01:19,901][ INFO] ===========> Epoch: 31, LR: 0.0003, Previous best: 75.50
[2022-11-26 03:01:21,159][ INFO] Iters: 0, Total loss: 0.050, Loss x: 0.049, Loss s: 0.099, Loss w_fp: 0.005, Mask: 0.923
[2022-11-26 03:03:52,176][ INFO] Iters: 310, Total loss: 0.056, Loss x: 0.059, Loss s: 0.097, Loss w_fp: 0.008, Mask: 0.829
[2022-11-26 03:06:23,637][ INFO] Iters: 620, Total loss: 0.056, Loss x: 0.061, Loss s: 0.095, Loss w_fp: 0.008, Mask: 0.830
[2022-11-26 03:08:55,391][ INFO] Iters: 930, Total loss: 0.056, Loss x: 0.061, Loss s: 0.094, Loss w_fp: 0.008, Mask: 0.831
[2022-11-26 03:11:26,776][ INFO] Iters: 1240, Total loss: 0.056, Loss x: 0.061, Loss s: 0.094, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 03:13:58,162][ INFO] Iters: 1550, Total loss: 0.057, Loss x: 0.061, Loss s: 0.096, Loss w_fp: 0.008, Mask: 0.832
[2022-11-26 03:16:29,520][ INFO] Iters: 1860, Total loss: 0.057, Loss x: 0.061, Loss s: 0.096, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 03:19:01,022][ INFO] Iters: 2170, Total loss: 0.056, Loss x: 0.061, Loss s: 0.094, Loss w_fp: 0.008, Mask: 0.834
[2022-11-26 03:22:35,723][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.38

[2022-11-26 03:22:35,723][ INFO] ===========> Epoch: 32, LR: 0.0003, Previous best: 75.50
[2022-11-26 03:22:36,886][ INFO] Iters: 0, Total loss: 0.066, Loss x: 0.079, Loss s: 0.097, Loss w_fp: 0.009, Mask: 0.688
[2022-11-26 03:25:04,079][ INFO] Iters: 310, Total loss: 0.053, Loss x: 0.058, Loss s: 0.088, Loss w_fp: 0.008, Mask: 0.832
[2022-11-26 03:27:31,214][ INFO] Iters: 620, Total loss: 0.053, Loss x: 0.057, Loss s: 0.090, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 03:29:58,622][ INFO] Iters: 930, Total loss: 0.053, Loss x: 0.058, Loss s: 0.090, Loss w_fp: 0.008, Mask: 0.834
[2022-11-26 03:32:30,543][ INFO] Iters: 1240, Total loss: 0.054, Loss x: 0.057, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.836
[2022-11-26 03:35:02,917][ INFO] Iters: 1550, Total loss: 0.054, Loss x: 0.057, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.838
[2022-11-26 03:37:35,553][ INFO] Iters: 1860, Total loss: 0.054, Loss x: 0.057, Loss s: 0.092, Loss w_fp: 0.008, Mask: 0.839
[2022-11-26 03:40:07,923][ INFO] Iters: 2170, Total loss: 0.053, Loss x: 0.057, Loss s: 0.092, Loss w_fp: 0.008, Mask: 0.839
[2022-11-26 03:43:45,001][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.81

[2022-11-26 03:43:45,001][ INFO] ===========> Epoch: 33, LR: 0.0003, Previous best: 75.50
[2022-11-26 03:43:46,280][ INFO] Iters: 0, Total loss: 0.046, Loss x: 0.062, Loss s: 0.056, Loss w_fp: 0.004, Mask: 0.851
[2022-11-26 03:46:18,370][ INFO] Iters: 310, Total loss: 0.049, Loss x: 0.053, Loss s: 0.084, Loss w_fp: 0.008, Mask: 0.847
[2022-11-26 03:48:50,543][ INFO] Iters: 620, Total loss: 0.050, Loss x: 0.055, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.848
[2022-11-26 03:51:22,944][ INFO] Iters: 930, Total loss: 0.051, Loss x: 0.055, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.845
[2022-11-26 03:53:55,383][ INFO] Iters: 1240, Total loss: 0.051, Loss x: 0.055, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.844
[2022-11-26 03:56:27,481][ INFO] Iters: 1550, Total loss: 0.052, Loss x: 0.055, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.843
[2022-11-26 03:58:59,236][ INFO] Iters: 1860, Total loss: 0.052, Loss x: 0.055, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.844
[2022-11-26 04:01:31,274][ INFO] Iters: 2170, Total loss: 0.052, Loss x: 0.056, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.844
[2022-11-26 04:05:08,973][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.42

[2022-11-26 04:05:08,974][ INFO] ===========> Epoch: 34, LR: 0.0003, Previous best: 75.50
[2022-11-26 04:05:10,274][ INFO] Iters: 0, Total loss: 0.017, Loss x: 0.023, Loss s: 0.019, Loss w_fp: 0.005, Mask: 0.923
[2022-11-26 04:07:42,168][ INFO] Iters: 310, Total loss: 0.055, Loss x: 0.059, Loss s: 0.092, Loss w_fp: 0.009, Mask: 0.834
[2022-11-26 04:10:14,693][ INFO] Iters: 620, Total loss: 0.055, Loss x: 0.060, Loss s: 0.092, Loss w_fp: 0.008, Mask: 0.833
[2022-11-26 04:12:46,611][ INFO] Iters: 930, Total loss: 0.054, Loss x: 0.059, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.837
[2022-11-26 04:15:18,449][ INFO] Iters: 1240, Total loss: 0.054, Loss x: 0.058, Loss s: 0.091, Loss w_fp: 0.007, Mask: 0.837
[2022-11-26 04:17:50,212][ INFO] Iters: 1550, Total loss: 0.053, Loss x: 0.058, Loss s: 0.091, Loss w_fp: 0.007, Mask: 0.838
[2022-11-26 04:20:22,630][ INFO] Iters: 1860, Total loss: 0.054, Loss x: 0.057, Loss s: 0.092, Loss w_fp: 0.007, Mask: 0.838
[2022-11-26 04:22:54,465][ INFO] Iters: 2170, Total loss: 0.054, Loss x: 0.057, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.837
[2022-11-26 04:26:31,838][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.96

[2022-11-26 04:26:31,838][ INFO] ===========> Epoch: 35, LR: 0.0003, Previous best: 75.50
[2022-11-26 04:26:33,074][ INFO] Iters: 0, Total loss: 0.068, Loss x: 0.064, Loss s: 0.129, Loss w_fp: 0.016, Mask: 0.818
[2022-11-26 04:29:05,422][ INFO] Iters: 310, Total loss: 0.052, Loss x: 0.056, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.841
[2022-11-26 04:31:37,662][ INFO] Iters: 620, Total loss: 0.053, Loss x: 0.058, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.839
[2022-11-26 04:34:10,203][ INFO] Iters: 930, Total loss: 0.053, Loss x: 0.057, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.840
[2022-11-26 04:36:42,221][ INFO] Iters: 1240, Total loss: 0.052, Loss x: 0.056, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.841
[2022-11-26 04:39:14,643][ INFO] Iters: 1550, Total loss: 0.052, Loss x: 0.056, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.842
[2022-11-26 04:41:46,537][ INFO] Iters: 1860, Total loss: 0.052, Loss x: 0.056, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.843
[2022-11-26 04:44:19,157][ INFO] Iters: 2170, Total loss: 0.051, Loss x: 0.055, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.843
[2022-11-26 04:47:55,833][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.02

