-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathlog.txt
122 lines (121 loc) · 8.1 KB
/
log.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
[ 2022-06-07 10:06:43,817 ] Saving folder path: ./workdir/2002_EfficientGCN-B0_ntu-xview/2022-06-07 10-06-43
[ 2022-06-07 10:06:43,817 ] Saving model name: ./60/70.pdparams
[ 2022-06-07 10:06:43,835 ] Dataset: ntu-xview
[ 2022-06-07 10:06:43,835 ] Batch size: train-16, eval-16
[ 2022-06-07 10:06:43,835 ] Data shape (branch, channel, frame, joint, person): [3, 6, 288, 25, 2]
[ 2022-06-07 10:06:43,835 ] Number of action classes: 60
[ 2022-06-07 10:06:46,693 ] Model: EfficientGCN-B0 {'stem_channel': 64, 'block_args': [[48, 1, 0.5], [24, 1, 0.5], [64, 2, 1], [128, 2, 1]], 'fusion_stage': 2, 'act_type': 'swish', 'att_type': 'stja', 'layer_type': 'SG', 'drop_prob': 0.25, 'kernel_size': [5, 2], 'scale_args': [1.2, 1.35], 'expand_ratio': 0, 'reduct_ratio': 2, 'bias': True, 'edge': True}
[ 2022-06-07 10:06:46,695 ] LR_Scheduler: cosine {'max_epoch': 70, 'warm_up': 10}
[ 2022-06-07 10:06:46,696 ] Optimizer: Momentum {'learning_rate': <paddle.optimizer.lr.LambdaDecay object at 0x7f3bc2a34c50>, 'momentum': 0.9, 'use_nesterov': True, 'weight_decay': 0.0001}
[ 2022-06-07 10:06:46,696 ] Loss function: CrossEntropyLoss
[ 2022-06-07 10:06:46,696 ] Successful!
[ 2022-06-07 10:06:46,697 ]
[ 2022-06-07 10:06:46,697 ] Loading pretrain model ...
[ 2022-06-07 10:06:46,760 ] Successful!
[ 2022-06-07 10:06:46,760 ]
[ 2022-06-07 10:06:46,760 ] Starting training ...
[ 2022-06-07 10:20:32,245 ] Epoch: 60/70, Training accuracy: 36806/37632(97.81%), Training time: 825.48s
[ 2022-06-07 10:20:32,245 ]
[ 2022-06-07 10:20:32,246 ] Evaluating for epoch 60/70 ...
[ 2022-06-07 10:22:54,946 ] Top-1 accuracy: 17865/18928(94.38%), Top-5 accuracy: 18783/18928(99.23%), Mean loss:0.1905
[ 2022-06-07 10:22:54,946 ] Evaluating time: 142.70s, Speed: 132.64 sequnces/(second*GPU)
[ 2022-06-07 10:22:54,946 ]
[ 2022-06-07 10:22:54,947 ] Saving model for epoch 60/70 ...
[ 2022-06-07 10:22:54,974 ] Best top-1 accuracy: 94.38%, Total time: 00d-00h-16m-08s
[ 2022-06-07 10:22:54,974 ]
[ 2022-06-07 10:36:42,489 ] Epoch: 61/70, Training accuracy: 36972/37632(98.25%), Training time: 827.51s
[ 2022-06-07 10:36:42,489 ]
[ 2022-06-07 10:36:42,490 ] Evaluating for epoch 61/70 ...
[ 2022-06-07 10:39:04,484 ] Top-1 accuracy: 17920/18928(94.67%), Top-5 accuracy: 18788/18928(99.26%), Mean loss:0.1816
[ 2022-06-07 10:39:04,484 ] Evaluating time: 141.99s, Speed: 133.30 sequnces/(second*GPU)
[ 2022-06-07 10:39:04,484 ]
[ 2022-06-07 10:39:04,484 ] Saving model for epoch 61/70 ...
[ 2022-06-07 10:39:04,511 ] Best top-1 accuracy: 94.67%, Total time: 00d-00h-32m-17s
[ 2022-06-07 10:39:04,511 ]
[ 2022-06-07 10:52:52,617 ] Epoch: 62/70, Training accuracy: 37082/37632(98.54%), Training time: 828.10s
[ 2022-06-07 10:52:52,618 ]
[ 2022-06-07 10:52:52,619 ] Evaluating for epoch 62/70 ...