[2022-11-26 04:47:55,834][ INFO] ===========> Epoch: 36, LR: 0.0003, Previous best: 75.50
[2022-11-26 04:47:56,947][ INFO] Iters: 0, Total loss: 0.043, Loss x: 0.053, Loss s: 0.064, Loss w_fp: 0.004, Mask: 0.848
[2022-11-26 04:50:29,061][ INFO] Iters: 310, Total loss: 0.052, Loss x: 0.056, Loss s: 0.087, Loss w_fp: 0.008, Mask: 0.841
[2022-11-26 04:52:56,854][ INFO] Iters: 620, Total loss: 0.052, Loss x: 0.054, Loss s: 0.090, Loss w_fp: 0.008, Mask: 0.841
[2022-11-26 04:55:24,589][ INFO] Iters: 930, Total loss: 0.052, Loss x: 0.055, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.841
[2022-11-26 04:57:52,093][ INFO] Iters: 1240, Total loss: 0.052, Loss x: 0.054, Loss s: 0.090, Loss w_fp: 0.007, Mask: 0.842
[2022-11-26 05:00:19,320][ INFO] Iters: 1550, Total loss: 0.051, Loss x: 0.054, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.843
[2022-11-26 05:02:46,526][ INFO] Iters: 1860, Total loss: 0.051, Loss x: 0.054, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.843
[2022-11-26 05:05:13,866][ INFO] Iters: 2170, Total loss: 0.051, Loss x: 0.055, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.844
[2022-11-26 05:08:39,956][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.34

[2022-11-26 05:08:39,956][ INFO] ===========> Epoch: 37, LR: 0.0003, Previous best: 75.50
[2022-11-26 05:08:41,038][ INFO] Iters: 0, Total loss: 0.055, Loss x: 0.042, Loss s: 0.135, Loss w_fp: 0.002, Mask: 0.713
[2022-11-26 05:11:08,250][ INFO] Iters: 310, Total loss: 0.050, Loss x: 0.054, Loss s: 0.081, Loss w_fp: 0.008, Mask: 0.837
[2022-11-26 05:13:35,415][ INFO] Iters: 620, Total loss: 0.051, Loss x: 0.053, Loss s: 0.088, Loss w_fp: 0.008, Mask: 0.841
[2022-11-26 05:16:02,777][ INFO] Iters: 930, Total loss: 0.051, Loss x: 0.053, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.845
[2022-11-26 05:18:29,924][ INFO] Iters: 1240, Total loss: 0.051, Loss x: 0.053, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.847
[2022-11-26 05:20:57,195][ INFO] Iters: 1550, Total loss: 0.051, Loss x: 0.053, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.846
[2022-11-26 05:23:24,363][ INFO] Iters: 1860, Total loss: 0.052, Loss x: 0.054, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.844
[2022-11-26 05:25:51,538][ INFO] Iters: 2170, Total loss: 0.052, Loss x: 0.055, Loss s: 0.090, Loss w_fp: 0.008, Mask: 0.844
[2022-11-26 05:29:15,612][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.90

[2022-11-26 05:29:15,613][ INFO] ===========> Epoch: 38, LR: 0.0003, Previous best: 75.50
[2022-11-26 05:29:16,730][ INFO] Iters: 0, Total loss: 0.072, Loss x: 0.073, Loss s: 0.138, Loss w_fp: 0.003, Mask: 0.870
[2022-11-26 05:31:43,928][ INFO] Iters: 310, Total loss: 0.050, Loss x: 0.052, Loss s: 0.090, Loss w_fp: 0.007, Mask: 0.841
[2022-11-26 05:34:11,015][ INFO] Iters: 620, Total loss: 0.051, Loss x: 0.054, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.841
[2022-11-26 05:36:38,046][ INFO] Iters: 930, Total loss: 0.051, Loss x: 0.054, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.842
[2022-11-26 05:39:05,319][ INFO] Iters: 1240, Total loss: 0.051, Loss x: 0.053, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.843
[2022-11-26 05:41:32,492][ INFO] Iters: 1550, Total loss: 0.051, Loss x: 0.053, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.844
[2022-11-26 05:43:59,655][ INFO] Iters: 1860, Total loss: 0.051, Loss x: 0.054, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.846
[2022-11-26 05:46:26,853][ INFO] Iters: 2170, Total loss: 0.051, Loss x: 0.053, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.848
[2022-11-26 05:49:50,750][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.73

[2022-11-26 05:49:50,750][ INFO] ===========> Epoch: 39, LR: 0.0003, Previous best: 75.50
[2022-11-26 05:49:51,794][ INFO] Iters: 0, Total loss: 0.037, Loss x: 0.023, Loss s: 0.093, Loss w_fp: 0.010, Mask: 0.905
[2022-11-26 05:52:19,005][ INFO] Iters: 310, Total loss: 0.048, Loss x: 0.052, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.846
[2022-11-26 05:54:46,391][ INFO] Iters: 620, Total loss: 0.051, Loss x: 0.053, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.844
[2022-11-26 05:57:13,646][ INFO] Iters: 930, Total loss: 0.051, Loss x: 0.053, Loss s: 0.091, Loss w_fp: 0.008, Mask: 0.844
[2022-11-26 05:59:41,046][ INFO] Iters: 1240, Total loss: 0.051, Loss x: 0.054, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.849
[2022-11-26 06:02:08,181][ INFO] Iters: 1550, Total loss: 0.051, Loss x: 0.053, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.850
[2022-11-26 06:04:35,452][ INFO] Iters: 1860, Total loss: 0.051, Loss x: 0.053, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.849
[2022-11-26 06:07:02,763][ INFO] Iters: 2170, Total loss: 0.051, Loss x: 0.054, Loss s: 0.089, Loss w_fp: 0.007, Mask: 0.849
[2022-11-26 06:10:26,666][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.32

[2022-11-26 06:10:26,666][ INFO] ===========> Epoch: 40, LR: 0.0003, Previous best: 75.50
[2022-11-26 06:10:27,756][ INFO] Iters: 0, Total loss: 0.117, Loss x: 0.091, Loss s: 0.276, Loss w_fp: 0.011, Mask: 0.765
[2022-11-26 06:12:54,928][ INFO] Iters: 310, Total loss: 0.050, Loss x: 0.052, Loss s: 0.087, Loss w_fp: 0.009, Mask: 0.846
[2022-11-26 06:15:22,432][ INFO] Iters: 620, Total loss: 0.050, Loss x: 0.052, Loss s: 0.086, Loss w_fp: 0.009, Mask: 0.846
[2022-11-26 06:17:49,693][ INFO] Iters: 930, Total loss: 0.049, Loss x: 0.051, Loss s: 0.087, Loss w_fp: 0.008, Mask: 0.847
[2022-11-26 06:20:17,024][ INFO] Iters: 1240, Total loss: 0.050, Loss x: 0.052, Loss s: 0.090, Loss w_fp: 0.008, Mask: 0.847
[2022-11-26 06:22:44,320][ INFO] Iters: 1550, Total loss: 0.050, Loss x: 0.052, Loss s: 0.088, Loss w_fp: 0.008, Mask: 0.849
[2022-11-26 06:25:11,534][ INFO] Iters: 1860, Total loss: 0.050, Loss x: 0.052, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.848
[2022-11-26 06:27:38,850][ INFO] Iters: 2170, Total loss: 0.050, Loss x: 0.052, Loss s: 0.089, Loss w_fp: 0.008, Mask: 0.849
[2022-11-26 06:31:02,729][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.43

[2022-11-26 06:31:02,729][ INFO] ===========> Epoch: 41, LR: 0.0003, Previous best: 75.50
[2022-11-26 06:31:03,893][ INFO] Iters: 0, Total loss: 0.033, Loss x: 0.041, Loss s: 0.050, Loss w_fp: 0.001, Mask: 0.920
[2022-11-26 06:33:31,166][ INFO] Iters: 310, Total loss: 0.053, Loss x: 0.055, Loss s: 0.093, Loss w_fp: 0.008, Mask: 0.849
[2022-11-26 06:35:58,148][ INFO] Iters: 620, Total loss: 0.051, Loss x: 0.053, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.846
[2022-11-26 06:38:24,089][ INFO] Iters: 930, Total loss: 0.050, Loss x: 0.053, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.848
[2022-11-26 06:40:49,982][ INFO] Iters: 1240, Total loss: 0.049, Loss x: 0.053, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.850
[2022-11-26 06:43:15,828][ INFO] Iters: 1550, Total loss: 0.049, Loss x: 0.052, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.851
[2022-11-26 06:45:41,294][ INFO] Iters: 1860, Total loss: 0.050, Loss x: 0.053, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.850
[2022-11-26 06:48:06,862][ INFO] Iters: 2170, Total loss: 0.049, Loss x: 0.052, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.852
[2022-11-26 06:51:17,615][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.23

[2022-11-26 06:51:17,615][ INFO] ===========> Epoch: 42, LR: 0.0003, Previous best: 75.50
[2022-11-26 06:51:18,636][ INFO] Iters: 0, Total loss: 0.082, Loss x: 0.122, Loss s: 0.080, Loss w_fp: 0.003, Mask: 0.934
[2022-11-26 06:53:44,099][ INFO] Iters: 310, Total loss: 0.046, Loss x: 0.051, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.860
[2022-11-26 06:56:09,596][ INFO] Iters: 620, Total loss: 0.047, Loss x: 0.050, Loss s: 0.080, Loss w_fp: 0.006, Mask: 0.861
[2022-11-26 06:58:35,110][ INFO] Iters: 930, Total loss: 0.047, Loss x: 0.050, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 07:01:00,495][ INFO] Iters: 1240, Total loss: 0.048, Loss x: 0.050, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.855
[2022-11-26 07:03:25,899][ INFO] Iters: 1550, Total loss: 0.048, Loss x: 0.051, Loss s: 0.084, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 07:05:51,458][ INFO] Iters: 1860, Total loss: 0.048, Loss x: 0.051, Loss s: 0.084, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 07:08:16,978][ INFO] Iters: 2170, Total loss: 0.048, Loss x: 0.051, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 07:11:28,024][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.96

[2022-11-26 07:11:28,024][ INFO] ===========> Epoch: 43, LR: 0.0002, Previous best: 75.50
[2022-11-26 07:11:28,965][ INFO] Iters: 0, Total loss: 0.066, Loss x: 0.053, Loss s: 0.147, Loss w_fp: 0.008, Mask: 0.662
[2022-11-26 07:13:54,466][ INFO] Iters: 310, Total loss: 0.050, Loss x: 0.052, Loss s: 0.087, Loss w_fp: 0.008, Mask: 0.853
[2022-11-26 07:16:19,959][ INFO] Iters: 620, Total loss: 0.049, Loss x: 0.052, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.855
[2022-11-26 07:18:45,415][ INFO] Iters: 930, Total loss: 0.049, Loss x: 0.052, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.856
[2022-11-26 07:21:10,911][ INFO] Iters: 1240, Total loss: 0.049, Loss x: 0.051, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.856
[2022-11-26 07:23:36,492][ INFO] Iters: 1550, Total loss: 0.049, Loss x: 0.052, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.856
[2022-11-26 07:26:02,107][ INFO] Iters: 1860, Total loss: 0.049, Loss x: 0.052, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.854
[2022-11-26 07:28:27,663][ INFO] Iters: 2170, Total loss: 0.049, Loss x: 0.052, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.853
[2022-11-26 07:31:38,971][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.28

[2022-11-26 07:31:38,971][ INFO] ===========> Epoch: 44, LR: 0.0002, Previous best: 75.50
[2022-11-26 07:31:39,927][ INFO] Iters: 0, Total loss: 0.028, Loss x: 0.043, Loss s: 0.023, Loss w_fp: 0.003, Mask: 0.564
[2022-11-26 07:34:05,276][ INFO] Iters: 310, Total loss: 0.051, Loss x: 0.052, Loss s: 0.092, Loss w_fp: 0.007, Mask: 0.847
[2022-11-26 07:36:30,847][ INFO] Iters: 620, Total loss: 0.050, Loss x: 0.051, Loss s: 0.092, Loss w_fp: 0.007, Mask: 0.848
[2022-11-26 07:38:56,425][ INFO] Iters: 930, Total loss: 0.049, Loss x: 0.051, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.852
[2022-11-26 07:41:21,879][ INFO] Iters: 1240, Total loss: 0.048, Loss x: 0.051, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.852
[2022-11-26 07:43:47,336][ INFO] Iters: 1550, Total loss: 0.049, Loss x: 0.050, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.852
[2022-11-26 07:46:12,760][ INFO] Iters: 1860, Total loss: 0.049, Loss x: 0.051, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.853
[2022-11-26 07:48:38,283][ INFO] Iters: 2170, Total loss: 0.049, Loss x: 0.051, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.853
[2022-11-26 07:51:50,323][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.96

[2022-11-26 07:51:50,323][ INFO] ===========> Epoch: 45, LR: 0.0002, Previous best: 75.50
[2022-11-26 07:51:51,311][ INFO] Iters: 0, Total loss: 0.048, Loss x: 0.041, Loss s: 0.105, Loss w_fp: 0.005, Mask: 0.853
[2022-11-26 07:54:16,848][ INFO] Iters: 310, Total loss: 0.049, Loss x: 0.051, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.856
[2022-11-26 07:56:42,407][ INFO] Iters: 620, Total loss: 0.048, Loss x: 0.050, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 07:59:07,930][ INFO] Iters: 930, Total loss: 0.049, Loss x: 0.050, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 08:01:33,449][ INFO] Iters: 1240, Total loss: 0.049, Loss x: 0.050, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 08:03:59,016][ INFO] Iters: 1550, Total loss: 0.049, Loss x: 0.050, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:06:24,576][ INFO] Iters: 1860, Total loss: 0.049, Loss x: 0.050, Loss s: 0.087, Loss w_fp: 0.008, Mask: 0.858
[2022-11-26 08:08:50,084][ INFO] Iters: 2170, Total loss: 0.048, Loss x: 0.050, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.858
[2022-11-26 08:12:01,618][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.92

[2022-11-26 08:12:01,619][ INFO] ===========> Epoch: 46, LR: 0.0002, Previous best: 75.50
[2022-11-26 08:12:02,567][ INFO] Iters: 0, Total loss: 0.041, Loss x: 0.043, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.894
[2022-11-26 08:14:28,025][ INFO] Iters: 310, Total loss: 0.048, Loss x: 0.049, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 08:16:53,576][ INFO] Iters: 620, Total loss: 0.048, Loss x: 0.049, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:19:19,101][ INFO] Iters: 930, Total loss: 0.048, Loss x: 0.049, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:21:44,638][ INFO] Iters: 1240, Total loss: 0.047, Loss x: 0.049, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:24:10,203][ INFO] Iters: 1550, Total loss: 0.047, Loss x: 0.049, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:26:35,756][ INFO] Iters: 1860, Total loss: 0.047, Loss x: 0.049, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 08:29:01,371][ INFO] Iters: 2170, Total loss: 0.047, Loss x: 0.050, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.858
[2022-11-26 08:32:11,983][ INFO] ***** Evaluation original ***** >>>> meanIOU: 75.14

[2022-11-26 08:32:11,983][ INFO] ===========> Epoch: 47, LR: 0.0002, Previous best: 75.50
[2022-11-26 08:32:12,975][ INFO] Iters: 0, Total loss: 0.058, Loss x: 0.101, Loss s: 0.024, Loss w_fp: 0.004, Mask: 0.937
[2022-11-26 08:34:38,364][ INFO] Iters: 310, Total loss: 0.047, Loss x: 0.050, Loss s: 0.080, Loss w_fp: 0.006, Mask: 0.860
[2022-11-26 08:37:03,967][ INFO] Iters: 620, Total loss: 0.048, Loss x: 0.050, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:39:29,659][ INFO] Iters: 930, Total loss: 0.047, Loss x: 0.050, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 08:41:55,223][ INFO] Iters: 1240, Total loss: 0.047, Loss x: 0.050, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 08:44:20,985][ INFO] Iters: 1550, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 08:46:46,526][ INFO] Iters: 1860, Total loss: 0.048, Loss x: 0.050, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 08:49:12,020][ INFO] Iters: 2170, Total loss: 0.047, Loss x: 0.050, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.858
[2022-11-26 08:52:22,672][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.65