[ 2022-06-07 10:55:13,542 ] Top-1 accuracy: 17814/18928(94.11%), Top-5 accuracy: 18790/18928(99.27%), Mean loss:0.1949
[ 2022-06-07 10:55:13,542 ] Evaluating time: 140.92s, Speed: 134.32 sequnces/(second*GPU)
[ 2022-06-07 10:55:13,542 ]
[ 2022-06-07 10:55:13,542 ] Saving model for epoch 62/70 ...
[ 2022-06-07 10:55:13,568 ] Best top-1 accuracy: 94.67%, Total time: 00d-00h-48m-26s
[ 2022-06-07 10:55:13,568 ]
[ 2022-06-07 11:08:59,108 ] Epoch: 63/70, Training accuracy: 37145/37632(98.71%), Training time: 825.54s
[ 2022-06-07 11:08:59,109 ]
[ 2022-06-07 11:08:59,110 ] Evaluating for epoch 63/70 ...
[ 2022-06-07 11:11:20,612 ] Top-1 accuracy: 17938/18928(94.77%), Top-5 accuracy: 18796/18928(99.30%), Mean loss:0.1821
[ 2022-06-07 11:11:20,612 ] Evaluating time: 141.50s, Speed: 133.77 sequnces/(second*GPU)
[ 2022-06-07 11:11:20,612 ]
[ 2022-06-07 11:11:20,612 ] Saving model for epoch 63/70 ...
[ 2022-06-07 11:11:20,638 ] Best top-1 accuracy: 94.77%, Total time: 00d-01h-04m-33s
[ 2022-06-07 11:11:20,638 ]
[ 2022-06-07 11:25:07,508 ] Epoch: 64/70, Training accuracy: 37253/37632(98.99%), Training time: 826.87s
[ 2022-06-07 11:25:07,508 ]
[ 2022-06-07 11:25:07,510 ] Evaluating for epoch 64/70 ...
[ 2022-06-07 11:27:30,794 ] Top-1 accuracy: 17902/18928(94.58%), Top-5 accuracy: 18786/18928(99.25%), Mean loss:0.1856
[ 2022-06-07 11:27:30,794 ] Evaluating time: 143.28s, Speed: 132.10 sequnces/(second*GPU)
[ 2022-06-07 11:27:30,794 ]
[ 2022-06-07 11:27:30,795 ] Saving model for epoch 64/70 ...
[ 2022-06-07 11:27:30,822 ] Best top-1 accuracy: 94.77%, Total time: 00d-01h-20m-44s
[ 2022-06-07 11:27:30,822 ]
[ 2022-06-07 11:41:16,431 ] Epoch: 65/70, Training accuracy: 37285/37632(99.08%), Training time: 825.61s
[ 2022-06-07 11:41:16,431 ]
[ 2022-06-07 11:41:16,432 ] Evaluating for epoch 65/70 ...
[ 2022-06-07 11:43:36,693 ] Top-1 accuracy: 17897/18928(94.55%), Top-5 accuracy: 18785/18928(99.24%), Mean loss:0.1884
[ 2022-06-07 11:43:36,693 ] Evaluating time: 140.26s, Speed: 134.95 sequnces/(second*GPU)
[ 2022-06-07 11:43:36,693 ]
[ 2022-06-07 11:43:36,693 ] Saving model for epoch 65/70 ...
[ 2022-06-07 11:43:36,722 ] Best top-1 accuracy: 94.77%, Total time: 00d-01h-36m-50s
[ 2022-06-07 11:43:36,722 ]
[ 2022-06-07 11:57:23,343 ] Epoch: 66/70, Training accuracy: 37288/37632(99.09%), Training time: 826.62s
[ 2022-06-07 11:57:23,343 ]
[ 2022-06-07 11:57:23,344 ] Evaluating for epoch 66/70 ...
[ 2022-06-07 11:59:44,752 ] Top-1 accuracy: 17957/18928(94.87%), Top-5 accuracy: 18791/18928(99.28%), Mean loss:0.1799
[ 2022-06-07 11:59:44,753 ] Evaluating time: 141.41s, Speed: 133.85 sequnces/(second*GPU)
[ 2022-06-07 11:59:44,753 ]
[ 2022-06-07 11:59:44,753 ] Saving model for epoch 66/70 ...