[2022-11-26 08:52:22,672][ INFO] ===========> Epoch: 48, LR: 0.0002, Previous best: 75.50
[2022-11-26 08:52:23,697][ INFO] Iters: 0, Total loss: 0.048, Loss x: 0.050, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.909
[2022-11-26 08:54:49,490][ INFO] Iters: 310, Total loss: 0.046, Loss x: 0.050, Loss s: 0.078, Loss w_fp: 0.006, Mask: 0.864
[2022-11-26 08:57:15,084][ INFO] Iters: 620, Total loss: 0.046, Loss x: 0.050, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.863
[2022-11-26 08:59:40,743][ INFO] Iters: 930, Total loss: 0.048, Loss x: 0.052, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 09:02:06,297][ INFO] Iters: 1240, Total loss: 0.049, Loss x: 0.051, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.855
[2022-11-26 09:04:32,002][ INFO] Iters: 1550, Total loss: 0.048, Loss x: 0.051, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.854
[2022-11-26 09:06:57,657][ INFO] Iters: 1860, Total loss: 0.048, Loss x: 0.050, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.854
[2022-11-26 09:09:23,299][ INFO] Iters: 2170, Total loss: 0.049, Loss x: 0.050, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.855
[2022-11-26 09:12:35,224][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.26

[2022-11-26 09:12:35,225][ INFO] ===========> Epoch: 49, LR: 0.0002, Previous best: 75.50
[2022-11-26 09:12:36,191][ INFO] Iters: 0, Total loss: 0.043, Loss x: 0.026, Loss s: 0.118, Loss w_fp: 0.003, Mask: 0.927
[2022-11-26 09:15:01,760][ INFO] Iters: 310, Total loss: 0.046, Loss x: 0.049, Loss s: 0.081, Loss w_fp: 0.006, Mask: 0.862
[2022-11-26 09:17:27,337][ INFO] Iters: 620, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 09:19:52,931][ INFO] Iters: 930, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 09:22:18,609][ INFO] Iters: 1240, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 09:24:44,144][ INFO] Iters: 1550, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 09:27:09,714][ INFO] Iters: 1860, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 09:29:35,470][ INFO] Iters: 2170, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.863
[2022-11-26 09:32:47,801][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.28

[2022-11-26 09:32:47,801][ INFO] ===========> Epoch: 50, LR: 0.0002, Previous best: 75.50
[2022-11-26 09:32:48,807][ INFO] Iters: 0, Total loss: 0.040, Loss x: 0.063, Loss s: 0.032, Loss w_fp: 0.001, Mask: 0.884
[2022-11-26 09:35:14,531][ INFO] Iters: 310, Total loss: 0.046, Loss x: 0.049, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.868
[2022-11-26 09:37:40,257][ INFO] Iters: 620, Total loss: 0.046, Loss x: 0.048, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.862
[2022-11-26 09:40:06,319][ INFO] Iters: 930, Total loss: 0.046, Loss x: 0.047, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 09:42:32,331][ INFO] Iters: 1240, Total loss: 0.047, Loss x: 0.048, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 09:44:58,249][ INFO] Iters: 1550, Total loss: 0.047, Loss x: 0.048, Loss s: 0.085, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 09:47:24,310][ INFO] Iters: 1860, Total loss: 0.046, Loss x: 0.047, Loss s: 0.084, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 09:49:50,410][ INFO] Iters: 2170, Total loss: 0.047, Loss x: 0.047, Loss s: 0.086, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 09:53:03,597][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.61

[2022-11-26 09:53:03,597][ INFO] ===========> Epoch: 51, LR: 0.0002, Previous best: 75.50
[2022-11-26 09:53:04,569][ INFO] Iters: 0, Total loss: 0.028, Loss x: 0.046, Loss s: 0.016, Loss w_fp: 0.003, Mask: 0.708
[2022-11-26 09:55:30,218][ INFO] Iters: 310, Total loss: 0.048, Loss x: 0.049, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.858
[2022-11-26 09:57:56,160][ INFO] Iters: 620, Total loss: 0.049, Loss x: 0.050, Loss s: 0.088, Loss w_fp: 0.007, Mask: 0.855
[2022-11-26 10:00:21,889][ INFO] Iters: 930, Total loss: 0.048, Loss x: 0.049, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 10:02:47,769][ INFO] Iters: 1240, Total loss: 0.047, Loss x: 0.048, Loss s: 0.084, Loss w_fp: 0.007, Mask: 0.857
[2022-11-26 10:05:13,434][ INFO] Iters: 1550, Total loss: 0.046, Loss x: 0.048, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.859
[2022-11-26 10:07:39,109][ INFO] Iters: 1860, Total loss: 0.046, Loss x: 0.047, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 10:10:04,923][ INFO] Iters: 2170, Total loss: 0.046, Loss x: 0.048, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 10:13:17,711][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.48

[2022-11-26 10:13:17,711][ INFO] ===========> Epoch: 52, LR: 0.0002, Previous best: 75.50
[2022-11-26 10:13:18,711][ INFO] Iters: 0, Total loss: 0.050, Loss x: 0.068, Loss s: 0.056, Loss w_fp: 0.008, Mask: 0.751
[2022-11-26 10:15:44,407][ INFO] Iters: 310, Total loss: 0.045, Loss x: 0.048, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.869
[2022-11-26 10:18:10,372][ INFO] Iters: 620, Total loss: 0.046, Loss x: 0.048, Loss s: 0.081, Loss w_fp: 0.006, Mask: 0.862
[2022-11-26 10:20:36,209][ INFO] Iters: 930, Total loss: 0.046, Loss x: 0.048, Loss s: 0.082, Loss w_fp: 0.006, Mask: 0.862
[2022-11-26 10:23:01,994][ INFO] Iters: 1240, Total loss: 0.046, Loss x: 0.048, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 10:25:27,779][ INFO] Iters: 1550, Total loss: 0.046, Loss x: 0.048, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.861
[2022-11-26 10:27:53,873][ INFO] Iters: 1860, Total loss: 0.047, Loss x: 0.049, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.862
[2022-11-26 10:30:19,787][ INFO] Iters: 2170, Total loss: 0.046, Loss x: 0.049, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.863
[2022-11-26 10:33:32,393][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.49

[2022-11-26 10:33:32,393][ INFO] ===========> Epoch: 53, LR: 0.0002, Previous best: 75.50
[2022-11-26 10:33:33,332][ INFO] Iters: 0, Total loss: 0.026, Loss x: 0.028, Loss s: 0.044, Loss w_fp: 0.003, Mask: 0.903
[2022-11-26 10:35:58,714][ INFO] Iters: 310, Total loss: 0.045, Loss x: 0.048, Loss s: 0.078, Loss w_fp: 0.006, Mask: 0.869
[2022-11-26 10:38:24,337][ INFO] Iters: 620, Total loss: 0.045, Loss x: 0.047, Loss s: 0.080, Loss w_fp: 0.006, Mask: 0.868
[2022-11-26 10:40:50,150][ INFO] Iters: 930, Total loss: 0.045, Loss x: 0.046, Loss s: 0.081, Loss w_fp: 0.006, Mask: 0.865
[2022-11-26 10:43:16,044][ INFO] Iters: 1240, Total loss: 0.045, Loss x: 0.047, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 10:45:41,914][ INFO] Iters: 1550, Total loss: 0.045, Loss x: 0.047, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 10:48:07,904][ INFO] Iters: 1860, Total loss: 0.045, Loss x: 0.047, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.864
[2022-11-26 10:50:33,652][ INFO] Iters: 2170, Total loss: 0.045, Loss x: 0.047, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.864
[2022-11-26 10:53:45,375][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.35

[2022-11-26 10:53:45,375][ INFO] ===========> Epoch: 54, LR: 0.0002, Previous best: 75.50
[2022-11-26 10:53:46,323][ INFO] Iters: 0, Total loss: 0.032, Loss x: 0.040, Loss s: 0.045, Loss w_fp: 0.002, Mask: 0.881
[2022-11-26 10:56:11,853][ INFO] Iters: 310, Total loss: 0.045, Loss x: 0.048, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.862
[2022-11-26 10:58:37,966][ INFO] Iters: 620, Total loss: 0.045, Loss x: 0.047, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.862
[2022-11-26 11:01:04,017][ INFO] Iters: 930, Total loss: 0.045, Loss x: 0.047, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 11:03:30,321][ INFO] Iters: 1240, Total loss: 0.044, Loss x: 0.046, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.866
[2022-11-26 11:05:56,510][ INFO] Iters: 1550, Total loss: 0.044, Loss x: 0.046, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.867
[2022-11-26 11:08:22,646][ INFO] Iters: 1860, Total loss: 0.045, Loss x: 0.047, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 11:10:48,504][ INFO] Iters: 2170, Total loss: 0.045, Loss x: 0.047, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.864
[2022-11-26 11:14:01,102][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.29

[2022-11-26 11:14:01,103][ INFO] ===========> Epoch: 55, LR: 0.0002, Previous best: 75.50
[2022-11-26 11:14:02,100][ INFO] Iters: 0, Total loss: 0.025, Loss x: 0.034, Loss s: 0.030, Loss w_fp: 0.001, Mask: 0.925
[2022-11-26 11:16:27,599][ INFO] Iters: 310, Total loss: 0.045, Loss x: 0.048, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.861
[2022-11-26 11:18:53,285][ INFO] Iters: 620, Total loss: 0.045, Loss x: 0.048, Loss s: 0.076, Loss w_fp: 0.006, Mask: 0.861
[2022-11-26 11:21:18,824][ INFO] Iters: 930, Total loss: 0.045, Loss x: 0.047, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.862
[2022-11-26 11:23:44,518][ INFO] Iters: 1240, Total loss: 0.044, Loss x: 0.047, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.863
[2022-11-26 11:26:10,227][ INFO] Iters: 1550, Total loss: 0.045, Loss x: 0.047, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.862
[2022-11-26 11:28:35,855][ INFO] Iters: 1860, Total loss: 0.045, Loss x: 0.047, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.863
[2022-11-26 11:31:01,421][ INFO] Iters: 2170, Total loss: 0.045, Loss x: 0.047, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.864
[2022-11-26 11:34:13,650][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.33

[2022-11-26 11:34:13,650][ INFO] ===========> Epoch: 56, LR: 0.0002, Previous best: 75.50
[2022-11-26 11:34:14,687][ INFO] Iters: 0, Total loss: 0.035, Loss x: 0.037, Loss s: 0.061, Loss w_fp: 0.005, Mask: 0.893
[2022-11-26 11:36:40,506][ INFO] Iters: 310, Total loss: 0.046, Loss x: 0.048, Loss s: 0.083, Loss w_fp: 0.006, Mask: 0.865
[2022-11-26 11:39:06,332][ INFO] Iters: 620, Total loss: 0.045, Loss x: 0.047, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 11:41:32,073][ INFO] Iters: 930, Total loss: 0.045, Loss x: 0.046, Loss s: 0.083, Loss w_fp: 0.006, Mask: 0.866
[2022-11-26 11:43:57,916][ INFO] Iters: 1240, Total loss: 0.045, Loss x: 0.046, Loss s: 0.082, Loss w_fp: 0.007, Mask: 0.867
[2022-11-26 11:46:23,908][ INFO] Iters: 1550, Total loss: 0.045, Loss x: 0.046, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.867
[2022-11-26 11:48:49,777][ INFO] Iters: 1860, Total loss: 0.045, Loss x: 0.046, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.867
[2022-11-26 11:51:15,520][ INFO] Iters: 2170, Total loss: 0.045, Loss x: 0.046, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.868
[2022-11-26 11:54:27,968][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.50

[2022-11-26 11:54:27,968][ INFO] ===========> Epoch: 57, LR: 0.0002, Previous best: 75.50
[2022-11-26 11:54:28,900][ INFO] Iters: 0, Total loss: 0.038, Loss x: 0.045, Loss s: 0.058, Loss w_fp: 0.002, Mask: 0.883
[2022-11-26 11:56:54,730][ INFO] Iters: 310, Total loss: 0.044, Loss x: 0.046, Loss s: 0.078, Loss w_fp: 0.005, Mask: 0.863
[2022-11-26 11:59:20,621][ INFO] Iters: 620, Total loss: 0.045, Loss x: 0.046, Loss s: 0.081, Loss w_fp: 0.006, Mask: 0.859
[2022-11-26 12:01:46,475][ INFO] Iters: 930, Total loss: 0.045, Loss x: 0.046, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.860
[2022-11-26 12:04:12,280][ INFO] Iters: 1240, Total loss: 0.044, Loss x: 0.046, Loss s: 0.080, Loss w_fp: 0.006, Mask: 0.863
[2022-11-26 12:06:38,034][ INFO] Iters: 1550, Total loss: 0.045, Loss x: 0.046, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 12:09:03,909][ INFO] Iters: 1860, Total loss: 0.044, Loss x: 0.046, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 12:11:30,059][ INFO] Iters: 2170, Total loss: 0.044, Loss x: 0.046, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 12:14:42,647][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.24

[2022-11-26 12:14:42,647][ INFO] ===========> Epoch: 58, LR: 0.0002, Previous best: 75.50
[2022-11-26 12:14:43,629][ INFO] Iters: 0, Total loss: 0.038, Loss x: 0.054, Loss s: 0.042, Loss w_fp: 0.003, Mask: 0.819
[2022-11-26 12:17:09,267][ INFO] Iters: 310, Total loss: 0.043, Loss x: 0.045, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.869
[2022-11-26 12:19:35,127][ INFO] Iters: 620, Total loss: 0.043, Loss x: 0.046, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.869
[2022-11-26 12:22:01,166][ INFO] Iters: 930, Total loss: 0.044, Loss x: 0.046, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.870
[2022-11-26 12:24:27,059][ INFO] Iters: 1240, Total loss: 0.044, Loss x: 0.046, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.871
[2022-11-26 12:26:53,013][ INFO] Iters: 1550, Total loss: 0.044, Loss x: 0.046, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.871
[2022-11-26 12:29:19,101][ INFO] Iters: 1860, Total loss: 0.044, Loss x: 0.046, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.871
[2022-11-26 12:31:45,082][ INFO] Iters: 2170, Total loss: 0.044, Loss x: 0.046, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 12:34:56,832][ INFO] ***** Evaluation original ***** >>>> meanIOU: 75.20

[2022-11-26 12:34:56,832][ INFO] ===========> Epoch: 59, LR: 0.0002, Previous best: 75.50
[2022-11-26 12:34:57,782][ INFO] Iters: 0, Total loss: 0.026, Loss x: 0.026, Loss s: 0.041, Loss w_fp: 0.012, Mask: 0.897
[2022-11-26 12:37:23,437][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.043, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 12:39:49,310][ INFO] Iters: 620, Total loss: 0.043, Loss x: 0.044, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 12:42:15,185][ INFO] Iters: 930, Total loss: 0.043, Loss x: 0.045, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.872
[2022-11-26 12:44:41,006][ INFO] Iters: 1240, Total loss: 0.043, Loss x: 0.044, Loss s: 0.078, Loss w_fp: 0.006, Mask: 0.871
[2022-11-26 12:47:06,909][ INFO] Iters: 1550, Total loss: 0.043, Loss x: 0.045, Loss s: 0.078, Loss w_fp: 0.006, Mask: 0.870
[2022-11-26 12:49:32,478][ INFO] Iters: 1860, Total loss: 0.044, Loss x: 0.045, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 12:51:57,936][ INFO] Iters: 2170, Total loss: 0.044, Loss x: 0.045, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 12:55:09,624][ INFO] ***** Evaluation original ***** >>>> meanIOU: 75.18