[ 2022-06-07 11:59:44,779 ] Best top-1 accuracy: 94.87%, Total time: 00d-01h-52m-58s
[ 2022-06-07 11:59:44,779 ]
[ 2022-06-07 12:13:30,770 ] Epoch: 67/70, Training accuracy: 37303/37632(99.13%), Training time: 825.99s
[ 2022-06-07 12:13:30,771 ]
[ 2022-06-07 12:13:30,772 ] Evaluating for epoch 67/70 ...
[ 2022-06-07 12:15:53,397 ] Top-1 accuracy: 17870/18928(94.41%), Top-5 accuracy: 18778/18928(99.21%), Mean loss:0.1943
[ 2022-06-07 12:15:53,397 ] Evaluating time: 142.62s, Speed: 132.71 sequnces/(second*GPU)
[ 2022-06-07 12:15:53,397 ]
[ 2022-06-07 12:15:53,397 ] Saving model for epoch 67/70 ...
[ 2022-06-07 12:15:53,426 ] Best top-1 accuracy: 94.87%, Total time: 00d-02h-09m-06s
[ 2022-06-07 12:15:53,426 ]
[ 2022-06-07 12:29:39,566 ] Epoch: 68/70, Training accuracy: 37322/37632(99.18%), Training time: 826.14s
[ 2022-06-07 12:29:39,566 ]
[ 2022-06-07 12:29:39,567 ] Evaluating for epoch 68/70 ...
[ 2022-06-07 12:32:00,241 ] Top-1 accuracy: 17964/18928(94.91%), Top-5 accuracy: 18791/18928(99.28%), Mean loss:0.1766
[ 2022-06-07 12:32:00,241 ] Evaluating time: 140.67s, Speed: 134.55 sequnces/(second*GPU)
[ 2022-06-07 12:32:00,241 ]
[ 2022-06-07 12:32:00,241 ] Saving model for epoch 68/70 ...
[ 2022-06-07 12:32:00,278 ] Best top-1 accuracy: 94.91%, Total time: 00d-02h-25m-13s
[ 2022-06-07 12:32:00,278 ]
[ 2022-06-07 12:45:46,589 ] Epoch: 69/70, Training accuracy: 37335/37632(99.21%), Training time: 826.31s
[ 2022-06-07 12:45:46,589 ]
[ 2022-06-07 12:45:46,590 ] Evaluating for epoch 69/70 ...
[ 2022-06-07 12:48:05,975 ] Top-1 accuracy: 17929/18928(94.72%), Top-5 accuracy: 18785/18928(99.24%), Mean loss:0.1839
[ 2022-06-07 12:48:05,975 ] Evaluating time: 139.38s, Speed: 135.80 sequnces/(second*GPU)
[ 2022-06-07 12:48:05,975 ]
[ 2022-06-07 12:48:05,975 ] Saving model for epoch 69/70 ...
[ 2022-06-07 12:48:06,004 ] Best top-1 accuracy: 94.91%, Total time: 00d-02h-41m-19s
[ 2022-06-07 12:48:06,004 ]
[ 2022-06-07 13:01:51,555 ] Epoch: 70/70, Training accuracy: 37354/37632(99.26%), Training time: 825.55s
[ 2022-06-07 13:01:51,555 ]
[ 2022-06-07 13:01:51,557 ] Evaluating for epoch 70/70 ...
[ 2022-06-07 13:04:12,181 ] Top-1 accuracy: 17935/18928(94.75%), Top-5 accuracy: 18788/18928(99.26%), Mean loss:0.1813
[ 2022-06-07 13:04:12,182 ] Evaluating time: 140.62s, Speed: 134.60 sequnces/(second*GPU)
[ 2022-06-07 13:04:12,182 ]
[ 2022-06-07 13:04:12,182 ] Saving model for epoch 70/70 ...
[ 2022-06-07 13:04:12,207 ] Best top-1 accuracy: 94.91%, Total time: 00d-02h-57m-25s
[ 2022-06-07 13:04:12,208 ]
[ 2022-06-07 13:04:12,208 ] Finish training!
[ 2022-06-07 13:04:12,208 ]
[ 2022-06-07 13:04:12,207 ] Best top-1 accuracy: 94.91%, Total time: 00d-02h-57m-25s
[ 2022-06-07 13:04:12,208 ]
[ 2022-06-07 13:04:12,208 ] Finish training!
[ 2022-06-07 13:04:12,208 ]