[2022-11-26 12:55:09,624][ INFO] ===========> Epoch: 60, LR: 0.0001, Previous best: 75.50
[2022-11-26 12:55:10,612][ INFO] Iters: 0, Total loss: 0.045, Loss x: 0.033, Loss s: 0.108, Loss w_fp: 0.006, Mask: 0.899
[2022-11-26 12:57:36,220][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.044, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.862
[2022-11-26 13:00:02,062][ INFO] Iters: 620, Total loss: 0.042, Loss x: 0.045, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.866
[2022-11-26 13:02:27,916][ INFO] Iters: 930, Total loss: 0.042, Loss x: 0.045, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.866
[2022-11-26 13:04:53,793][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.044, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.868
[2022-11-26 13:07:19,528][ INFO] Iters: 1550, Total loss: 0.043, Loss x: 0.045, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.868
[2022-11-26 13:09:45,282][ INFO] Iters: 1860, Total loss: 0.043, Loss x: 0.045, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 13:12:11,231][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.045, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 13:15:22,930][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.15

[2022-11-26 13:15:22,931][ INFO] ===========> Epoch: 61, LR: 0.0001, Previous best: 75.50
[2022-11-26 13:15:23,886][ INFO] Iters: 0, Total loss: 0.044, Loss x: 0.063, Loss s: 0.046, Loss w_fp: 0.004, Mask: 0.843
[2022-11-26 13:17:49,647][ INFO] Iters: 310, Total loss: 0.046, Loss x: 0.046, Loss s: 0.087, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 13:20:15,348][ INFO] Iters: 620, Total loss: 0.045, Loss x: 0.046, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.868
[2022-11-26 13:22:41,402][ INFO] Iters: 930, Total loss: 0.045, Loss x: 0.046, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.867
[2022-11-26 13:25:06,971][ INFO] Iters: 1240, Total loss: 0.044, Loss x: 0.046, Loss s: 0.078, Loss w_fp: 0.006, Mask: 0.868
[2022-11-26 13:27:33,111][ INFO] Iters: 1550, Total loss: 0.044, Loss x: 0.045, Loss s: 0.079, Loss w_fp: 0.006, Mask: 0.869
[2022-11-26 13:29:58,903][ INFO] Iters: 1860, Total loss: 0.043, Loss x: 0.045, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 13:32:24,713][ INFO] Iters: 2170, Total loss: 0.043, Loss x: 0.045, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.871
[2022-11-26 13:35:36,523][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.47

[2022-11-26 13:35:36,523][ INFO] ===========> Epoch: 62, LR: 0.0001, Previous best: 75.50
[2022-11-26 13:35:37,471][ INFO] Iters: 0, Total loss: 0.014, Loss x: 0.004, Loss s: 0.034, Loss w_fp: 0.013, Mask: 0.875
[2022-11-26 13:38:03,005][ INFO] Iters: 310, Total loss: 0.043, Loss x: 0.046, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 13:40:28,712][ INFO] Iters: 620, Total loss: 0.044, Loss x: 0.045, Loss s: 0.080, Loss w_fp: 0.007, Mask: 0.865
[2022-11-26 13:42:54,664][ INFO] Iters: 930, Total loss: 0.044, Loss x: 0.045, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 13:45:20,332][ INFO] Iters: 1240, Total loss: 0.043, Loss x: 0.044, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.871
[2022-11-26 13:47:46,133][ INFO] Iters: 1550, Total loss: 0.043, Loss x: 0.044, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 13:50:11,919][ INFO] Iters: 1860, Total loss: 0.043, Loss x: 0.044, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.871
[2022-11-26 13:52:37,632][ INFO] Iters: 2170, Total loss: 0.043, Loss x: 0.044, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 13:55:49,728][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.41

[2022-11-26 13:55:49,728][ INFO] ===========> Epoch: 63, LR: 0.0001, Previous best: 75.50
[2022-11-26 13:55:50,670][ INFO] Iters: 0, Total loss: 0.029, Loss x: 0.051, Loss s: 0.015, Loss w_fp: 0.001, Mask: 0.958
[2022-11-26 13:58:16,224][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.044, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.876
[2022-11-26 14:00:42,107][ INFO] Iters: 620, Total loss: 0.043, Loss x: 0.045, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 14:03:07,846][ INFO] Iters: 930, Total loss: 0.043, Loss x: 0.045, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.871
[2022-11-26 14:05:33,621][ INFO] Iters: 1240, Total loss: 0.043, Loss x: 0.045, Loss s: 0.076, Loss w_fp: 0.006, Mask: 0.869
[2022-11-26 14:07:59,267][ INFO] Iters: 1550, Total loss: 0.043, Loss x: 0.045, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 14:10:24,949][ INFO] Iters: 1860, Total loss: 0.043, Loss x: 0.044, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.871
[2022-11-26 14:12:50,610][ INFO] Iters: 2170, Total loss: 0.043, Loss x: 0.045, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.870
[2022-11-26 14:16:01,629][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.38

[2022-11-26 14:16:01,630][ INFO] ===========> Epoch: 64, LR: 0.0001, Previous best: 75.50
[2022-11-26 14:16:02,649][ INFO] Iters: 0, Total loss: 0.043, Loss x: 0.064, Loss s: 0.031, Loss w_fp: 0.011, Mask: 0.900
[2022-11-26 14:18:28,307][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.043, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 14:20:54,204][ INFO] Iters: 620, Total loss: 0.041, Loss x: 0.043, Loss s: 0.071, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 14:23:20,072][ INFO] Iters: 930, Total loss: 0.041, Loss x: 0.044, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 14:25:45,909][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.044, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.872
[2022-11-26 14:28:11,749][ INFO] Iters: 1550, Total loss: 0.042, Loss x: 0.043, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 14:30:37,476][ INFO] Iters: 1860, Total loss: 0.042, Loss x: 0.044, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.872
[2022-11-26 14:33:03,438][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.043, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 14:36:15,417][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.00

[2022-11-26 14:36:15,418][ INFO] ===========> Epoch: 65, LR: 0.0001, Previous best: 75.50
[2022-11-26 14:36:16,383][ INFO] Iters: 0, Total loss: 0.120, Loss x: 0.051, Loss s: 0.358, Loss w_fp: 0.017, Mask: 0.739
[2022-11-26 14:38:41,883][ INFO] Iters: 310, Total loss: 0.045, Loss x: 0.045, Loss s: 0.083, Loss w_fp: 0.007, Mask: 0.867
[2022-11-26 14:41:07,715][ INFO] Iters: 620, Total loss: 0.044, Loss x: 0.044, Loss s: 0.081, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 14:43:33,642][ INFO] Iters: 930, Total loss: 0.043, Loss x: 0.044, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 14:45:59,682][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.044, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 14:48:25,603][ INFO] Iters: 1550, Total loss: 0.042, Loss x: 0.044, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 14:50:51,235][ INFO] Iters: 1860, Total loss: 0.042, Loss x: 0.044, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 14:53:16,984][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.044, Loss s: 0.073, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 14:56:29,377][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.41

[2022-11-26 14:56:29,377][ INFO] ===========> Epoch: 66, LR: 0.0001, Previous best: 75.50
[2022-11-26 14:56:30,308][ INFO] Iters: 0, Total loss: 0.037, Loss x: 0.030, Loss s: 0.085, Loss w_fp: 0.004, Mask: 0.910
[2022-11-26 14:58:56,270][ INFO] Iters: 310, Total loss: 0.039, Loss x: 0.042, Loss s: 0.068, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:01:22,247][ INFO] Iters: 620, Total loss: 0.040, Loss x: 0.042, Loss s: 0.071, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 15:03:48,455][ INFO] Iters: 930, Total loss: 0.042, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 15:06:14,616][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:08:40,501][ INFO] Iters: 1550, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 15:11:06,328][ INFO] Iters: 1860, Total loss: 0.042, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 15:13:32,193][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 15:16:45,312][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.90

[2022-11-26 15:16:45,312][ INFO] ===========> Epoch: 67, LR: 0.0001, Previous best: 75.50
[2022-11-26 15:16:46,296][ INFO] Iters: 0, Total loss: 0.028, Loss x: 0.037, Loss s: 0.030, Loss w_fp: 0.009, Mask: 0.900
[2022-11-26 15:19:12,017][ INFO] Iters: 310, Total loss: 0.043, Loss x: 0.045, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.883
[2022-11-26 15:21:37,881][ INFO] Iters: 620, Total loss: 0.042, Loss x: 0.044, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:24:03,621][ INFO] Iters: 930, Total loss: 0.042, Loss x: 0.043, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:26:29,487][ INFO] Iters: 1240, Total loss: 0.041, Loss x: 0.043, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 15:28:55,299][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.043, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.875
[2022-11-26 15:31:20,999][ INFO] Iters: 1860, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:33:46,605][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:36:58,628][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.40

[2022-11-26 15:36:58,628][ INFO] ===========> Epoch: 68, LR: 0.0001, Previous best: 75.50
[2022-11-26 15:36:59,587][ INFO] Iters: 0, Total loss: 0.036, Loss x: 0.031, Loss s: 0.078, Loss w_fp: 0.003, Mask: 0.893
[2022-11-26 15:39:25,359][ INFO] Iters: 310, Total loss: 0.043, Loss x: 0.043, Loss s: 0.078, Loss w_fp: 0.006, Mask: 0.873
[2022-11-26 15:41:51,302][ INFO] Iters: 620, Total loss: 0.044, Loss x: 0.043, Loss s: 0.082, Loss w_fp: 0.006, Mask: 0.871
[2022-11-26 15:44:17,272][ INFO] Iters: 930, Total loss: 0.043, Loss x: 0.044, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:46:43,027][ INFO] Iters: 1240, Total loss: 0.043, Loss x: 0.043, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 15:49:08,844][ INFO] Iters: 1550, Total loss: 0.043, Loss x: 0.043, Loss s: 0.079, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 15:51:34,646][ INFO] Iters: 1860, Total loss: 0.043, Loss x: 0.043, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.873
[2022-11-26 15:54:00,462][ INFO] Iters: 2170, Total loss: 0.043, Loss x: 0.043, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 15:57:12,666][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.46

[2022-11-26 15:57:12,667][ INFO] ===========> Epoch: 69, LR: 0.0001, Previous best: 75.50
[2022-11-26 15:57:13,584][ INFO] Iters: 0, Total loss: 0.020, Loss x: 0.019, Loss s: 0.040, Loss w_fp: 0.002, Mask: 0.936
[2022-11-26 15:59:39,474][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.042, Loss s: 0.078, Loss w_fp: 0.007, Mask: 0.883
[2022-11-26 16:02:05,623][ INFO] Iters: 620, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.880
[2022-11-26 16:04:31,799][ INFO] Iters: 930, Total loss: 0.042, Loss x: 0.043, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.879
[2022-11-26 16:06:57,971][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.043, Loss s: 0.077, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 16:09:23,926][ INFO] Iters: 1550, Total loss: 0.042, Loss x: 0.043, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.877
[2022-11-26 16:11:49,954][ INFO] Iters: 1860, Total loss: 0.042, Loss x: 0.043, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.876
[2022-11-26 16:14:16,064][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 16:17:27,339][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.04

[2022-11-26 16:17:27,339][ INFO] ===========> Epoch: 70, LR: 0.0001, Previous best: 75.50
[2022-11-26 16:17:28,286][ INFO] Iters: 0, Total loss: 0.067, Loss x: 0.055, Loss s: 0.154, Loss w_fp: 0.004, Mask: 0.897
[2022-11-26 16:19:53,816][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.044, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 16:22:19,542][ INFO] Iters: 620, Total loss: 0.041, Loss x: 0.043, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 16:24:45,129][ INFO] Iters: 930, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 16:27:10,740][ INFO] Iters: 1240, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.875
[2022-11-26 16:29:36,433][ INFO] Iters: 1550, Total loss: 0.042, Loss x: 0.042, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.873
[2022-11-26 16:32:02,092][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 16:34:27,946][ INFO] Iters: 2170, Total loss: 0.042, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 16:37:40,700][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.45

[2022-11-26 16:37:40,700][ INFO] ===========> Epoch: 71, LR: 0.0001, Previous best: 75.50
[2022-11-26 16:37:41,687][ INFO] Iters: 0, Total loss: 0.034, Loss x: 0.042, Loss s: 0.050, Loss w_fp: 0.003, Mask: 0.958
[2022-11-26 16:40:07,409][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.040, Loss s: 0.079, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 16:42:33,304][ INFO] Iters: 620, Total loss: 0.041, Loss x: 0.041, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.875
[2022-11-26 16:44:59,122][ INFO] Iters: 930, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 16:47:25,134][ INFO] Iters: 1240, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 16:49:51,268][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.876
[2022-11-26 16:52:17,082][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.042, Loss s: 0.073, Loss w_fp: 0.007, Mask: 0.876
[2022-11-26 16:54:42,751][ INFO] Iters: 2170, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.876
[2022-11-26 16:57:54,588][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.28

[2022-11-26 16:57:54,589][ INFO] ===========> Epoch: 72, LR: 0.0001, Previous best: 75.50
[2022-11-26 16:57:55,606][ INFO] Iters: 0, Total loss: 0.021, Loss x: 0.024, Loss s: 0.031, Loss w_fp: 0.003, Mask: 0.906
[2022-11-26 17:00:21,296][ INFO] Iters: 310, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.873
[2022-11-26 17:02:47,255][ INFO] Iters: 620, Total loss: 0.040, Loss x: 0.042, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.872
[2022-11-26 17:05:13,095][ INFO] Iters: 930, Total loss: 0.041, Loss x: 0.042, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.872
[2022-11-26 17:07:38,972][ INFO] Iters: 1240, Total loss: 0.040, Loss x: 0.042, Loss s: 0.071, Loss w_fp: 0.007, Mask: 0.872
[2022-11-26 17:10:04,906][ INFO] Iters: 1550, Total loss: 0.040, Loss x: 0.042, Loss s: 0.071, Loss w_fp: 0.007, Mask: 0.874
[2022-11-26 17:12:30,715][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.042, Loss s: 0.071, Loss w_fp: 0.007, Mask: 0.875
[2022-11-26 17:14:56,443][ INFO] Iters: 2170, Total loss: 0.041, Loss x: 0.042, Loss s: 0.071, Loss w_fp: 0.007, Mask: 0.875
[2022-11-26 17:18:08,370][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.72

[2022-11-26 17:18:08,370][ INFO] ===========> Epoch: 73, LR: 0.0001, Previous best: 75.50
[2022-11-26 17:18:09,303][ INFO] Iters: 0, Total loss: 0.036, Loss x: 0.055, Loss s: 0.021, Loss w_fp: 0.012, Mask: 0.858
[2022-11-26 17:20:35,162][ INFO] Iters: 310, Total loss: 0.042, Loss x: 0.043, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.869
[2022-11-26 17:23:01,109][ INFO] Iters: 620, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 17:25:27,178][ INFO] Iters: 930, Total loss: 0.041, Loss x: 0.043, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.874
[2022-11-26 17:27:53,151][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.043, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 17:30:19,070][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.875
[2022-11-26 17:32:44,909][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.876
[2022-11-26 17:35:11,114][ INFO] Iters: 2170, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.876
[2022-11-26 17:38:25,497][ INFO] ***** Evaluation original ***** >>>> meanIOU: 75.23

[2022-11-26 17:38:25,498][ INFO] ===========> Epoch: 74, LR: 0.0000, Previous best: 75.50
[2022-11-26 17:38:26,570][ INFO] Iters: 0, Total loss: 0.027, Loss x: 0.040, Loss s: 0.012, Loss w_fp: 0.017, Mask: 0.965
[2022-11-26 17:40:52,452][ INFO] Iters: 310, Total loss: 0.039, Loss x: 0.040, Loss s: 0.070, Loss w_fp: 0.006, Mask: 0.879
[2022-11-26 17:43:18,544][ INFO] Iters: 620, Total loss: 0.040, Loss x: 0.042, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 17:45:44,546][ INFO] Iters: 930, Total loss: 0.040, Loss x: 0.042, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 17:48:10,223][ INFO] Iters: 1240, Total loss: 0.041, Loss x: 0.042, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 17:50:36,061][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.042, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 17:53:01,906][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 17:55:27,752][ INFO] Iters: 2170, Total loss: 0.041, Loss x: 0.041, Loss s: 0.074, Loss w_fp: 0.007, Mask: 0.878
[2022-11-26 17:58:40,788][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.61

[2022-11-26 17:58:40,789][ INFO] ===========> Epoch: 75, LR: 0.0000, Previous best: 75.50
[2022-11-26 17:58:41,776][ INFO] Iters: 0, Total loss: 0.044, Loss x: 0.064, Loss s: 0.040, Loss w_fp: 0.007, Mask: 0.950
[2022-11-26 18:01:07,456][ INFO] Iters: 310, Total loss: 0.039, Loss x: 0.042, Loss s: 0.067, Loss w_fp: 0.006, Mask: 0.885
[2022-11-26 18:03:33,128][ INFO] Iters: 620, Total loss: 0.039, Loss x: 0.041, Loss s: 0.067, Loss w_fp: 0.006, Mask: 0.881
[2022-11-26 18:05:59,445][ INFO] Iters: 930, Total loss: 0.040, Loss x: 0.042, Loss s: 0.070, Loss w_fp: 0.006, Mask: 0.880
[2022-11-26 18:08:25,245][ INFO] Iters: 1240, Total loss: 0.041, Loss x: 0.042, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 18:10:51,085][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.042, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.879
[2022-11-26 18:13:16,942][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.041, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 18:15:42,945][ INFO] Iters: 2170, Total loss: 0.040, Loss x: 0.041, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.879
[2022-11-26 18:18:56,812][ INFO] ***** Evaluation original ***** >>>> meanIOU: 73.65

[2022-11-26 18:18:56,813][ INFO] ===========> Epoch: 76, LR: 0.0000, Previous best: 75.50
[2022-11-26 18:18:57,858][ INFO] Iters: 0, Total loss: 0.019, Loss x: 0.028, Loss s: 0.017, Loss w_fp: 0.002, Mask: 0.947
[2022-11-26 18:21:23,727][ INFO] Iters: 310, Total loss: 0.038, Loss x: 0.040, Loss s: 0.065, Loss w_fp: 0.006, Mask: 0.883
[2022-11-26 18:23:49,527][ INFO] Iters: 620, Total loss: 0.039, Loss x: 0.041, Loss s: 0.068, Loss w_fp: 0.006, Mask: 0.883
[2022-11-26 18:26:15,655][ INFO] Iters: 930, Total loss: 0.040, Loss x: 0.041, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.881
[2022-11-26 18:28:41,375][ INFO] Iters: 1240, Total loss: 0.040, Loss x: 0.041, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.881
[2022-11-26 18:31:07,247][ INFO] Iters: 1550, Total loss: 0.040, Loss x: 0.041, Loss s: 0.072, Loss w_fp: 0.006, Mask: 0.880
[2022-11-26 18:33:32,964][ INFO] Iters: 1860, Total loss: 0.040, Loss x: 0.041, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.880
[2022-11-26 18:35:58,774][ INFO] Iters: 2170, Total loss: 0.040, Loss x: 0.041, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.879
[2022-11-26 18:39:12,715][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.40

[2022-11-26 18:39:12,715][ INFO] ===========> Epoch: 77, LR: 0.0000, Previous best: 75.50
[2022-11-26 18:39:13,694][ INFO] Iters: 0, Total loss: 0.023, Loss x: 0.024, Loss s: 0.040, Loss w_fp: 0.005, Mask: 0.893
[2022-11-26 18:41:39,716][ INFO] Iters: 310, Total loss: 0.039, Loss x: 0.040, Loss s: 0.068, Loss w_fp: 0.006, Mask: 0.879
[2022-11-26 18:44:05,647][ INFO] Iters: 620, Total loss: 0.040, Loss x: 0.041, Loss s: 0.071, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 18:46:31,521][ INFO] Iters: 930, Total loss: 0.040, Loss x: 0.041, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 18:48:57,314][ INFO] Iters: 1240, Total loss: 0.041, Loss x: 0.042, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 18:51:23,071][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.006, Mask: 0.877
[2022-11-26 18:53:48,921][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.042, Loss s: 0.074, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 18:56:14,890][ INFO] Iters: 2170, Total loss: 0.041, Loss x: 0.041, Loss s: 0.073, Loss w_fp: 0.006, Mask: 0.878
[2022-11-26 18:59:28,298][ INFO] ***** Evaluation original ***** >>>> meanIOU: 74.88

[2022-11-26 18:59:28,298][ INFO] ===========> Epoch: 78, LR: 0.0000, Previous best: 75.50
[2022-11-26 18:59:29,233][ INFO] Iters: 0, Total loss: 0.042, Loss x: 0.070, Loss s: 0.027, Loss w_fp: 0.002, Mask: 0.941
[2022-11-26 19:01:55,023][ INFO] Iters: 310, Total loss: 0.040, Loss x: 0.042, Loss s: 0.071, Loss w_fp: 0.007, Mask: 0.876
[2022-11-26 19:04:21,088][ INFO] Iters: 620, Total loss: 0.042, Loss x: 0.042, Loss s: 0.076, Loss w_fp: 0.007, Mask: 0.877
[2022-11-26 19:06:46,743][ INFO] Iters: 930, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.876
[2022-11-26 19:09:12,492][ INFO] Iters: 1240, Total loss: 0.042, Loss x: 0.042, Loss s: 0.077, Loss w_fp: 0.007, Mask: 0.876
[2022-11-26 19:11:38,280][ INFO] Iters: 1550, Total loss: 0.041, Loss x: 0.042, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.878
[2022-11-26 19:14:04,024][ INFO] Iters: 1860, Total loss: 0.041, Loss x: 0.041, Loss s: 0.075, Loss w_fp: 0.007, Mask: 0.879

from unimatch.

LiheYoung commented on September 24, 2024

Since the batch size is reduced by 2 times, I think it can be expected that the performance drops by 1%. You can try to use two GPUs, or reduce the cropping size from 321 to 281 and maintain the batch size as 8 if GPU memory is allowed.

from unimatch.

jerrywyn commented on September 24, 2024

I ran the unimatch with batch size 4 and initial lr 0.0005 on a single GPU for the 1/16 partition of pascal and I took 78.63 mIOU without other change except eliminated distributed code. The best result is trained in epoch 38. This is a bit of a huge difference from your result of 75.50？

from unimatch.

nysp78 commented on September 24, 2024

@jerrywyn Which part of the distributed code do you cross out? Also, what crop size do you use?

Thanks

from unimatch.

About the reproduction of results on Pascal dataset about unimatch HOT 7 CLOSED

Comments (7)

Related Issues (20)

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent