Hugging Face's logo

zwa73
/
SoulTide-AudioData-Model

Japanese
Model card Files
xet
Community
zwa73 commited on
Commit
983c6fa
·
1 Parent(s): 1807e40

sovits 4

Browse files
Files changed (17) hide show
  1. so-vits-svc-4.0/20230325-Dreizehn-44k/best/D_13600.pth +3 -0
  2. so-vits-svc-4.0/20230325-Dreizehn-44k/best/G_13600.pth +3 -0
  3. so-vits-svc-4.0/20230325-Dreizehn-44k/best/config.json +93 -0
  4. so-vits-svc-4.0/20230325-Dreizehn-44k/config.json +93 -0
  5. so-vits-svc-4.0/20230325-Dreizehn-44k/kmeans_10000.pt +3 -0
  6. so-vits-svc-4.0/20230325-Dreizehn-44k/pruned/P_G_13600.pth +3 -0
  7. so-vits-svc-4.0/20230325-Dreizehn-44k/pruned/config.json +93 -0
  8. so-vits-svc-4.0/20230325-Dreizehn-44k/train.log +405 -0
  9. so-vits-svc-4.0/20230325-DreizehnNS-44k/best/D_13600.pth +3 -0
  10. so-vits-svc-4.0/20230325-DreizehnNS-44k/best/G_13600.pth +3 -0
  11. so-vits-svc-4.0/20230325-DreizehnNS-44k/best/config.json +93 -0
  12. so-vits-svc-4.0/20230325-DreizehnNS-44k/config.json +93 -0
  13. so-vits-svc-4.0/20230325-DreizehnNS-44k/kmeans_10000.pt +3 -0
  14. so-vits-svc-4.0/20230325-DreizehnNS-44k/pruned/P_G_13600.pth +3 -0
  15. so-vits-svc-4.0/20230325-DreizehnNS-44k/pruned/config.json +93 -0
  16. so-vits-svc-4.0/20230325-DreizehnNS-44k/train.log +592 -0
  17. so-vits-svc-4.0/README.md +5 -0
so-vits-svc-4.0/20230325-Dreizehn-44k/best/D_13600.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba7a99daaf96c141fa25a929f2fb5219925113e628481dd415b19704c1d72b50
3
+ size 561098185
so-vits-svc-4.0/20230325-Dreizehn-44k/best/G_13600.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3080cc0577b0947cd0c3b28fb99ef315efda3d78f8e380e2f0f65bf6607d6c6d
3
+ size 542789405
so-vits-svc-4.0/20230325-Dreizehn-44k/best/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 800,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 0.0001,
8
+ "betas": [
9
+ 0.8,
10
+ 0.99
11
+ ],
12
+ "eps": 1e-09,
13
+ "batch_size": 6,
14
+ "fp16_run": false,
15
+ "lr_decay": 0.999875,
16
+ "segment_size": 10240,
17
+ "init_lr_ratio": 1,
18
+ "warmup_epochs": 0,
19
+ "c_mel": 45,
20
+ "c_kl": 1.0,
21
+ "use_sr": true,
22
+ "max_speclen": 512,
23
+ "port": "8001",
24
+ "keep_ckpts": 99
25
+ },
26
+ "data": {
27
+ "training_files": "filelists/train.txt",
28
+ "validation_files": "filelists/val.txt",
29
+ "max_wav_value": 32768.0,
30
+ "sampling_rate": 44100,
31
+ "filter_length": 2048,
32
+ "hop_length": 512,
33
+ "win_length": 2048,
34
+ "n_mel_channels": 80,
35
+ "mel_fmin": 0.0,
36
+ "mel_fmax": 22050
37
+ },
38
+ "model": {
39
+ "inter_channels": 192,
40
+ "hidden_channels": 192,
41
+ "filter_channels": 768,
42
+ "n_heads": 2,
43
+ "n_layers": 6,
44
+ "kernel_size": 3,
45
+ "p_dropout": 0.1,
46
+ "resblock": "1",
47
+ "resblock_kernel_sizes": [
48
+ 3,
49
+ 7,
50
+ 11
51
+ ],
52
+ "resblock_dilation_sizes": [
53
+ [
54
+ 1,
55
+ 3,
56
+ 5
57
+ ],
58
+ [
59
+ 1,
60
+ 3,
61
+ 5
62
+ ],
63
+ [
64
+ 1,
65
+ 3,
66
+ 5
67
+ ]
68
+ ],
69
+ "upsample_rates": [
70
+ 8,
71
+ 8,
72
+ 2,
73
+ 2,
74
+ 2
75
+ ],
76
+ "upsample_initial_channel": 512,
77
+ "upsample_kernel_sizes": [
78
+ 16,
79
+ 16,
80
+ 4,
81
+ 4,
82
+ 4
83
+ ],
84
+ "n_layers_q": 3,
85
+ "use_spectral_norm": false,
86
+ "gin_channels": 256,
87
+ "ssl_dim": 256,
88
+ "n_speakers": 200
89
+ },
90
+ "spk": {
91
+ "Dreizehn": 0
92
+ }
93
+ }
so-vits-svc-4.0/20230325-Dreizehn-44k/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 800,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 0.0001,
8
+ "betas": [
9
+ 0.8,
10
+ 0.99
11
+ ],
12
+ "eps": 1e-09,
13
+ "batch_size": 6,
14
+ "fp16_run": false,
15
+ "lr_decay": 0.999875,
16
+ "segment_size": 10240,
17
+ "init_lr_ratio": 1,
18
+ "warmup_epochs": 0,
19
+ "c_mel": 45,
20
+ "c_kl": 1.0,
21
+ "use_sr": true,
22
+ "max_speclen": 512,
23
+ "port": "8001",
24
+ "keep_ckpts": 99
25
+ },
26
+ "data": {
27
+ "training_files": "filelists/train.txt",
28
+ "validation_files": "filelists/val.txt",
29
+ "max_wav_value": 32768.0,
30
+ "sampling_rate": 44100,
31
+ "filter_length": 2048,
32
+ "hop_length": 512,
33
+ "win_length": 2048,
34
+ "n_mel_channels": 80,
35
+ "mel_fmin": 0.0,
36
+ "mel_fmax": 22050
37
+ },
38
+ "model": {
39
+ "inter_channels": 192,
40
+ "hidden_channels": 192,
41
+ "filter_channels": 768,
42
+ "n_heads": 2,
43
+ "n_layers": 6,
44
+ "kernel_size": 3,
45
+ "p_dropout": 0.1,
46
+ "resblock": "1",
47
+ "resblock_kernel_sizes": [
48
+ 3,
49
+ 7,
50
+ 11
51
+ ],
52
+ "resblock_dilation_sizes": [
53
+ [
54
+ 1,
55
+ 3,
56
+ 5
57
+ ],
58
+ [
59
+ 1,
60
+ 3,
61
+ 5
62
+ ],
63
+ [
64
+ 1,
65
+ 3,
66
+ 5
67
+ ]
68
+ ],
69
+ "upsample_rates": [
70
+ 8,
71
+ 8,
72
+ 2,
73
+ 2,
74
+ 2
75
+ ],
76
+ "upsample_initial_channel": 512,
77
+ "upsample_kernel_sizes": [
78
+ 16,
79
+ 16,
80
+ 4,
81
+ 4,
82
+ 4
83
+ ],
84
+ "n_layers_q": 3,
85
+ "use_spectral_norm": false,
86
+ "gin_channels": 256,
87
+ "ssl_dim": 256,
88
+ "n_speakers": 200
89
+ },
90
+ "spk": {
91
+ "Dreizehn": 0
92
+ }
93
+ }
so-vits-svc-4.0/20230325-Dreizehn-44k/kmeans_10000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c0f201ce3cf292c9d7fc36bd8fc6a4b19626d4975c29c705a153aeccf8d996c
3
+ size 15447609
so-vits-svc-4.0/20230325-Dreizehn-44k/pruned/P_G_13600.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60b2d6d3272113095017126fd69b4298d53428c66168986c245d4b8875e6fd7d
3
+ size 180885829
so-vits-svc-4.0/20230325-Dreizehn-44k/pruned/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 800,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 0.0001,
8
+ "betas": [
9
+ 0.8,
10
+ 0.99
11
+ ],
12
+ "eps": 1e-09,
13
+ "batch_size": 6,
14
+ "fp16_run": false,
15
+ "lr_decay": 0.999875,
16
+ "segment_size": 10240,
17
+ "init_lr_ratio": 1,
18
+ "warmup_epochs": 0,
19
+ "c_mel": 45,
20
+ "c_kl": 1.0,
21
+ "use_sr": true,
22
+ "max_speclen": 512,
23
+ "port": "8001",
24
+ "keep_ckpts": 99
25
+ },
26
+ "data": {
27
+ "training_files": "filelists/train.txt",
28
+ "validation_files": "filelists/val.txt",
29
+ "max_wav_value": 32768.0,
30
+ "sampling_rate": 44100,
31
+ "filter_length": 2048,
32
+ "hop_length": 512,
33
+ "win_length": 2048,
34
+ "n_mel_channels": 80,
35
+ "mel_fmin": 0.0,
36
+ "mel_fmax": 22050
37
+ },
38
+ "model": {
39
+ "inter_channels": 192,
40
+ "hidden_channels": 192,
41
+ "filter_channels": 768,
42
+ "n_heads": 2,
43
+ "n_layers": 6,
44
+ "kernel_size": 3,
45
+ "p_dropout": 0.1,
46
+ "resblock": "1",
47
+ "resblock_kernel_sizes": [
48
+ 3,
49
+ 7,
50
+ 11
51
+ ],
52
+ "resblock_dilation_sizes": [
53
+ [
54
+ 1,
55
+ 3,
56
+ 5
57
+ ],
58
+ [
59
+ 1,
60
+ 3,
61
+ 5
62
+ ],
63
+ [
64
+ 1,
65
+ 3,
66
+ 5
67
+ ]
68
+ ],
69
+ "upsample_rates": [
70
+ 8,
71
+ 8,
72
+ 2,
73
+ 2,
74
+ 2
75
+ ],
76
+ "upsample_initial_channel": 512,
77
+ "upsample_kernel_sizes": [
78
+ 16,
79
+ 16,
80
+ 4,
81
+ 4,
82
+ 4
83
+ ],
84
+ "n_layers_q": 3,
85
+ "use_spectral_norm": false,
86
+ "gin_channels": 256,
87
+ "ssl_dim": 256,
88
+ "n_speakers": 200
89
+ },
90
+ "spk": {
91
+ "Dreizehn": 0
92
+ }
93
+ }
so-vits-svc-4.0/20230325-Dreizehn-44k/train.log ADDED
@@ -0,0 +1,405 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-03-25 14:44:39,506 44k INFO {'train': {'log_interval': 200, 'eval_interval': 800, 'seed': 1234, 'epochs': 10000, 'learning_rate': 0.0001, 'betas': [0.8, 0.99], 'eps': 1e-09, 'batch_size': 6, 'fp16_run': False, 'lr_decay': 0.999875, 'segment_size': 10240, 'init_lr_ratio': 1, 'warmup_epochs': 0, 'c_mel': 45, 'c_kl': 1.0, 'use_sr': True, 'max_speclen': 512, 'port': '8001', 'keep_ckpts': 99}, 'data': {'training_files': 'filelists/train.txt', 'validation_files': 'filelists/val.txt', 'max_wav_value': 32768.0, 'sampling_rate': 44100, 'filter_length': 2048, 'hop_length': 512, 'win_length': 2048, 'n_mel_channels': 80, 'mel_fmin': 0.0, 'mel_fmax': 22050}, 'model': {'inter_channels': 192, 'hidden_channels': 192, 'filter_channels': 768, 'n_heads': 2, 'n_layers': 6, 'kernel_size': 3, 'p_dropout': 0.1, 'resblock': '1', 'resblock_kernel_sizes': [3, 7, 11], 'resblock_dilation_sizes': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'upsample_rates': [8, 8, 2, 2, 2], 'upsample_initial_channel': 512, 'upsample_kernel_sizes': [16, 16, 4, 4, 4], 'n_layers_q': 3, 'use_spectral_norm': False, 'gin_channels': 256, 'ssl_dim': 256, 'n_speakers': 200}, 'spk': {'Dreizehn': 0}, 'model_dir': './logs/44k'}
2
+ 2023-03-25 14:44:39,506 44k WARNING /root/so-vits-svc-4.0 is not a git repository, therefore hash value comparison will be ignored.
3
+ 2023-03-25 14:44:42,455 44k INFO emb_g.weight is not in the checkpoint
4
+ 2023-03-25 14:44:42,528 44k INFO Loaded checkpoint './logs/44k/G_0.pth' (iteration 0)
5
+ 2023-03-25 14:44:42,698 44k INFO Loaded checkpoint './logs/44k/D_0.pth' (iteration 0)
6
+ 2023-03-25 14:44:49,806 44k INFO Train Epoch: 1 [0%]
7
+ 2023-03-25 14:44:49,807 44k INFO Losses: [2.940145492553711, 2.266042709350586, 7.757871150970459, 30.300212860107422, 4.057180881500244], step: 0, lr: 0.0001
8
+ 2023-03-25 14:44:54,651 44k INFO Saving model and optimizer state at iteration 1 to ./logs/44k/G_0.pth
9
+ 2023-03-25 14:44:56,000 44k INFO Saving model and optimizer state at iteration 1 to ./logs/44k/D_0.pth
10
+ 2023-03-25 14:45:24,127 44k INFO ====> Epoch: 1, cost 44.62 s
11
+ 2023-03-25 14:45:50,654 44k INFO ====> Epoch: 2, cost 26.53 s
12
+ 2023-03-25 14:46:17,848 44k INFO ====> Epoch: 3, cost 27.19 s
13
+ 2023-03-25 14:46:29,013 44k INFO Train Epoch: 4 [33%]
14
+ 2023-03-25 14:46:29,013 44k INFO Losses: [2.512408494949341, 2.084155321121216, 12.769187927246094, 23.224327087402344, 1.4044638872146606], step: 200, lr: 9.996250468730469e-05
15
+ 2023-03-25 14:46:45,744 44k INFO ====> Epoch: 4, cost 27.90 s
16
+ 2023-03-25 14:47:13,993 44k INFO ====> Epoch: 5, cost 28.25 s
17
+ 2023-03-25 14:47:42,175 44k INFO ====> Epoch: 6, cost 28.18 s
18
+ 2023-03-25 14:48:01,068 44k INFO Train Epoch: 7 [67%]
19
+ 2023-03-25 14:48:01,069 44k INFO Losses: [2.7341504096984863, 2.1355395317077637, 7.289322853088379, 17.03099250793457, 1.2272381782531738], step: 400, lr: 9.99250234335941e-05
20
+ 2023-03-25 14:48:09,326 44k INFO ====> Epoch: 7, cost 27.15 s
21
+ 2023-03-25 14:48:36,038 44k INFO ====> Epoch: 8, cost 26.71 s
22
+ 2023-03-25 14:49:03,725 44k INFO ====> Epoch: 9, cost 27.69 s
23
+ 2023-03-25 14:49:31,705 44k INFO ====> Epoch: 10, cost 27.98 s
24
+ 2023-03-25 14:49:35,219 44k INFO Train Epoch: 11 [0%]
25
+ 2023-03-25 14:49:35,220 44k INFO Losses: [2.4548428058624268, 2.215726137161255, 9.75304889678955, 19.071210861206055, 1.1653872728347778], step: 600, lr: 9.987507028906759e-05
26
+ 2023-03-25 14:49:59,796 44k INFO ====> Epoch: 11, cost 28.09 s
27
+ 2023-03-25 14:50:26,328 44k INFO ====> Epoch: 12, cost 26.53 s
28
+ 2023-03-25 14:50:53,108 44k INFO ====> Epoch: 13, cost 26.78 s
29
+ 2023-03-25 14:51:04,486 44k INFO Train Epoch: 14 [33%]
30
+ 2023-03-25 14:51:04,487 44k INFO Losses: [2.5523459911346436, 2.3060622215270996, 13.709832191467285, 22.02556610107422, 1.4591848850250244], step: 800, lr: 9.983762181915804e-05
31
+ 2023-03-25 14:51:08,756 44k INFO Saving model and optimizer state at iteration 14 to ./logs/44k/G_800.pth
32
+ 2023-03-25 14:51:09,936 44k INFO Saving model and optimizer state at iteration 14 to ./logs/44k/D_800.pth
33
+ 2023-03-25 14:51:25,899 44k INFO ====> Epoch: 14, cost 32.79 s
34
+ 2023-03-25 14:51:52,420 44k INFO ====> Epoch: 15, cost 26.52 s
35
+ 2023-03-25 14:52:19,045 44k INFO ====> Epoch: 16, cost 26.62 s
36
+ 2023-03-25 14:52:37,909 44k INFO Train Epoch: 17 [67%]
37
+ 2023-03-25 14:52:37,910 44k INFO Losses: [2.1840286254882812, 2.7090253829956055, 14.09952449798584, 20.535053253173828, 0.9362862706184387], step: 1000, lr: 9.980018739066937e-05
38
+ 2023-03-25 14:52:46,301 44k INFO ====> Epoch: 17, cost 27.26 s
39
+ 2023-03-25 14:53:12,974 44k INFO ====> Epoch: 18, cost 26.67 s
40
+ 2023-03-25 14:53:39,790 44k INFO ====> Epoch: 19, cost 26.82 s
41
+ 2023-03-25 14:54:06,380 44k INFO ====> Epoch: 20, cost 26.59 s
42
+ 2023-03-25 14:54:09,784 44k INFO Train Epoch: 21 [0%]
43
+ 2023-03-25 14:54:09,785 44k INFO Losses: [2.469074010848999, 2.314445972442627, 7.137829303741455, 14.518792152404785, 1.2171446084976196], step: 1200, lr: 9.975029665246193e-05
44
+ 2023-03-25 14:54:33,576 44k INFO ====> Epoch: 21, cost 27.20 s
45
+ 2023-03-25 14:55:00,354 44k INFO ====> Epoch: 22, cost 26.78 s
46
+ 2023-03-25 14:55:27,049 44k INFO ====> Epoch: 23, cost 26.70 s
47
+ 2023-03-25 14:55:38,268 44k INFO Train Epoch: 24 [33%]
48
+ 2023-03-25 14:55:38,269 44k INFO Losses: [2.2463409900665283, 2.576353073120117, 12.585086822509766, 20.897579193115234, 1.2720139026641846], step: 1400, lr: 9.971289496681757e-05
49
+ 2023-03-25 14:55:54,212 44k INFO ====> Epoch: 24, cost 27.16 s
50
+ 2023-03-25 14:56:20,732 44k INFO ====> Epoch: 25, cost 26.52 s
51
+ 2023-03-25 14:56:47,308 44k INFO ====> Epoch: 26, cost 26.58 s
52
+ 2023-03-25 14:57:06,114 44k INFO Train Epoch: 27 [67%]
53
+ 2023-03-25 14:57:06,115 44k INFO Losses: [2.5702061653137207, 2.543508768081665, 8.91149616241455, 18.796783447265625, 0.8504246473312378], step: 1600, lr: 9.967550730505221e-05
54
+ 2023-03-25 14:57:10,496 44k INFO Saving model and optimizer state at iteration 27 to ./logs/44k/G_1600.pth
55
+ 2023-03-25 14:57:11,683 44k INFO Saving model and optimizer state at iteration 27 to ./logs/44k/D_1600.pth
56
+ 2023-03-25 14:57:20,013 44k INFO ====> Epoch: 27, cost 32.71 s
57
+ 2023-03-25 14:57:46,724 44k INFO ====> Epoch: 28, cost 26.71 s
58
+ 2023-03-25 14:58:13,413 44k INFO ====> Epoch: 29, cost 26.69 s
59
+ 2023-03-25 14:58:39,922 44k INFO ====> Epoch: 30, cost 26.51 s
60
+ 2023-03-25 14:58:43,385 44k INFO Train Epoch: 31 [0%]
61
+ 2023-03-25 14:58:43,386 44k INFO Losses: [2.177131175994873, 2.665175437927246, 12.00287914276123, 19.802133560180664, 0.9337674379348755], step: 1800, lr: 9.962567889519979e-05
62
+ 2023-03-25 14:59:07,188 44k INFO ====> Epoch: 31, cost 27.27 s
63
+ 2023-03-25 14:59:33,991 44k INFO ====> Epoch: 32, cost 26.80 s
64
+ 2023-03-25 15:00:00,700 44k INFO ====> Epoch: 33, cost 26.71 s
65
+ 2023-03-25 15:00:11,887 44k INFO Train Epoch: 34 [33%]
66
+ 2023-03-25 15:00:11,888 44k INFO Losses: [2.2040858268737793, 2.457867383956909, 12.169000625610352, 20.236183166503906, 1.0930821895599365], step: 2000, lr: 9.95883239353732e-05
67
+ 2023-03-25 15:00:27,890 44k INFO ====> Epoch: 34, cost 27.19 s
68
+ 2023-03-25 15:00:54,577 44k INFO ====> Epoch: 35, cost 26.69 s
69
+ 2023-03-25 15:01:21,275 44k INFO ====> Epoch: 36, cost 26.70 s
70
+ 2023-03-25 15:01:40,077 44k INFO Train Epoch: 37 [67%]
71
+ 2023-03-25 15:01:40,078 44k INFO Losses: [2.5743024349212646, 2.2537405490875244, 6.321502208709717, 14.697388648986816, 0.5774912238121033], step: 2200, lr: 9.95509829819056e-05
72
+ 2023-03-25 15:01:48,399 44k INFO ====> Epoch: 37, cost 27.12 s
73
+ 2023-03-25 15:02:14,993 44k INFO ====> Epoch: 38, cost 26.59 s
74
+ 2023-03-25 15:02:41,823 44k INFO ====> Epoch: 39, cost 26.83 s
75
+ 2023-03-25 15:03:08,943 44k INFO ====> Epoch: 40, cost 27.12 s
76
+ 2023-03-25 15:03:12,327 44k INFO Train Epoch: 41 [0%]
77
+ 2023-03-25 15:03:12,328 44k INFO Losses: [2.5939323902130127, 2.1589159965515137, 5.888221263885498, 14.312878608703613, 0.8646751046180725], step: 2400, lr: 9.950121682254156e-05
78
+ 2023-03-25 15:03:16,742 44k INFO Saving model and optimizer state at iteration 41 to ./logs/44k/G_2400.pth
79
+ 2023-03-25 15:03:17,938 44k INFO Saving model and optimizer state at iteration 41 to ./logs/44k/D_2400.pth
80
+ 2023-03-25 15:03:41,768 44k INFO ====> Epoch: 41, cost 32.82 s
81
+ 2023-03-25 15:04:08,550 44k INFO ====> Epoch: 42, cost 26.78 s
82
+ 2023-03-25 15:04:37,085 44k INFO ====> Epoch: 43, cost 28.53 s
83
+ 2023-03-25 15:04:49,635 44k INFO Train Epoch: 44 [33%]
84
+ 2023-03-25 15:04:49,636 44k INFO Losses: [2.496485710144043, 2.323936700820923, 9.929926872253418, 19.4571533203125, 0.8525826334953308], step: 2600, lr: 9.94639085301583e-05
85
+ 2023-03-25 15:05:08,311 44k INFO ====> Epoch: 44, cost 31.23 s
86
+ 2023-03-25 15:05:38,820 44k INFO ====> Epoch: 45, cost 30.51 s
87
+ 2023-03-25 15:06:09,384 44k INFO ====> Epoch: 46, cost 30.56 s
88
+ 2023-03-25 15:06:29,222 44k INFO Train Epoch: 47 [67%]
89
+ 2023-03-25 15:06:29,223 44k INFO Losses: [2.4428462982177734, 2.2649857997894287, 11.198545455932617, 16.910449981689453, 0.8590737581253052], step: 2800, lr: 9.942661422663591e-05
90
+ 2023-03-25 15:06:37,524 44k INFO ====> Epoch: 47, cost 28.14 s
91
+ 2023-03-25 15:07:04,582 44k INFO ====> Epoch: 48, cost 27.06 s
92
+ 2023-03-25 15:07:31,254 44k INFO ====> Epoch: 49, cost 26.67 s
93
+ 2023-03-25 15:08:01,616 44k INFO ====> Epoch: 50, cost 30.36 s
94
+ 2023-03-25 15:08:04,934 44k INFO Train Epoch: 51 [0%]
95
+ 2023-03-25 15:08:04,936 44k INFO Losses: [2.133681535720825, 2.7440037727355957, 10.468965530395508, 16.335269927978516, 0.6788036227226257], step: 3000, lr: 9.937691023999092e-05
96
+ 2023-03-25 15:08:28,670 44k INFO ====> Epoch: 51, cost 27.05 s
97
+ 2023-03-25 15:08:55,226 44k INFO ====> Epoch: 52, cost 26.56 s
98
+ 2023-03-25 15:09:25,896 44k INFO ====> Epoch: 53, cost 30.67 s
99
+ 2023-03-25 15:09:37,230 44k INFO Train Epoch: 54 [33%]
100
+ 2023-03-25 15:09:37,231 44k INFO Losses: [2.2836685180664062, 2.4881534576416016, 14.617388725280762, 21.667917251586914, 0.8856773972511292], step: 3200, lr: 9.933964855674948e-05
101
+ 2023-03-25 15:09:41,465 44k INFO Saving model and optimizer state at iteration 54 to ./logs/44k/G_3200.pth
102
+ 2023-03-25 15:09:42,753 44k INFO Saving model and optimizer state at iteration 54 to ./logs/44k/D_3200.pth
103
+ 2023-03-25 15:09:58,674 44k INFO ====> Epoch: 54, cost 32.78 s
104
+ 2023-03-25 15:10:25,308 44k INFO ====> Epoch: 55, cost 26.63 s
105
+ 2023-03-25 15:10:52,009 44k INFO ====> Epoch: 56, cost 26.70 s
106
+ 2023-03-25 15:11:10,887 44k INFO Train Epoch: 57 [67%]
107
+ 2023-03-25 15:11:10,888 44k INFO Losses: [1.9383201599121094, 2.8536083698272705, 8.996301651000977, 15.366815567016602, 0.528129518032074], step: 3400, lr: 9.930240084489267e-05
108
+ 2023-03-25 15:11:19,170 44k INFO ====> Epoch: 57, cost 27.16 s
109
+ 2023-03-25 15:11:45,974 44k INFO ====> Epoch: 58, cost 26.80 s
110
+ 2023-03-25 15:12:12,588 44k INFO ====> Epoch: 59, cost 26.61 s
111
+ 2023-03-25 15:12:39,155 44k INFO ====> Epoch: 60, cost 26.57 s
112
+ 2023-03-25 15:12:42,558 44k INFO Train Epoch: 61 [0%]
113
+ 2023-03-25 15:12:42,559 44k INFO Losses: [2.4830005168914795, 2.3222198486328125, 9.070521354675293, 17.854820251464844, 1.0685895681381226], step: 3600, lr: 9.92527589532945e-05
114
+ 2023-03-25 15:13:06,196 44k INFO ====> Epoch: 61, cost 27.04 s
115
+ 2023-03-25 15:13:32,981 44k INFO ====> Epoch: 62, cost 26.79 s
116
+ 2023-03-25 15:13:59,808 44k INFO ====> Epoch: 63, cost 26.83 s
117
+ 2023-03-25 15:14:12,306 44k INFO Train Epoch: 64 [33%]
118
+ 2023-03-25 15:14:12,307 44k INFO Losses: [2.3387303352355957, 2.5378148555755615, 11.396452903747559, 21.508935928344727, 0.9471877217292786], step: 3800, lr: 9.921554382096622e-05
119
+ 2023-03-25 15:14:30,820 44k INFO ====> Epoch: 64, cost 31.01 s
120
+ 2023-03-25 15:14:57,504 44k INFO ====> Epoch: 65, cost 26.68 s
121
+ 2023-03-25 15:15:24,355 44k INFO ====> Epoch: 66, cost 26.85 s
122
+ 2023-03-25 15:15:43,113 44k INFO Train Epoch: 67 [67%]
123
+ 2023-03-25 15:15:43,114 44k INFO Losses: [2.4443154335021973, 2.425118923187256, 9.618949890136719, 19.272972106933594, 0.7614860534667969], step: 4000, lr: 9.917834264256819e-05
124
+ 2023-03-25 15:15:47,233 44k INFO Saving model and optimizer state at iteration 67 to ./logs/44k/G_4000.pth
125
+ 2023-03-25 15:15:48,506 44k INFO Saving model and optimizer state at iteration 67 to ./logs/44k/D_4000.pth
126
+ 2023-03-25 15:15:56,835 44k INFO ====> Epoch: 67, cost 32.48 s
127
+ 2023-03-25 15:16:23,426 44k INFO ====> Epoch: 68, cost 26.59 s
128
+ 2023-03-25 15:16:50,090 44k INFO ====> Epoch: 69, cost 26.66 s
129
+ 2023-03-25 15:17:16,871 44k INFO ====> Epoch: 70, cost 26.78 s
130
+ 2023-03-25 15:17:20,255 44k INFO Train Epoch: 71 [0%]
131
+ 2023-03-25 15:17:20,256 44k INFO Losses: [2.5252299308776855, 2.3594260215759277, 8.788890838623047, 17.564353942871094, 0.23537257313728333], step: 4200, lr: 9.912876276844171e-05
132
+ 2023-03-25 15:17:43,990 44k INFO ====> Epoch: 71, cost 27.12 s
133
+ 2023-03-25 15:18:10,744 44k INFO ====> Epoch: 72, cost 26.75 s
134
+ 2023-03-25 15:18:37,327 44k INFO ====> Epoch: 73, cost 26.58 s
135
+ 2023-03-25 15:18:48,595 44k INFO Train Epoch: 74 [33%]
136
+ 2023-03-25 15:18:48,596 44k INFO Losses: [2.375034809112549, 2.3302433490753174, 10.43783950805664, 13.910572052001953, 0.8278229236602783], step: 4400, lr: 9.909159412887068e-05
137
+ 2023-03-25 15:19:04,613 44k INFO ====> Epoch: 74, cost 27.29 s
138
+ 2023-03-25 15:19:31,231 44k INFO ====> Epoch: 75, cost 26.62 s
139
+ 2023-03-25 15:19:57,771 44k INFO ====> Epoch: 76, cost 26.54 s
140
+ 2023-03-25 15:20:16,714 44k INFO Train Epoch: 77 [67%]
141
+ 2023-03-25 15:20:16,715 44k INFO Losses: [2.7217864990234375, 1.882816195487976, 4.675782203674316, 8.256917953491211, 0.8678979277610779], step: 4600, lr: 9.905443942579728e-05
142
+ 2023-03-25 15:20:25,166 44k INFO ====> Epoch: 77, cost 27.40 s
143
+ 2023-03-25 15:20:51,789 44k INFO ====> Epoch: 78, cost 26.62 s
144
+ 2023-03-25 15:21:18,445 44k INFO ====> Epoch: 79, cost 26.66 s
145
+ 2023-03-25 15:21:46,733 44k INFO ====> Epoch: 80, cost 28.29 s
146
+ 2023-03-25 15:21:50,085 44k INFO Train Epoch: 81 [0%]
147
+ 2023-03-25 15:21:50,086 44k INFO Losses: [2.6070265769958496, 2.4579570293426514, 6.947027206420898, 16.77017593383789, 0.5974054932594299], step: 4800, lr: 9.900492149166423e-05
148
+ 2023-03-25 15:21:54,255 44k INFO Saving model and optimizer state at iteration 81 to ./logs/44k/G_4800.pth
149
+ 2023-03-25 15:21:55,531 44k INFO Saving model and optimizer state at iteration 81 to ./logs/44k/D_4800.pth
150
+ 2023-03-25 15:22:19,263 44k INFO ====> Epoch: 81, cost 32.53 s
151
+ 2023-03-25 15:22:46,437 44k INFO ====> Epoch: 82, cost 27.17 s
152
+ 2023-03-25 15:23:13,154 44k INFO ====> Epoch: 83, cost 26.72 s
153
+ 2023-03-25 15:23:24,324 44k INFO Train Epoch: 84 [33%]
154
+ 2023-03-25 15:23:24,325 44k INFO Losses: [2.5022735595703125, 2.247422456741333, 10.291221618652344, 19.483619689941406, 0.974785566329956], step: 5000, lr: 9.896779928676716e-05
155
+ 2023-03-25 15:23:40,217 44k INFO ====> Epoch: 84, cost 27.06 s
156
+ 2023-03-25 15:24:06,814 44k INFO ====> Epoch: 85, cost 26.60 s
157
+ 2023-03-25 15:24:33,777 44k INFO ====> Epoch: 86, cost 26.96 s
158
+ 2023-03-25 15:24:52,565 44k INFO Train Epoch: 87 [67%]
159
+ 2023-03-25 15:24:52,566 44k INFO Losses: [2.3922224044799805, 2.1541266441345215, 9.643497467041016, 14.642343521118164, 1.0375347137451172], step: 5200, lr: 9.89306910009569e-05
160
+ 2023-03-25 15:25:00,791 44k INFO ====> Epoch: 87, cost 27.01 s
161
+ 2023-03-25 15:25:27,394 44k INFO ====> Epoch: 88, cost 26.60 s
162
+ 2023-03-25 15:25:53,905 44k INFO ====> Epoch: 89, cost 26.51 s
163
+ 2023-03-25 15:26:20,528 44k INFO ====> Epoch: 90, cost 26.62 s
164
+ 2023-03-25 15:26:23,949 44k INFO Train Epoch: 91 [0%]
165
+ 2023-03-25 15:26:23,950 44k INFO Losses: [2.1768221855163574, 2.630232334136963, 11.107681274414062, 23.186044692993164, 0.9275988340377808], step: 5400, lr: 9.888123492943583e-05
166
+ 2023-03-25 15:26:47,862 44k INFO ====> Epoch: 91, cost 27.33 s
167
+ 2023-03-25 15:27:14,428 44k INFO ====> Epoch: 92, cost 26.57 s
168
+ 2023-03-25 15:27:41,017 44k INFO ====> Epoch: 93, cost 26.59 s
169
+ 2023-03-25 15:27:52,175 44k INFO Train Epoch: 94 [33%]
170
+ 2023-03-25 15:27:52,175 44k INFO Losses: [2.031240463256836, 2.6001627445220947, 12.831860542297363, 17.43545150756836, 0.6656062006950378], step: 5600, lr: 9.884415910120204e-05
171
+ 2023-03-25 15:27:56,288 44k INFO Saving model and optimizer state at iteration 94 to ./logs/44k/G_5600.pth
172
+ 2023-03-25 15:27:57,451 44k INFO Saving model and optimizer state at iteration 94 to ./logs/44k/D_5600.pth
173
+ 2023-03-25 15:28:13,565 44k INFO ====> Epoch: 94, cost 32.55 s
174
+ 2023-03-25 15:28:40,077 44k INFO ====> Epoch: 95, cost 26.51 s
175
+ 2023-03-25 15:29:06,692 44k INFO ====> Epoch: 96, cost 26.61 s
176
+ 2023-03-25 15:29:25,411 44k INFO Train Epoch: 97 [67%]
177
+ 2023-03-25 15:29:25,412 44k INFO Losses: [2.6305477619171143, 2.0891010761260986, 7.081687927246094, 14.060254096984863, 0.7252136468887329], step: 5800, lr: 9.880709717466598e-05
178
+ 2023-03-25 15:29:33,688 44k INFO ====> Epoch: 97, cost 27.00 s
179
+ 2023-03-25 15:30:00,268 44k INFO ====> Epoch: 98, cost 26.58 s
180
+ 2023-03-25 15:30:26,927 44k INFO ====> Epoch: 99, cost 26.66 s
181
+ 2023-03-25 15:30:53,739 44k INFO ====> Epoch: 100, cost 26.81 s
182
+ 2023-03-25 15:30:57,158 44k INFO Train Epoch: 101 [0%]
183
+ 2023-03-25 15:30:57,159 44k INFO Losses: [2.57724928855896, 2.4419920444488525, 9.626336097717285, 20.45102882385254, 0.8625250458717346], step: 6000, lr: 9.875770288847208e-05
184
+ 2023-03-25 15:31:20,916 44k INFO ====> Epoch: 101, cost 27.18 s
185
+ 2023-03-25 15:31:47,606 44k INFO ====> Epoch: 102, cost 26.69 s
186
+ 2023-03-25 15:32:14,189 44k INFO ====> Epoch: 103, cost 26.58 s
187
+ 2023-03-25 15:32:25,493 44k INFO Train Epoch: 104 [33%]
188
+ 2023-03-25 15:32:25,494 44k INFO Losses: [2.366825819015503, 2.278965711593628, 9.511296272277832, 17.809585571289062, 0.773294985294342], step: 6200, lr: 9.872067337896332e-05
189
+ 2023-03-25 15:32:41,421 44k INFO ====> Epoch: 104, cost 27.23 s
190
+ 2023-03-25 15:33:10,300 44k INFO ====> Epoch: 105, cost 28.88 s
191
+ 2023-03-25 15:33:37,022 44k INFO ====> Epoch: 106, cost 26.72 s
192
+ 2023-03-25 15:33:56,020 44k INFO Train Epoch: 107 [67%]
193
+ 2023-03-25 15:33:56,021 44k INFO Losses: [2.5079104900360107, 2.207770347595215, 8.176128387451172, 14.970914840698242, 0.7647779583930969], step: 6400, lr: 9.868365775378495e-05
194
+ 2023-03-25 15:34:00,185 44k INFO Saving model and optimizer state at iteration 107 to ./logs/44k/G_6400.pth
195
+ 2023-03-25 15:34:01,302 44k INFO Saving model and optimizer state at iteration 107 to ./logs/44k/D_6400.pth
196
+ 2023-03-25 15:34:09,595 44k INFO ====> Epoch: 107, cost 32.57 s
197
+ 2023-03-25 15:34:36,317 44k INFO ====> Epoch: 108, cost 26.72 s
198
+ 2023-03-25 15:35:03,186 44k INFO ====> Epoch: 109, cost 26.87 s
199
+ 2023-03-25 15:35:30,271 44k INFO ====> Epoch: 110, cost 27.09 s
200
+ 2023-03-25 15:35:33,805 44k INFO Train Epoch: 111 [0%]
201
+ 2023-03-25 15:35:33,806 44k INFO Losses: [2.3396925926208496, 2.208421230316162, 8.759860038757324, 13.542329788208008, 0.7297613620758057], step: 6600, lr: 9.863432517573002e-05
202
+ 2023-03-25 15:35:58,905 44k INFO ====> Epoch: 111, cost 28.63 s
203
+ 2023-03-25 15:36:26,578 44k INFO ====> Epoch: 112, cost 27.67 s
204
+ 2023-03-25 15:36:54,361 44k INFO ====> Epoch: 113, cost 27.78 s
205
+ 2023-03-25 15:37:05,864 44k INFO Train Epoch: 114 [33%]
206
+ 2023-03-25 15:37:05,865 44k INFO Losses: [2.2866804599761963, 2.403698444366455, 14.162016868591309, 20.087860107421875, 0.8307042121887207], step: 6800, lr: 9.859734192708044e-05
207
+ 2023-03-25 15:37:22,493 44k INFO ====> Epoch: 114, cost 28.13 s
208
+ 2023-03-25 15:37:50,437 44k INFO ====> Epoch: 115, cost 27.94 s
209
+ 2023-03-25 15:38:17,891 44k INFO ====> Epoch: 116, cost 27.45 s
210
+ 2023-03-25 15:38:37,530 44k INFO Train Epoch: 117 [67%]
211
+ 2023-03-25 15:38:37,531 44k INFO Losses: [2.4838879108428955, 2.1345486640930176, 9.153286933898926, 19.457439422607422, 0.8745102882385254], step: 7000, lr: 9.85603725454156e-05
212
+ 2023-03-25 15:38:45,975 44k INFO ====> Epoch: 117, cost 28.08 s
213
+ 2023-03-25 15:39:13,778 44k INFO ====> Epoch: 118, cost 27.80 s
214
+ 2023-03-25 15:39:41,699 44k INFO ====> Epoch: 119, cost 27.92 s
215
+ 2023-03-25 15:40:09,333 44k INFO ====> Epoch: 120, cost 27.63 s
216
+ 2023-03-25 15:40:12,966 44k INFO Train Epoch: 121 [0%]
217
+ 2023-03-25 15:40:12,967 44k INFO Losses: [2.2715530395507812, 2.6235780715942383, 12.198266983032227, 21.405956268310547, 1.187570333480835], step: 7200, lr: 9.851110159840781e-05
218
+ 2023-03-25 15:40:17,172 44k INFO Saving model and optimizer state at iteration 121 to ./logs/44k/G_7200.pth
219
+ 2023-03-25 15:40:18,433 44k INFO Saving model and optimizer state at iteration 121 to ./logs/44k/D_7200.pth
220
+ 2023-03-25 15:40:43,478 44k INFO ====> Epoch: 121, cost 34.15 s
221
+ 2023-03-25 15:41:11,318 44k INFO ====> Epoch: 122, cost 27.84 s
222
+ 2023-03-25 15:41:38,933 44k INFO ====> Epoch: 123, cost 27.62 s
223
+ 2023-03-25 15:41:50,572 44k INFO Train Epoch: 124 [33%]
224
+ 2023-03-25 15:41:50,573 44k INFO Losses: [2.6159331798553467, 1.9919030666351318, 10.321922302246094, 17.80287742614746, 0.8138943910598755], step: 7400, lr: 9.847416455282387e-05
225
+ 2023-03-25 15:42:07,225 44k INFO ====> Epoch: 124, cost 28.29 s
226
+ 2023-03-25 15:42:35,119 44k INFO ====> Epoch: 125, cost 27.89 s
227
+ 2023-03-25 15:43:02,554 44k INFO ====> Epoch: 126, cost 27.43 s
228
+ 2023-03-25 15:43:22,175 44k INFO Train Epoch: 127 [67%]
229
+ 2023-03-25 15:43:22,176 44k INFO Losses: [2.354686737060547, 2.4857425689697266, 8.663324356079102, 13.985154151916504, 0.27259084582328796], step: 7600, lr: 9.84372413569007e-05
230
+ 2023-03-25 15:43:30,787 44k INFO ====> Epoch: 127, cost 28.23 s
231
+ 2023-03-25 15:43:58,989 44k INFO ====> Epoch: 128, cost 28.20 s
232
+ 2023-03-25 15:44:26,506 44k INFO ====> Epoch: 129, cost 27.52 s
233
+ 2023-03-25 15:44:54,120 44k INFO ====> Epoch: 130, cost 27.61 s
234
+ 2023-03-25 15:44:57,800 44k INFO Train Epoch: 131 [0%]
235
+ 2023-03-25 15:44:57,801 44k INFO Losses: [2.362338066101074, 2.521932363510132, 10.726180076599121, 18.021610260009766, 0.8564594984054565], step: 7800, lr: 9.838803196394459e-05
236
+ 2023-03-25 15:45:22,279 44k INFO ====> Epoch: 131, cost 28.16 s
237
+ 2023-03-25 15:45:50,028 44k INFO ====> Epoch: 132, cost 27.75 s
238
+ 2023-03-25 15:46:17,576 44k INFO ====> Epoch: 133, cost 27.55 s
239
+ 2023-03-25 15:46:29,163 44k INFO Train Epoch: 134 [33%]
240
+ 2023-03-25 15:46:29,164 44k INFO Losses: [2.1983914375305176, 2.453049898147583, 10.3521089553833, 19.192773818969727, 1.035571575164795], step: 8000, lr: 9.835114106370493e-05
241
+ 2023-03-25 15:46:33,555 44k INFO Saving model and optimizer state at iteration 134 to ./logs/44k/G_8000.pth
242
+ 2023-03-25 15:46:34,883 44k INFO Saving model and optimizer state at iteration 134 to ./logs/44k/D_8000.pth
243
+ 2023-03-25 15:46:51,651 44k INFO ====> Epoch: 134, cost 34.07 s
244
+ 2023-03-25 15:47:19,182 44k INFO ====> Epoch: 135, cost 27.53 s
245
+ 2023-03-25 15:47:47,310 44k INFO ====> Epoch: 136, cost 28.13 s
246
+ 2023-03-25 15:48:07,405 44k INFO Train Epoch: 137 [67%]
247
+ 2023-03-25 15:48:07,406 44k INFO Losses: [2.5634498596191406, 2.1999950408935547, 9.676583290100098, 19.181659698486328, 0.4880225360393524], step: 8200, lr: 9.831426399582366e-05
248
+ 2023-03-25 15:48:16,176 44k INFO ====> Epoch: 137, cost 28.87 s
249
+ 2023-03-25 15:48:44,491 44k INFO ====> Epoch: 138, cost 28.32 s
250
+ 2023-03-25 15:49:12,145 44k INFO ====> Epoch: 139, cost 27.65 s
251
+ 2023-03-25 15:49:39,691 44k INFO ====> Epoch: 140, cost 27.55 s
252
+ 2023-03-25 15:49:43,562 44k INFO Train Epoch: 141 [0%]
253
+ 2023-03-25 15:49:43,563 44k INFO Losses: [2.1003589630126953, 2.6904215812683105, 11.157882690429688, 18.755420684814453, 0.7207870483398438], step: 8400, lr: 9.826511608001993e-05
254
+ 2023-03-25 15:50:08,579 44k INFO ====> Epoch: 141, cost 28.89 s
255
+ 2023-03-25 15:50:36,778 44k INFO ====> Epoch: 142, cost 28.20 s
256
+ 2023-03-25 15:51:04,578 44k INFO ====> Epoch: 143, cost 27.80 s
257
+ 2023-03-25 15:51:16,196 44k INFO Train Epoch: 144 [33%]
258
+ 2023-03-25 15:51:16,197 44k INFO Losses: [2.6339874267578125, 2.1276330947875977, 10.583199501037598, 17.78839874267578, 0.9267172813415527], step: 8600, lr: 9.822827126747529e-05
259
+ 2023-03-25 15:51:33,003 44k INFO ====> Epoch: 144, cost 28.43 s
260
+ 2023-03-25 15:52:01,405 44k INFO ====> Epoch: 145, cost 28.40 s
261
+ 2023-03-25 15:52:28,990 44k INFO ====> Epoch: 146, cost 27.58 s
262
+ 2023-03-25 15:52:48,861 44k INFO Train Epoch: 147 [67%]
263
+ 2023-03-25 15:52:48,862 44k INFO Losses: [2.5725722312927246, 2.5007855892181396, 7.138617515563965, 13.22614860534668, 0.8027871251106262], step: 8800, lr: 9.819144027000834e-05
264
+ 2023-03-25 15:52:53,303 44k INFO Saving model and optimizer state at iteration 147 to ./logs/44k/G_8800.pth
265
+ 2023-03-25 15:52:54,482 44k INFO Saving model and optimizer state at iteration 147 to ./logs/44k/D_8800.pth
266
+ 2023-03-25 15:53:03,077 44k INFO ====> Epoch: 147, cost 34.09 s
267
+ 2023-03-25 15:53:30,602 44k INFO ====> Epoch: 148, cost 27.53 s
268
+ 2023-03-25 15:53:58,172 44k INFO ====> Epoch: 149, cost 27.57 s
269
+ 2023-03-25 15:54:25,707 44k INFO ====> Epoch: 150, cost 27.54 s
270
+ 2023-03-25 15:54:29,284 44k INFO Train Epoch: 151 [0%]
271
+ 2023-03-25 15:54:29,285 44k INFO Losses: [2.169490337371826, 2.8044533729553223, 11.827469825744629, 18.990873336791992, 0.9448695778846741], step: 9000, lr: 9.814235375455375e-05
272
+ 2023-03-25 15:54:54,505 44k INFO ====> Epoch: 151, cost 28.80 s
273
+ 2023-03-25 15:55:22,519 44k INFO ====> Epoch: 152, cost 28.01 s
274
+ 2023-03-25 15:55:50,002 44k INFO ====> Epoch: 153, cost 27.48 s
275
+ 2023-03-25 15:56:01,731 44k INFO Train Epoch: 154 [33%]
276
+ 2023-03-25 15:56:01,732 44k INFO Losses: [2.2980504035949707, 2.498497247695923, 9.984493255615234, 16.98595428466797, 0.7363108992576599], step: 9200, lr: 9.810555497212693e-05
277
+ 2023-03-25 15:56:18,105 44k INFO ====> Epoch: 154, cost 28.10 s
278
+ 2023-03-25 15:56:45,928 44k INFO ====> Epoch: 155, cost 27.82 s
279
+ 2023-03-25 15:57:13,862 44k INFO ====> Epoch: 156, cost 27.93 s
280
+ 2023-03-25 15:57:33,710 44k INFO Train Epoch: 157 [67%]
281
+ 2023-03-25 15:57:33,711 44k INFO Losses: [2.4687323570251465, 2.260420083999634, 7.67680025100708, 14.242744445800781, 0.7204187512397766], step: 9400, lr: 9.806876998751865e-05
282
+ 2023-03-25 15:57:42,365 44k INFO ====> Epoch: 157, cost 28.50 s
283
+ 2023-03-25 15:58:10,840 44k INFO ====> Epoch: 158, cost 28.47 s
284
+ 2023-03-25 15:58:38,494 44k INFO ====> Epoch: 159, cost 27.65 s
285
+ 2023-03-25 15:59:06,177 44k INFO ====> Epoch: 160, cost 27.68 s
286
+ 2023-03-25 15:59:10,017 44k INFO Train Epoch: 161 [0%]
287
+ 2023-03-25 15:59:10,018 44k INFO Losses: [2.2403976917266846, 2.689331531524658, 10.701247215270996, 18.250370025634766, 0.5913345217704773], step: 9600, lr: 9.801974479570593e-05
288
+ 2023-03-25 15:59:14,725 44k INFO Saving model and optimizer state at iteration 161 to ./logs/44k/G_9600.pth
289
+ 2023-03-25 15:59:15,941 44k INFO Saving model and optimizer state at iteration 161 to ./logs/44k/D_9600.pth
290
+ 2023-03-25 15:59:40,465 44k INFO ====> Epoch: 161, cost 34.29 s
291
+ 2023-03-25 16:00:08,135 44k INFO ====> Epoch: 162, cost 27.67 s
292
+ 2023-03-25 16:00:36,280 44k INFO ====> Epoch: 163, cost 28.15 s
293
+ 2023-03-25 16:00:47,950 44k INFO Train Epoch: 164 [33%]
294
+ 2023-03-25 16:00:47,951 44k INFO Losses: [2.291417360305786, 2.2027976512908936, 9.914422035217285, 14.098004341125488, 0.9870694875717163], step: 9800, lr: 9.798299198589162e-05
295
+ 2023-03-25 16:01:04,546 44k INFO ====> Epoch: 164, cost 28.27 s
296
+ 2023-03-25 16:01:32,268 44k INFO ====> Epoch: 165, cost 27.72 s
297
+ 2023-03-25 16:02:00,779 44k INFO ====> Epoch: 166, cost 28.51 s
298
+ 2023-03-25 16:02:20,253 44k INFO Train Epoch: 167 [67%]
299
+ 2023-03-25 16:02:20,255 44k INFO Losses: [2.5774083137512207, 2.229320764541626, 11.070889472961426, 18.494190216064453, 0.5609847903251648], step: 10000, lr: 9.794625295665828e-05
300
+ 2023-03-25 16:02:29,005 44k INFO ====> Epoch: 167, cost 28.23 s
301
+ 2023-03-25 16:02:57,331 44k INFO ====> Epoch: 168, cost 28.33 s
302
+ 2023-03-25 16:03:25,540 44k INFO ====> Epoch: 169, cost 28.21 s
303
+ 2023-03-25 16:03:53,268 44k INFO ====> Epoch: 170, cost 27.73 s
304
+ 2023-03-25 16:03:56,836 44k INFO Train Epoch: 171 [0%]
305
+ 2023-03-25 16:03:56,837 44k INFO Losses: [2.506603717803955, 2.397548198699951, 10.223786354064941, 18.886367797851562, 0.70171058177948], step: 10200, lr: 9.789728901187598e-05
306
+ 2023-03-25 16:04:21,639 44k INFO ====> Epoch: 171, cost 28.37 s
307
+ 2023-03-25 16:04:49,952 44k INFO ====> Epoch: 172, cost 28.31 s
308
+ 2023-03-25 16:05:18,292 44k INFO ====> Epoch: 173, cost 28.34 s
309
+ 2023-03-25 16:05:30,033 44k INFO Train Epoch: 174 [33%]
310
+ 2023-03-25 16:05:30,035 44k INFO Losses: [2.3508377075195312, 2.1366045475006104, 10.397187232971191, 17.74555015563965, 0.7728736996650696], step: 10400, lr: 9.786058211724074e-05
311
+ 2023-03-25 16:05:34,367 44k INFO Saving model and optimizer state at iteration 174 to ./logs/44k/G_10400.pth
312
+ 2023-03-25 16:05:35,703 44k INFO Saving model and optimizer state at iteration 174 to ./logs/44k/D_10400.pth
313
+ 2023-03-25 16:05:52,335 44k INFO ====> Epoch: 174, cost 34.04 s
314
+ 2023-03-25 16:06:20,613 44k INFO ====> Epoch: 175, cost 28.28 s
315
+ 2023-03-25 16:06:48,223 44k INFO ====> Epoch: 176, cost 27.61 s
316
+ 2023-03-25 16:07:07,774 44k INFO Train Epoch: 177 [67%]
317
+ 2023-03-25 16:07:07,776 44k INFO Losses: [2.3116414546966553, 2.3159842491149902, 11.632226943969727, 18.547286987304688, 0.9227203726768494], step: 10600, lr: 9.782388898597041e-05
318
+ 2023-03-25 16:07:16,338 44k INFO ====> Epoch: 177, cost 28.11 s
319
+ 2023-03-25 16:07:44,185 44k INFO ====> Epoch: 178, cost 27.85 s
320
+ 2023-03-25 16:08:11,900 44k INFO ====> Epoch: 179, cost 27.71 s
321
+ 2023-03-25 16:08:40,228 44k INFO ====> Epoch: 180, cost 28.33 s
322
+ 2023-03-25 16:08:43,879 44k INFO Train Epoch: 181 [0%]
323
+ 2023-03-25 16:08:43,880 44k INFO Losses: [2.40667724609375, 2.235522985458374, 9.322698593139648, 17.760761260986328, 0.6746690273284912], step: 10800, lr: 9.777498621170277e-05
324
+ 2023-03-25 16:09:08,780 44k INFO ====> Epoch: 181, cost 28.55 s
325
+ 2023-03-25 16:09:35,925 44k INFO ====> Epoch: 182, cost 27.15 s
326
+ 2023-03-25 16:10:02,656 44k INFO ====> Epoch: 183, cost 26.73 s
327
+ 2023-03-25 16:10:14,074 44k INFO Train Epoch: 184 [33%]
328
+ 2023-03-25 16:10:14,075 44k INFO Losses: [2.315093755722046, 2.372858762741089, 11.837750434875488, 19.7078800201416, 0.28122562170028687], step: 11000, lr: 9.773832517488488e-05
329
+ 2023-03-25 16:10:30,227 44k INFO ====> Epoch: 184, cost 27.57 s
330
+ 2023-03-25 16:10:56,929 44k INFO ====> Epoch: 185, cost 26.70 s
331
+ 2023-03-25 16:11:23,751 44k INFO ====> Epoch: 186, cost 26.82 s
332
+ 2023-03-25 16:11:44,391 44k INFO Train Epoch: 187 [67%]
333
+ 2023-03-25 16:11:44,392 44k INFO Losses: [2.32973575592041, 2.6208066940307617, 10.395020484924316, 16.26879119873047, 0.376319944858551], step: 11200, lr: 9.77016778842374e-05
334
+ 2023-03-25 16:11:48,550 44k INFO Saving model and optimizer state at iteration 187 to ./logs/44k/G_11200.pth
335
+ 2023-03-25 16:11:49,725 44k INFO Saving model and optimizer state at iteration 187 to ./logs/44k/D_11200.pth
336
+ 2023-03-25 16:11:58,048 44k INFO ====> Epoch: 187, cost 34.30 s
337
+ 2023-03-25 16:12:24,750 44k INFO ====> Epoch: 188, cost 26.70 s
338
+ 2023-03-25 16:12:51,376 44k INFO ====> Epoch: 189, cost 26.63 s
339
+ 2023-03-25 16:13:18,199 44k INFO ====> Epoch: 190, cost 26.82 s
340
+ 2023-03-25 16:13:21,590 44k INFO Train Epoch: 191 [0%]
341
+ 2023-03-25 16:13:21,591 44k INFO Losses: [2.3605268001556396, 2.5424208641052246, 7.856480121612549, 14.165212631225586, 0.7268055081367493], step: 11400, lr: 9.765283620406429e-05
342
+ 2023-03-25 16:13:45,845 44k INFO ====> Epoch: 191, cost 27.65 s
343
+ 2023-03-25 16:14:12,604 44k INFO ====> Epoch: 192, cost 26.76 s
344
+ 2023-03-25 16:14:43,116 44k INFO ====> Epoch: 193, cost 30.51 s
345
+ 2023-03-25 16:14:54,428 44k INFO Train Epoch: 194 [33%]
346
+ 2023-03-25 16:14:54,429 44k INFO Losses: [2.7298760414123535, 1.8387227058410645, 6.033370494842529, 13.283061981201172, 0.9616314768791199], step: 11600, lr: 9.761622096777372e-05
347
+ 2023-03-25 16:15:10,675 44k INFO ====> Epoch: 194, cost 27.56 s
348
+ 2023-03-25 16:15:37,838 44k INFO ====> Epoch: 195, cost 27.16 s
349
+ 2023-03-25 16:16:05,901 44k INFO ====> Epoch: 196, cost 28.06 s
350
+ 2023-03-25 16:16:25,641 44k INFO Train Epoch: 197 [67%]
351
+ 2023-03-25 16:16:25,642 44k INFO Losses: [2.3175127506256104, 2.385716199874878, 11.155668258666992, 17.777097702026367, 0.3992789685726166], step: 11800, lr: 9.757961946048049e-05
352
+ 2023-03-25 16:16:34,258 44k INFO ====> Epoch: 197, cost 28.36 s
353
+ 2023-03-25 16:17:02,036 44k INFO ====> Epoch: 198, cost 27.78 s
354
+ 2023-03-25 16:17:29,701 44k INFO ====> Epoch: 199, cost 27.67 s
355
+ 2023-03-25 16:17:56,622 44k INFO ====> Epoch: 200, cost 26.92 s
356
+ 2023-03-25 16:18:00,038 44k INFO Train Epoch: 201 [0%]
357
+ 2023-03-25 16:18:00,039 44k INFO Losses: [2.126671314239502, 2.953439712524414, 10.01817798614502, 18.810077667236328, 0.6492265462875366], step: 12000, lr: 9.753083879807726e-05
358
+ 2023-03-25 16:18:04,307 44k INFO Saving model and optimizer state at iteration 201 to ./logs/44k/G_12000.pth
359
+ 2023-03-25 16:18:05,443 44k INFO Saving model and optimizer state at iteration 201 to ./logs/44k/D_12000.pth
360
+ 2023-03-25 16:18:30,206 44k INFO ====> Epoch: 201, cost 33.58 s
361
+ 2023-03-25 16:18:58,235 44k INFO ====> Epoch: 202, cost 28.03 s
362
+ 2023-03-25 16:19:26,346 44k INFO ====> Epoch: 203, cost 28.11 s
363
+ 2023-03-25 16:19:38,537 44k INFO Train Epoch: 204 [33%]
364
+ 2023-03-25 16:19:38,538 44k INFO Losses: [2.341325521469116, 2.3583366870880127, 9.117012023925781, 14.948786735534668, 0.8733038902282715], step: 12200, lr: 9.749426930509556e-05
365
+ 2023-03-25 16:19:55,293 44k INFO ====> Epoch: 204, cost 28.95 s
366
+ 2023-03-25 16:20:24,093 44k INFO ====> Epoch: 205, cost 28.80 s
367
+ 2023-03-25 16:20:52,270 44k INFO ====> Epoch: 206, cost 28.18 s
368
+ 2023-03-25 16:21:12,828 44k INFO Train Epoch: 207 [67%]
369
+ 2023-03-25 16:21:12,829 44k INFO Losses: [2.518537759780884, 2.0252599716186523, 8.098270416259766, 14.455513000488281, 0.4925019145011902], step: 12400, lr: 9.745771352395957e-05
370
+ 2023-03-25 16:21:21,530 44k INFO ====> Epoch: 207, cost 29.26 s
371
+ 2023-03-25 16:21:49,885 44k INFO ====> Epoch: 208, cost 28.36 s
372
+ 2023-03-25 16:22:18,738 44k INFO ====> Epoch: 209, cost 28.85 s
373
+ 2023-03-25 16:22:46,546 44k INFO ====> Epoch: 210, cost 27.81 s
374
+ 2023-03-25 16:22:50,385 44k INFO Train Epoch: 211 [0%]
375
+ 2023-03-25 16:22:50,386 44k INFO Losses: [2.324171781539917, 2.519193649291992, 9.139647483825684, 19.630090713500977, 0.3635931611061096], step: 12600, lr: 9.740899380309685e-05
376
+ 2023-03-25 16:23:15,533 44k INFO ====> Epoch: 211, cost 28.99 s
377
+ 2023-03-25 16:23:44,387 44k INFO ====> Epoch: 212, cost 28.85 s
378
+ 2023-03-25 16:24:12,486 44k INFO ====> Epoch: 213, cost 28.10 s
379
+ 2023-03-25 16:24:24,567 44k INFO Train Epoch: 214 [33%]
380
+ 2023-03-25 16:24:24,569 44k INFO Losses: [2.4054665565490723, 2.550511121749878, 12.541123390197754, 19.724668502807617, 1.1664975881576538], step: 12800, lr: 9.7372469996277e-05
381
+ 2023-03-25 16:24:29,234 44k INFO Saving model and optimizer state at iteration 214 to ./logs/44k/G_12800.pth
382
+ 2023-03-25 16:24:30,661 44k INFO Saving model and optimizer state at iteration 214 to ./logs/44k/D_12800.pth
383
+ 2023-03-25 16:24:47,567 44k INFO ====> Epoch: 214, cost 35.08 s
384
+ 2023-03-25 16:25:16,212 44k INFO ====> Epoch: 215, cost 28.65 s
385
+ 2023-03-25 16:25:44,316 44k INFO ====> Epoch: 216, cost 28.10 s
386
+ 2023-03-25 16:26:04,228 44k INFO Train Epoch: 217 [67%]
387
+ 2023-03-25 16:26:04,229 44k INFO Losses: [2.41851544380188, 2.3247714042663574, 9.172240257263184, 16.376054763793945, 0.6352742910385132], step: 13000, lr: 9.733595988417275e-05
388
+ 2023-03-25 16:26:12,954 44k INFO ====> Epoch: 217, cost 28.64 s
389
+ 2023-03-25 16:26:41,838 44k INFO ====> Epoch: 218, cost 28.88 s
390
+ 2023-03-25 16:27:10,745 44k INFO ====> Epoch: 219, cost 28.91 s
391
+ 2023-03-25 16:27:38,859 44k INFO ====> Epoch: 220, cost 28.11 s
392
+ 2023-03-25 16:27:42,563 44k INFO Train Epoch: 221 [0%]
393
+ 2023-03-25 16:27:42,564 44k INFO Losses: [2.5801851749420166, 2.454096555709839, 8.478157043457031, 17.067672729492188, 0.8212155103683472], step: 13200, lr: 9.728730102871649e-05
394
+ 2023-03-25 16:28:07,414 44k INFO ====> Epoch: 221, cost 28.55 s
395
+ 2023-03-25 16:28:35,465 44k INFO ====> Epoch: 222, cost 28.05 s
396
+ 2023-03-25 16:29:04,152 44k INFO ====> Epoch: 223, cost 28.69 s
397
+ 2023-03-25 16:29:16,136 44k INFO Train Epoch: 224 [33%]
398
+ 2023-03-25 16:29:16,137 44k INFO Losses: [2.606571674346924, 2.4211103916168213, 9.568647384643555, 16.520126342773438, 0.5133330225944519], step: 13400, lr: 9.725082285098293e-05
399
+ 2023-03-25 16:29:33,043 44k INFO ====> Epoch: 224, cost 28.89 s
400
+ 2023-03-25 16:30:01,279 44k INFO ====> Epoch: 225, cost 28.24 s
401
+ 2023-03-25 16:30:29,534 44k INFO ====> Epoch: 226, cost 28.26 s
402
+ 2023-03-25 16:30:49,224 44k INFO Train Epoch: 227 [67%]
403
+ 2023-03-25 16:30:49,226 44k INFO Losses: [2.2993545532226562, 2.6229958534240723, 10.669507026672363, 18.169801712036133, 0.5818897485733032], step: 13600, lr: 9.721435835085619e-05
404
+ 2023-03-25 16:30:53,905 44k INFO Saving model and optimizer state at iteration 227 to ./logs/44k/G_13600.pth
405
+ 2023-03-25 16:30:55,146 44k INFO Saving model and optimizer state at iteration 227 to ./logs/44k/D_13600.pth
so-vits-svc-4.0/20230325-DreizehnNS-44k/best/D_13600.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa8a86282eef397e4078906c7447c91bd9b29ddefb6aef61cb261ec0bd10fc30
3
+ size 561098185
so-vits-svc-4.0/20230325-DreizehnNS-44k/best/G_13600.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24849a2d9adcd52bfea6cb2d8b24cd9b2a3cb27dac1b3afec79c2e46b2415721
3
+ size 542789405
so-vits-svc-4.0/20230325-DreizehnNS-44k/best/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 800,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 0.0001,
8
+ "betas": [
9
+ 0.8,
10
+ 0.99
11
+ ],
12
+ "eps": 1e-09,
13
+ "batch_size": 6,
14
+ "fp16_run": false,
15
+ "lr_decay": 0.999875,
16
+ "segment_size": 10240,
17
+ "init_lr_ratio": 1,
18
+ "warmup_epochs": 0,
19
+ "c_mel": 45,
20
+ "c_kl": 1.0,
21
+ "use_sr": true,
22
+ "max_speclen": 512,
23
+ "port": "8001",
24
+ "keep_ckpts": 99
25
+ },
26
+ "data": {
27
+ "training_files": "filelists/train.txt",
28
+ "validation_files": "filelists/val.txt",
29
+ "max_wav_value": 32768.0,
30
+ "sampling_rate": 44100,
31
+ "filter_length": 2048,
32
+ "hop_length": 512,
33
+ "win_length": 2048,
34
+ "n_mel_channels": 80,
35
+ "mel_fmin": 0.0,
36
+ "mel_fmax": 22050
37
+ },
38
+ "model": {
39
+ "inter_channels": 192,
40
+ "hidden_channels": 192,
41
+ "filter_channels": 768,
42
+ "n_heads": 2,
43
+ "n_layers": 6,
44
+ "kernel_size": 3,
45
+ "p_dropout": 0.1,
46
+ "resblock": "1",
47
+ "resblock_kernel_sizes": [
48
+ 3,
49
+ 7,
50
+ 11
51
+ ],
52
+ "resblock_dilation_sizes": [
53
+ [
54
+ 1,
55
+ 3,
56
+ 5
57
+ ],
58
+ [
59
+ 1,
60
+ 3,
61
+ 5
62
+ ],
63
+ [
64
+ 1,
65
+ 3,
66
+ 5
67
+ ]
68
+ ],
69
+ "upsample_rates": [
70
+ 8,
71
+ 8,
72
+ 2,
73
+ 2,
74
+ 2
75
+ ],
76
+ "upsample_initial_channel": 512,
77
+ "upsample_kernel_sizes": [
78
+ 16,
79
+ 16,
80
+ 4,
81
+ 4,
82
+ 4
83
+ ],
84
+ "n_layers_q": 3,
85
+ "use_spectral_norm": false,
86
+ "gin_channels": 256,
87
+ "ssl_dim": 256,
88
+ "n_speakers": 200
89
+ },
90
+ "spk": {
91
+ "DreizehnNS": 0
92
+ }
93
+ }
so-vits-svc-4.0/20230325-DreizehnNS-44k/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 800,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 0.0001,
8
+ "betas": [
9
+ 0.8,
10
+ 0.99
11
+ ],
12
+ "eps": 1e-09,
13
+ "batch_size": 6,
14
+ "fp16_run": false,
15
+ "lr_decay": 0.999875,
16
+ "segment_size": 10240,
17
+ "init_lr_ratio": 1,
18
+ "warmup_epochs": 0,
19
+ "c_mel": 45,
20
+ "c_kl": 1.0,
21
+ "use_sr": true,
22
+ "max_speclen": 512,
23
+ "port": "8001",
24
+ "keep_ckpts": 99
25
+ },
26
+ "data": {
27
+ "training_files": "filelists/train.txt",
28
+ "validation_files": "filelists/val.txt",
29
+ "max_wav_value": 32768.0,
30
+ "sampling_rate": 44100,
31
+ "filter_length": 2048,
32
+ "hop_length": 512,
33
+ "win_length": 2048,
34
+ "n_mel_channels": 80,
35
+ "mel_fmin": 0.0,
36
+ "mel_fmax": 22050
37
+ },
38
+ "model": {
39
+ "inter_channels": 192,
40
+ "hidden_channels": 192,
41
+ "filter_channels": 768,
42
+ "n_heads": 2,
43
+ "n_layers": 6,
44
+ "kernel_size": 3,
45
+ "p_dropout": 0.1,
46
+ "resblock": "1",
47
+ "resblock_kernel_sizes": [
48
+ 3,
49
+ 7,
50
+ 11
51
+ ],
52
+ "resblock_dilation_sizes": [
53
+ [
54
+ 1,
55
+ 3,
56
+ 5
57
+ ],
58
+ [
59
+ 1,
60
+ 3,
61
+ 5
62
+ ],
63
+ [
64
+ 1,
65
+ 3,
66
+ 5
67
+ ]
68
+ ],
69
+ "upsample_rates": [
70
+ 8,
71
+ 8,
72
+ 2,
73
+ 2,
74
+ 2
75
+ ],
76
+ "upsample_initial_channel": 512,
77
+ "upsample_kernel_sizes": [
78
+ 16,
79
+ 16,
80
+ 4,
81
+ 4,
82
+ 4
83
+ ],
84
+ "n_layers_q": 3,
85
+ "use_spectral_norm": false,
86
+ "gin_channels": 256,
87
+ "ssl_dim": 256,
88
+ "n_speakers": 200
89
+ },
90
+ "spk": {
91
+ "DreizehnNS": 0
92
+ }
93
+ }
so-vits-svc-4.0/20230325-DreizehnNS-44k/kmeans_10000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:742a89428e88e07a30f3bcb88e5a24444b3b628cf17130e15b3017f9c60b8272
3
+ size 15457593
so-vits-svc-4.0/20230325-DreizehnNS-44k/pruned/P_G_13600.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d603d1a052d95afc6ecab31f9f633701a72f05c621f40ede6f3d74054cc9fce9
3
+ size 180885829
so-vits-svc-4.0/20230325-DreizehnNS-44k/pruned/config.json ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 800,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 0.0001,
8
+ "betas": [
9
+ 0.8,
10
+ 0.99
11
+ ],
12
+ "eps": 1e-09,
13
+ "batch_size": 6,
14
+ "fp16_run": false,
15
+ "lr_decay": 0.999875,
16
+ "segment_size": 10240,
17
+ "init_lr_ratio": 1,
18
+ "warmup_epochs": 0,
19
+ "c_mel": 45,
20
+ "c_kl": 1.0,
21
+ "use_sr": true,
22
+ "max_speclen": 512,
23
+ "port": "8001",
24
+ "keep_ckpts": 99
25
+ },
26
+ "data": {
27
+ "training_files": "filelists/train.txt",
28
+ "validation_files": "filelists/val.txt",
29
+ "max_wav_value": 32768.0,
30
+ "sampling_rate": 44100,
31
+ "filter_length": 2048,
32
+ "hop_length": 512,
33
+ "win_length": 2048,
34
+ "n_mel_channels": 80,
35
+ "mel_fmin": 0.0,
36
+ "mel_fmax": 22050
37
+ },
38
+ "model": {
39
+ "inter_channels": 192,
40
+ "hidden_channels": 192,
41
+ "filter_channels": 768,
42
+ "n_heads": 2,
43
+ "n_layers": 6,
44
+ "kernel_size": 3,
45
+ "p_dropout": 0.1,
46
+ "resblock": "1",
47
+ "resblock_kernel_sizes": [
48
+ 3,
49
+ 7,
50
+ 11
51
+ ],
52
+ "resblock_dilation_sizes": [
53
+ [
54
+ 1,
55
+ 3,
56
+ 5
57
+ ],
58
+ [
59
+ 1,
60
+ 3,
61
+ 5
62
+ ],
63
+ [
64
+ 1,
65
+ 3,
66
+ 5
67
+ ]
68
+ ],
69
+ "upsample_rates": [
70
+ 8,
71
+ 8,
72
+ 2,
73
+ 2,
74
+ 2
75
+ ],
76
+ "upsample_initial_channel": 512,
77
+ "upsample_kernel_sizes": [
78
+ 16,
79
+ 16,
80
+ 4,
81
+ 4,
82
+ 4
83
+ ],
84
+ "n_layers_q": 3,
85
+ "use_spectral_norm": false,
86
+ "gin_channels": 256,
87
+ "ssl_dim": 256,
88
+ "n_speakers": 200
89
+ },
90
+ "spk": {
91
+ "DreizehnNS": 0
92
+ }
93
+ }
so-vits-svc-4.0/20230325-DreizehnNS-44k/train.log ADDED
@@ -0,0 +1,592 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2023-03-25 16:47:27,269 44k INFO {'train': {'log_interval': 200, 'eval_interval': 800, 'seed': 1234, 'epochs': 10000, 'learning_rate': 0.0001, 'betas': [0.8, 0.99], 'eps': 1e-09, 'batch_size': 6, 'fp16_run': False, 'lr_decay': 0.999875, 'segment_size': 10240, 'init_lr_ratio': 1, 'warmup_epochs': 0, 'c_mel': 45, 'c_kl': 1.0, 'use_sr': True, 'max_speclen': 512, 'port': '8001', 'keep_ckpts': 99}, 'data': {'training_files': 'filelists/train.txt', 'validation_files': 'filelists/val.txt', 'max_wav_value': 32768.0, 'sampling_rate': 44100, 'filter_length': 2048, 'hop_length': 512, 'win_length': 2048, 'n_mel_channels': 80, 'mel_fmin': 0.0, 'mel_fmax': 22050}, 'model': {'inter_channels': 192, 'hidden_channels': 192, 'filter_channels': 768, 'n_heads': 2, 'n_layers': 6, 'kernel_size': 3, 'p_dropout': 0.1, 'resblock': '1', 'resblock_kernel_sizes': [3, 7, 11], 'resblock_dilation_sizes': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'upsample_rates': [8, 8, 2, 2, 2], 'upsample_initial_channel': 512, 'upsample_kernel_sizes': [16, 16, 4, 4, 4], 'n_layers_q': 3, 'use_spectral_norm': False, 'gin_channels': 256, 'ssl_dim': 256, 'n_speakers': 200}, 'spk': {'DreizehnNS': 0}, 'model_dir': './logs/44k'}
2
+ 2023-03-25 16:47:27,269 44k WARNING /root/so-vits-svc-4.0 is not a git repository, therefore hash value comparison will be ignored.
3
+ 2023-03-25 16:47:30,338 44k INFO emb_g.weight is not in the checkpoint
4
+ 2023-03-25 16:47:30,394 44k INFO Loaded checkpoint './logs/44k/G_0.pth' (iteration 0)
5
+ 2023-03-25 16:47:30,551 44k INFO Loaded checkpoint './logs/44k/D_0.pth' (iteration 0)
6
+ 2023-03-25 16:47:38,222 44k INFO Train Epoch: 1 [0%]
7
+ 2023-03-25 16:47:38,223 44k INFO Losses: [2.6974782943725586, 2.4092516899108887, 14.672739028930664, 37.46104431152344, 4.290388584136963], step: 0, lr: 0.0001
8
+ 2023-03-25 16:47:43,503 44k INFO Saving model and optimizer state at iteration 1 to ./logs/44k/G_0.pth
9
+ 2023-03-25 16:47:44,765 44k INFO Saving model and optimizer state at iteration 1 to ./logs/44k/D_0.pth
10
+ 2023-03-25 16:48:02,181 44k INFO ====> Epoch: 1, cost 34.91 s
11
+ 2023-03-25 16:48:19,046 44k INFO ====> Epoch: 2, cost 16.86 s
12
+ 2023-03-25 16:48:35,900 44k INFO ====> Epoch: 3, cost 16.85 s
13
+ 2023-03-25 16:48:53,235 44k INFO ====> Epoch: 4, cost 17.33 s
14
+ 2023-03-25 16:49:09,884 44k INFO ====> Epoch: 5, cost 16.65 s
15
+ 2023-03-25 16:49:26,410 44k INFO ====> Epoch: 6, cost 16.53 s
16
+ 2023-03-25 16:49:30,788 44k INFO Train Epoch: 7 [6%]
17
+ 2023-03-25 16:49:30,790 44k INFO Losses: [2.2324411869049072, 2.718120574951172, 8.101375579833984, 15.547173500061035, 0.7784972190856934], step: 200, lr: 9.99250234335941e-05
18
+ 2023-03-25 16:49:43,790 44k INFO ====> Epoch: 7, cost 17.38 s
19
+ 2023-03-25 16:50:00,579 44k INFO ====> Epoch: 8, cost 16.79 s
20
+ 2023-03-25 16:50:17,010 44k INFO ====> Epoch: 9, cost 16.43 s
21
+ 2023-03-25 16:50:33,841 44k INFO ====> Epoch: 10, cost 16.83 s
22
+ 2023-03-25 16:50:50,880 44k INFO ====> Epoch: 11, cost 17.04 s
23
+ 2023-03-25 16:51:07,417 44k INFO ====> Epoch: 12, cost 16.54 s
24
+ 2023-03-25 16:51:12,576 44k INFO Train Epoch: 13 [12%]
25
+ 2023-03-25 16:51:12,577 44k INFO Losses: [2.325521469116211, 2.3794703483581543, 8.109458923339844, 15.86378002166748, 1.0891475677490234], step: 400, lr: 9.98501030820433e-05
26
+ 2023-03-25 16:51:25,360 44k INFO ====> Epoch: 13, cost 17.94 s
27
+ 2023-03-25 16:51:42,186 44k INFO ====> Epoch: 14, cost 16.83 s
28
+ 2023-03-25 16:51:59,000 44k INFO ====> Epoch: 15, cost 16.81 s
29
+ 2023-03-25 16:52:15,819 44k INFO ====> Epoch: 16, cost 16.82 s
30
+ 2023-03-25 16:52:32,638 44k INFO ====> Epoch: 17, cost 16.82 s
31
+ 2023-03-25 16:52:49,372 44k INFO ====> Epoch: 18, cost 16.73 s
32
+ 2023-03-25 16:52:55,578 44k INFO Train Epoch: 19 [18%]
33
+ 2023-03-25 16:52:55,579 44k INFO Losses: [2.597564220428467, 2.6969094276428223, 12.184596061706543, 19.04696273803711, 0.6627763509750366], step: 600, lr: 9.977523890319963e-05
34
+ 2023-03-25 16:53:06,934 44k INFO ====> Epoch: 19, cost 17.56 s
35
+ 2023-03-25 16:53:23,846 44k INFO ====> Epoch: 20, cost 16.91 s
36
+ 2023-03-25 16:53:40,397 44k INFO ====> Epoch: 21, cost 16.55 s
37
+ 2023-03-25 16:53:57,546 44k INFO ====> Epoch: 22, cost 17.15 s
38
+ 2023-03-25 16:54:14,372 44k INFO ====> Epoch: 23, cost 16.83 s
39
+ 2023-03-25 16:54:31,052 44k INFO ====> Epoch: 24, cost 16.68 s
40
+ 2023-03-25 16:54:37,717 44k INFO Train Epoch: 25 [24%]
41
+ 2023-03-25 16:54:37,718 44k INFO Losses: [2.5308003425598145, 2.7162580490112305, 9.232569694519043, 15.944351196289062, 0.7880488038063049], step: 800, lr: 9.970043085494672e-05
42
+ 2023-03-25 16:54:42,256 44k INFO Saving model and optimizer state at iteration 25 to ./logs/44k/G_800.pth
43
+ 2023-03-25 16:54:43,576 44k INFO Saving model and optimizer state at iteration 25 to ./logs/44k/D_800.pth
44
+ 2023-03-25 16:54:54,478 44k INFO ====> Epoch: 25, cost 23.43 s
45
+ 2023-03-25 16:55:11,205 44k INFO ====> Epoch: 26, cost 16.73 s
46
+ 2023-03-25 16:55:28,258 44k INFO ====> Epoch: 27, cost 17.05 s
47
+ 2023-03-25 16:55:44,954 44k INFO ====> Epoch: 28, cost 16.70 s
48
+ 2023-03-25 16:56:01,987 44k INFO ====> Epoch: 29, cost 17.03 s
49
+ 2023-03-25 16:56:18,844 44k INFO ====> Epoch: 30, cost 16.86 s
50
+ 2023-03-25 16:56:26,243 44k INFO Train Epoch: 31 [30%]
51
+ 2023-03-25 16:56:26,244 44k INFO Losses: [2.428475856781006, 2.539283037185669, 9.996885299682617, 21.852493286132812, 0.8732186555862427], step: 1000, lr: 9.962567889519979e-05
52
+ 2023-03-25 16:56:36,533 44k INFO ====> Epoch: 31, cost 17.69 s
53
+ 2023-03-25 16:56:53,197 44k INFO ====> Epoch: 32, cost 16.66 s
54
+ 2023-03-25 16:57:09,628 44k INFO ====> Epoch: 33, cost 16.43 s
55
+ 2023-03-25 16:57:26,475 44k INFO ====> Epoch: 34, cost 16.85 s
56
+ 2023-03-25 16:57:42,775 44k INFO ====> Epoch: 35, cost 16.30 s
57
+ 2023-03-25 16:57:59,365 44k INFO ====> Epoch: 36, cost 16.59 s
58
+ 2023-03-25 16:58:07,766 44k INFO Train Epoch: 37 [36%]
59
+ 2023-03-25 16:58:07,767 44k INFO Losses: [2.5228097438812256, 2.288203001022339, 11.588765144348145, 20.175125122070312, 0.6351476311683655], step: 1200, lr: 9.95509829819056e-05
60
+ 2023-03-25 16:58:16,719 44k INFO ====> Epoch: 37, cost 17.35 s
61
+ 2023-03-25 16:58:33,397 44k INFO ====> Epoch: 38, cost 16.68 s
62
+ 2023-03-25 16:58:49,877 44k INFO ====> Epoch: 39, cost 16.48 s
63
+ 2023-03-25 16:59:06,535 44k INFO ====> Epoch: 40, cost 16.66 s
64
+ 2023-03-25 16:59:23,000 44k INFO ====> Epoch: 41, cost 16.47 s
65
+ 2023-03-25 16:59:40,216 44k INFO ====> Epoch: 42, cost 17.22 s
66
+ 2023-03-25 16:59:49,295 44k INFO Train Epoch: 43 [42%]
67
+ 2023-03-25 16:59:49,297 44k INFO Losses: [2.5152947902679443, 2.0511679649353027, 7.733269214630127, 15.723326683044434, 0.8497045040130615], step: 1400, lr: 9.947634307304244e-05
68
+ 2023-03-25 16:59:58,173 44k INFO ====> Epoch: 43, cost 17.96 s
69
+ 2023-03-25 17:00:14,901 44k INFO ====> Epoch: 44, cost 16.73 s
70
+ 2023-03-25 17:00:31,479 44k INFO ====> Epoch: 45, cost 16.58 s
71
+ 2023-03-25 17:00:48,089 44k INFO ====> Epoch: 46, cost 16.61 s
72
+ 2023-03-25 17:01:04,848 44k INFO ====> Epoch: 47, cost 16.76 s
73
+ 2023-03-25 17:01:21,377 44k INFO ====> Epoch: 48, cost 16.53 s
74
+ 2023-03-25 17:01:31,234 44k INFO Train Epoch: 49 [48%]
75
+ 2023-03-25 17:01:31,235 44k INFO Losses: [2.0633249282836914, 2.451162576675415, 13.993906021118164, 21.134328842163086, 0.9099781513214111], step: 1600, lr: 9.940175912662009e-05
76
+ 2023-03-25 17:01:35,874 44k INFO Saving model and optimizer state at iteration 49 to ./logs/44k/G_1600.pth
77
+ 2023-03-25 17:01:37,075 44k INFO Saving model and optimizer state at iteration 49 to ./logs/44k/D_1600.pth
78
+ 2023-03-25 17:01:44,622 44k INFO ====> Epoch: 49, cost 23.24 s
79
+ 2023-03-25 17:02:01,509 44k INFO ====> Epoch: 50, cost 16.89 s
80
+ 2023-03-25 17:02:18,009 44k INFO ====> Epoch: 51, cost 16.50 s
81
+ 2023-03-25 17:02:35,156 44k INFO ====> Epoch: 52, cost 17.15 s
82
+ 2023-03-25 17:02:51,799 44k INFO ====> Epoch: 53, cost 16.64 s
83
+ 2023-03-25 17:03:08,333 44k INFO ====> Epoch: 54, cost 16.53 s
84
+ 2023-03-25 17:03:18,881 44k INFO Train Epoch: 55 [55%]
85
+ 2023-03-25 17:03:18,882 44k INFO Losses: [2.7328269481658936, 2.0261311531066895, 10.961967468261719, 17.05063819885254, 1.2199965715408325], step: 1800, lr: 9.932723110067987e-05
86
+ 2023-03-25 17:03:25,395 44k INFO ====> Epoch: 55, cost 17.06 s
87
+ 2023-03-25 17:03:41,918 44k INFO ====> Epoch: 56, cost 16.52 s
88
+ 2023-03-25 17:03:58,504 44k INFO ====> Epoch: 57, cost 16.59 s
89
+ 2023-03-25 17:04:15,265 44k INFO ====> Epoch: 58, cost 16.76 s
90
+ 2023-03-25 17:04:31,725 44k INFO ====> Epoch: 59, cost 16.46 s
91
+ 2023-03-25 17:04:48,454 44k INFO ====> Epoch: 60, cost 16.73 s
92
+ 2023-03-25 17:04:59,685 44k INFO Train Epoch: 61 [61%]
93
+ 2023-03-25 17:04:59,686 44k INFO Losses: [2.453200101852417, 2.5324313640594482, 10.589628219604492, 22.162248611450195, 1.0994495153427124], step: 2000, lr: 9.92527589532945e-05
94
+ 2023-03-25 17:05:05,700 44k INFO ====> Epoch: 61, cost 17.25 s
95
+ 2023-03-25 17:05:22,609 44k INFO ====> Epoch: 62, cost 16.91 s
96
+ 2023-03-25 17:05:39,301 44k INFO ====> Epoch: 63, cost 16.69 s
97
+ 2023-03-25 17:05:55,873 44k INFO ====> Epoch: 64, cost 16.57 s
98
+ 2023-03-25 17:06:13,560 44k INFO ====> Epoch: 65, cost 17.69 s
99
+ 2023-03-25 17:06:30,219 44k INFO ====> Epoch: 66, cost 16.66 s
100
+ 2023-03-25 17:06:42,692 44k INFO Train Epoch: 67 [67%]
101
+ 2023-03-25 17:06:42,693 44k INFO Losses: [2.295553684234619, 2.4008660316467285, 9.892032623291016, 17.572246551513672, 0.5077621936798096], step: 2200, lr: 9.917834264256819e-05
102
+ 2023-03-25 17:06:47,857 44k INFO ====> Epoch: 67, cost 17.64 s
103
+ 2023-03-25 17:07:04,741 44k INFO ====> Epoch: 68, cost 16.88 s
104
+ 2023-03-25 17:07:21,623 44k INFO ====> Epoch: 69, cost 16.88 s
105
+ 2023-03-25 17:07:38,193 44k INFO ====> Epoch: 70, cost 16.57 s
106
+ 2023-03-25 17:07:54,870 44k INFO ====> Epoch: 71, cost 16.68 s
107
+ 2023-03-25 17:08:11,664 44k INFO ====> Epoch: 72, cost 16.79 s
108
+ 2023-03-25 17:08:25,468 44k INFO Train Epoch: 73 [73%]
109
+ 2023-03-25 17:08:25,470 44k INFO Losses: [2.599912405014038, 2.0814945697784424, 6.905517101287842, 13.569595336914062, 0.7708091735839844], step: 2400, lr: 9.910398212663652e-05
110
+ 2023-03-25 17:08:29,876 44k INFO Saving model and optimizer state at iteration 73 to ./logs/44k/G_2400.pth
111
+ 2023-03-25 17:08:31,092 44k INFO Saving model and optimizer state at iteration 73 to ./logs/44k/D_2400.pth
112
+ 2023-03-25 17:08:35,432 44k INFO ====> Epoch: 73, cost 23.77 s
113
+ 2023-03-25 17:08:51,850 44k INFO ====> Epoch: 74, cost 16.42 s
114
+ 2023-03-25 17:09:08,371 44k INFO ====> Epoch: 75, cost 16.52 s
115
+ 2023-03-25 17:09:25,135 44k INFO ====> Epoch: 76, cost 16.76 s
116
+ 2023-03-25 17:09:41,551 44k INFO ====> Epoch: 77, cost 16.42 s
117
+ 2023-03-25 17:09:58,751 44k INFO ====> Epoch: 78, cost 17.20 s
118
+ 2023-03-25 17:10:12,408 44k INFO Train Epoch: 79 [79%]
119
+ 2023-03-25 17:10:12,409 44k INFO Losses: [2.381700038909912, 2.7567107677459717, 7.031135082244873, 11.056401252746582, 0.745064914226532], step: 2600, lr: 9.902967736366644e-05
120
+ 2023-03-25 17:10:15,809 44k INFO ====> Epoch: 79, cost 17.06 s
121
+ 2023-03-25 17:10:32,728 44k INFO ====> Epoch: 80, cost 16.92 s
122
+ 2023-03-25 17:10:49,486 44k INFO ====> Epoch: 81, cost 16.76 s
123
+ 2023-03-25 17:11:06,316 44k INFO ====> Epoch: 82, cost 16.83 s
124
+ 2023-03-25 17:11:23,200 44k INFO ====> Epoch: 83, cost 16.88 s
125
+ 2023-03-25 17:11:39,866 44k INFO ====> Epoch: 84, cost 16.67 s
126
+ 2023-03-25 17:11:54,418 44k INFO Train Epoch: 85 [85%]
127
+ 2023-03-25 17:11:54,419 44k INFO Losses: [2.4440665245056152, 2.255263328552246, 6.086838722229004, 10.995237350463867, 0.8823586106300354], step: 2800, lr: 9.895542831185631e-05
128
+ 2023-03-25 17:11:57,170 44k INFO ====> Epoch: 85, cost 17.30 s
129
+ 2023-03-25 17:12:14,015 44k INFO ====> Epoch: 86, cost 16.84 s
130
+ 2023-03-25 17:12:31,117 44k INFO ====> Epoch: 87, cost 17.10 s
131
+ 2023-03-25 17:12:47,612 44k INFO ====> Epoch: 88, cost 16.50 s
132
+ 2023-03-25 17:13:04,423 44k INFO ====> Epoch: 89, cost 16.81 s
133
+ 2023-03-25 17:13:21,588 44k INFO ====> Epoch: 90, cost 17.17 s
134
+ 2023-03-25 17:13:36,796 44k INFO Train Epoch: 91 [91%]
135
+ 2023-03-25 17:13:36,797 44k INFO Losses: [2.305187940597534, 2.6297731399536133, 10.468490600585938, 22.139997482299805, 0.8177512288093567], step: 3000, lr: 9.888123492943583e-05
136
+ 2023-03-25 17:13:38,623 44k INFO ====> Epoch: 91, cost 17.04 s
137
+ 2023-03-25 17:13:55,148 44k INFO ====> Epoch: 92, cost 16.52 s
138
+ 2023-03-25 17:14:12,020 44k INFO ====> Epoch: 93, cost 16.87 s
139
+ 2023-03-25 17:14:28,756 44k INFO ====> Epoch: 94, cost 16.74 s
140
+ 2023-03-25 17:14:45,594 44k INFO ====> Epoch: 95, cost 16.84 s
141
+ 2023-03-25 17:15:02,370 44k INFO ====> Epoch: 96, cost 16.78 s
142
+ 2023-03-25 17:15:18,311 44k INFO Train Epoch: 97 [97%]
143
+ 2023-03-25 17:15:18,312 44k INFO Losses: [2.6312460899353027, 2.329620838165283, 7.526452541351318, 11.057599067687988, 0.6342429518699646], step: 3200, lr: 9.880709717466598e-05
144
+ 2023-03-25 17:15:22,926 44k INFO Saving model and optimizer state at iteration 97 to ./logs/44k/G_3200.pth
145
+ 2023-03-25 17:15:24,087 44k INFO Saving model and optimizer state at iteration 97 to ./logs/44k/D_3200.pth
146
+ 2023-03-25 17:15:25,306 44k INFO ====> Epoch: 97, cost 22.94 s
147
+ 2023-03-25 17:15:42,129 44k INFO ====> Epoch: 98, cost 16.82 s
148
+ 2023-03-25 17:15:58,976 44k INFO ====> Epoch: 99, cost 16.85 s
149
+ 2023-03-25 17:16:15,653 44k INFO ====> Epoch: 100, cost 16.68 s
150
+ 2023-03-25 17:16:32,914 44k INFO ====> Epoch: 101, cost 17.26 s
151
+ 2023-03-25 17:16:50,033 44k INFO ====> Epoch: 102, cost 17.12 s
152
+ 2023-03-25 17:17:06,781 44k INFO ====> Epoch: 103, cost 16.75 s
153
+ 2023-03-25 17:17:10,819 44k INFO Train Epoch: 104 [3%]
154
+ 2023-03-25 17:17:10,821 44k INFO Losses: [2.338289737701416, 2.391996145248413, 10.017964363098145, 17.78423309326172, 0.8379142880439758], step: 3400, lr: 9.872067337896332e-05
155
+ 2023-03-25 17:17:24,042 44k INFO ====> Epoch: 104, cost 17.26 s
156
+ 2023-03-25 17:17:40,807 44k INFO ====> Epoch: 105, cost 16.76 s
157
+ 2023-03-25 17:17:57,366 44k INFO ====> Epoch: 106, cost 16.56 s
158
+ 2023-03-25 17:18:14,377 44k INFO ====> Epoch: 107, cost 17.01 s
159
+ 2023-03-25 17:18:31,046 44k INFO ====> Epoch: 108, cost 16.67 s
160
+ 2023-03-25 17:18:47,635 44k INFO ====> Epoch: 109, cost 16.59 s
161
+ 2023-03-25 17:18:52,568 44k INFO Train Epoch: 110 [9%]
162
+ 2023-03-25 17:18:52,569 44k INFO Losses: [2.822373390197754, 2.349123239517212, 4.166832447052002, 13.0591459274292, 1.1201118230819702], step: 3600, lr: 9.864665600773098e-05
163
+ 2023-03-25 17:19:04,926 44k INFO ====> Epoch: 110, cost 17.29 s
164
+ 2023-03-25 17:19:21,696 44k INFO ====> Epoch: 111, cost 16.77 s
165
+ 2023-03-25 17:19:38,208 44k INFO ====> Epoch: 112, cost 16.51 s
166
+ 2023-03-25 17:19:54,872 44k INFO ====> Epoch: 113, cost 16.66 s
167
+ 2023-03-25 17:20:11,555 44k INFO ====> Epoch: 114, cost 16.68 s
168
+ 2023-03-25 17:20:28,509 44k INFO ====> Epoch: 115, cost 16.95 s
169
+ 2023-03-25 17:20:34,295 44k INFO Train Epoch: 116 [15%]
170
+ 2023-03-25 17:20:34,296 44k INFO Losses: [2.072659492492676, 2.792006015777588, 11.296175003051758, 15.94166374206543, 0.21819163858890533], step: 3800, lr: 9.857269413218213e-05
171
+ 2023-03-25 17:20:46,053 44k INFO ====> Epoch: 116, cost 17.54 s
172
+ 2023-03-25 17:21:03,288 44k INFO ====> Epoch: 117, cost 17.24 s
173
+ 2023-03-25 17:21:19,708 44k INFO ====> Epoch: 118, cost 16.42 s
174
+ 2023-03-25 17:21:36,111 44k INFO ====> Epoch: 119, cost 16.40 s
175
+ 2023-03-25 17:21:52,543 44k INFO ====> Epoch: 120, cost 16.43 s
176
+ 2023-03-25 17:22:09,435 44k INFO ====> Epoch: 121, cost 16.89 s
177
+ 2023-03-25 17:22:15,757 44k INFO Train Epoch: 122 [21%]
178
+ 2023-03-25 17:22:15,758 44k INFO Losses: [2.796874523162842, 2.1552910804748535, 7.83256196975708, 11.650614738464355, 0.48031163215637207], step: 4000, lr: 9.8498787710708e-05
179
+ 2023-03-25 17:22:20,248 44k INFO Saving model and optimizer state at iteration 122 to ./logs/44k/G_4000.pth
180
+ 2023-03-25 17:22:21,400 44k INFO Saving model and optimizer state at iteration 122 to ./logs/44k/D_4000.pth
181
+ 2023-03-25 17:22:33,009 44k INFO ====> Epoch: 122, cost 23.57 s
182
+ 2023-03-25 17:22:50,276 44k INFO ====> Epoch: 123, cost 17.27 s
183
+ 2023-03-25 17:23:06,706 44k INFO ====> Epoch: 124, cost 16.43 s
184
+ 2023-03-25 17:23:23,424 44k INFO ====> Epoch: 125, cost 16.72 s
185
+ 2023-03-25 17:23:39,997 44k INFO ====> Epoch: 126, cost 16.57 s
186
+ 2023-03-25 17:23:56,588 44k INFO ====> Epoch: 127, cost 16.59 s
187
+ 2023-03-25 17:24:03,701 44k INFO Train Epoch: 128 [27%]
188
+ 2023-03-25 17:24:03,702 44k INFO Losses: [2.1375842094421387, 3.02132511138916, 10.848799705505371, 14.873762130737305, 0.668673574924469], step: 4200, lr: 9.842493670173108e-05
189
+ 2023-03-25 17:24:13,908 44k INFO ====> Epoch: 128, cost 17.32 s
190
+ 2023-03-25 17:24:30,679 44k INFO ====> Epoch: 129, cost 16.77 s
191
+ 2023-03-25 17:24:47,504 44k INFO ====> Epoch: 130, cost 16.82 s
192
+ 2023-03-25 17:25:04,152 44k INFO ====> Epoch: 131, cost 16.65 s
193
+ 2023-03-25 17:25:21,742 44k INFO ====> Epoch: 132, cost 17.59 s
194
+ 2023-03-25 17:25:39,064 44k INFO ====> Epoch: 133, cost 17.32 s
195
+ 2023-03-25 17:25:46,939 44k INFO Train Epoch: 134 [33%]
196
+ 2023-03-25 17:25:46,940 44k INFO Losses: [2.4770865440368652, 2.198981761932373, 5.717620849609375, 12.555487632751465, 0.7245680093765259], step: 4400, lr: 9.835114106370493e-05
197
+ 2023-03-25 17:25:56,287 44k INFO ====> Epoch: 134, cost 17.22 s
198
+ 2023-03-25 17:26:13,283 44k INFO ====> Epoch: 135, cost 17.00 s
199
+ 2023-03-25 17:26:29,684 44k INFO ====> Epoch: 136, cost 16.40 s
200
+ 2023-03-25 17:26:46,292 44k INFO ====> Epoch: 137, cost 16.61 s
201
+ 2023-03-25 17:27:03,013 44k INFO ====> Epoch: 138, cost 16.72 s
202
+ 2023-03-25 17:27:19,649 44k INFO ====> Epoch: 139, cost 16.64 s
203
+ 2023-03-25 17:27:28,415 44k INFO Train Epoch: 140 [39%]
204
+ 2023-03-25 17:27:28,417 44k INFO Losses: [2.2845919132232666, 3.015265941619873, 8.828485488891602, 16.861722946166992, 1.0159695148468018], step: 4600, lr: 9.827740075511432e-05
205
+ 2023-03-25 17:27:37,169 44k INFO ====> Epoch: 140, cost 17.52 s
206
+ 2023-03-25 17:27:53,821 44k INFO ====> Epoch: 141, cost 16.65 s
207
+ 2023-03-25 17:28:10,387 44k INFO ====> Epoch: 142, cost 16.57 s
208
+ 2023-03-25 17:28:27,169 44k INFO ====> Epoch: 143, cost 16.78 s
209
+ 2023-03-25 17:28:43,398 44k INFO ====> Epoch: 144, cost 16.23 s
210
+ 2023-03-25 17:28:59,928 44k INFO ====> Epoch: 145, cost 16.53 s
211
+ 2023-03-25 17:29:09,400 44k INFO Train Epoch: 146 [45%]
212
+ 2023-03-25 17:29:09,401 44k INFO Losses: [2.016758918762207, 2.6839516162872314, 10.109888076782227, 12.951353073120117, 0.9170368909835815], step: 4800, lr: 9.820371573447515e-05
213
+ 2023-03-25 17:29:13,880 44k INFO Saving model and optimizer state at iteration 146 to ./logs/44k/G_4800.pth
214
+ 2023-03-25 17:29:15,181 44k INFO Saving model and optimizer state at iteration 146 to ./logs/44k/D_4800.pth
215
+ 2023-03-25 17:29:23,012 44k INFO ====> Epoch: 146, cost 23.08 s
216
+ 2023-03-25 17:29:39,655 44k INFO ====> Epoch: 147, cost 16.64 s
217
+ 2023-03-25 17:29:56,259 44k INFO ====> Epoch: 148, cost 16.60 s
218
+ 2023-03-25 17:30:13,013 44k INFO ====> Epoch: 149, cost 16.75 s
219
+ 2023-03-25 17:30:29,995 44k INFO ====> Epoch: 150, cost 16.98 s
220
+ 2023-03-25 17:30:46,364 44k INFO ====> Epoch: 151, cost 16.37 s
221
+ 2023-03-25 17:30:56,410 44k INFO Train Epoch: 152 [52%]
222
+ 2023-03-25 17:30:56,411 44k INFO Losses: [2.8300862312316895, 2.162816286087036, 7.240455150604248, 13.705415725708008, 0.9281750321388245], step: 5000, lr: 9.813008596033443e-05
223
+ 2023-03-25 17:31:03,427 44k INFO ====> Epoch: 152, cost 17.06 s
224
+ 2023-03-25 17:31:20,293 44k INFO ====> Epoch: 153, cost 16.87 s
225
+ 2023-03-25 17:31:36,853 44k INFO ====> Epoch: 154, cost 16.56 s
226
+ 2023-03-25 17:31:53,575 44k INFO ====> Epoch: 155, cost 16.72 s
227
+ 2023-03-25 17:32:10,656 44k INFO ====> Epoch: 156, cost 17.08 s
228
+ 2023-03-25 17:32:27,958 44k INFO ====> Epoch: 157, cost 17.30 s
229
+ 2023-03-25 17:32:39,382 44k INFO Train Epoch: 158 [58%]
230
+ 2023-03-25 17:32:39,383 44k INFO Losses: [2.3936424255371094, 2.3674676418304443, 9.325252532958984, 17.687271118164062, 0.44196417927742004], step: 5200, lr: 9.80565113912702e-05
231
+ 2023-03-25 17:32:45,648 44k INFO ====> Epoch: 158, cost 17.69 s
232
+ 2023-03-25 17:33:02,438 44k INFO ====> Epoch: 159, cost 16.79 s
233
+ 2023-03-25 17:33:19,033 44k INFO ====> Epoch: 160, cost 16.60 s
234
+ 2023-03-25 17:33:35,733 44k INFO ====> Epoch: 161, cost 16.70 s
235
+ 2023-03-25 17:33:52,506 44k INFO ====> Epoch: 162, cost 16.77 s
236
+ 2023-03-25 17:34:09,468 44k INFO ====> Epoch: 163, cost 16.96 s
237
+ 2023-03-25 17:34:21,265 44k INFO Train Epoch: 164 [64%]
238
+ 2023-03-25 17:34:21,267 44k INFO Losses: [2.301771879196167, 2.4344451427459717, 9.849881172180176, 19.910600662231445, 0.5363461971282959], step: 5400, lr: 9.798299198589162e-05
239
+ 2023-03-25 17:34:26,576 44k INFO ====> Epoch: 164, cost 17.11 s
240
+ 2023-03-25 17:34:43,835 44k INFO ====> Epoch: 165, cost 17.26 s
241
+ 2023-03-25 17:35:00,502 44k INFO ====> Epoch: 166, cost 16.67 s
242
+ 2023-03-25 17:35:17,098 44k INFO ====> Epoch: 167, cost 16.60 s
243
+ 2023-03-25 17:35:33,473 44k INFO ====> Epoch: 168, cost 16.38 s
244
+ 2023-03-25 17:35:50,479 44k INFO ====> Epoch: 169, cost 17.01 s
245
+ 2023-03-25 17:36:02,978 44k INFO Train Epoch: 170 [70%]
246
+ 2023-03-25 17:36:02,979 44k INFO Losses: [2.5066232681274414, 2.6114039421081543, 5.712583065032959, 10.28527545928955, 0.8232187628746033], step: 5600, lr: 9.790952770283884e-05
247
+ 2023-03-25 17:36:07,446 44k INFO Saving model and optimizer state at iteration 170 to ./logs/44k/G_5600.pth
248
+ 2023-03-25 17:36:08,606 44k INFO Saving model and optimizer state at iteration 170 to ./logs/44k/D_5600.pth
249
+ 2023-03-25 17:36:13,349 44k INFO ====> Epoch: 170, cost 22.87 s
250
+ 2023-03-25 17:36:30,174 44k INFO ====> Epoch: 171, cost 16.82 s
251
+ 2023-03-25 17:36:46,923 44k INFO ====> Epoch: 172, cost 16.75 s
252
+ 2023-03-25 17:37:03,606 44k INFO ====> Epoch: 173, cost 16.68 s
253
+ 2023-03-25 17:37:20,121 44k INFO ====> Epoch: 174, cost 16.51 s
254
+ 2023-03-25 17:37:36,799 44k INFO ====> Epoch: 175, cost 16.68 s
255
+ 2023-03-25 17:37:50,758 44k INFO Train Epoch: 176 [76%]
256
+ 2023-03-25 17:37:50,759 44k INFO Losses: [2.715217351913452, 2.0549123287200928, 6.569155693054199, 15.182696342468262, 0.7629079222679138], step: 5800, lr: 9.783611850078301e-05
257
+ 2023-03-25 17:37:54,542 44k INFO ====> Epoch: 176, cost 17.74 s
258
+ 2023-03-25 17:38:11,174 44k INFO ====> Epoch: 177, cost 16.63 s
259
+ 2023-03-25 17:38:28,217 44k INFO ====> Epoch: 178, cost 17.04 s
260
+ 2023-03-25 17:38:45,162 44k INFO ====> Epoch: 179, cost 16.95 s
261
+ 2023-03-25 17:39:01,783 44k INFO ====> Epoch: 180, cost 16.62 s
262
+ 2023-03-25 17:39:18,572 44k INFO ====> Epoch: 181, cost 16.79 s
263
+ 2023-03-25 17:39:32,891 44k INFO Train Epoch: 182 [82%]
264
+ 2023-03-25 17:39:32,893 44k INFO Losses: [2.1133475303649902, 2.5383996963500977, 11.810857772827148, 19.861572265625, 0.6742956042289734], step: 6000, lr: 9.776276433842631e-05
265
+ 2023-03-25 17:39:36,059 44k INFO ====> Epoch: 182, cost 17.49 s
266
+ 2023-03-25 17:39:53,021 44k INFO ====> Epoch: 183, cost 16.96 s
267
+ 2023-03-25 17:40:09,474 44k INFO ====> Epoch: 184, cost 16.45 s
268
+ 2023-03-25 17:40:25,931 44k INFO ====> Epoch: 185, cost 16.46 s
269
+ 2023-03-25 17:40:42,785 44k INFO ====> Epoch: 186, cost 16.85 s
270
+ 2023-03-25 17:40:59,892 44k INFO ====> Epoch: 187, cost 17.11 s
271
+ 2023-03-25 17:41:14,757 44k INFO Train Epoch: 188 [88%]
272
+ 2023-03-25 17:41:14,759 44k INFO Losses: [2.27668833732605, 2.766618490219116, 12.279425621032715, 20.261022567749023, 0.8307116031646729], step: 6200, lr: 9.768946517450186e-05
273
+ 2023-03-25 17:41:16,971 44k INFO ====> Epoch: 188, cost 17.08 s
274
+ 2023-03-25 17:41:34,110 44k INFO ====> Epoch: 189, cost 17.14 s
275
+ 2023-03-25 17:41:50,411 44k INFO ====> Epoch: 190, cost 16.30 s
276
+ 2023-03-25 17:42:07,037 44k INFO ====> Epoch: 191, cost 16.63 s
277
+ 2023-03-25 17:42:24,065 44k INFO ====> Epoch: 192, cost 17.03 s
278
+ 2023-03-25 17:42:40,495 44k INFO ====> Epoch: 193, cost 16.43 s
279
+ 2023-03-25 17:42:56,123 44k INFO Train Epoch: 194 [94%]
280
+ 2023-03-25 17:42:56,124 44k INFO Losses: [2.5047290325164795, 2.0337984561920166, 10.720749855041504, 17.247852325439453, 0.6880492568016052], step: 6400, lr: 9.761622096777372e-05
281
+ 2023-03-25 17:43:00,648 44k INFO Saving model and optimizer state at iteration 194 to ./logs/44k/G_6400.pth
282
+ 2023-03-25 17:43:01,842 44k INFO Saving model and optimizer state at iteration 194 to ./logs/44k/D_6400.pth
283
+ 2023-03-25 17:43:03,376 44k INFO ====> Epoch: 194, cost 22.88 s
284
+ 2023-03-25 17:43:19,858 44k INFO ====> Epoch: 195, cost 16.48 s
285
+ 2023-03-25 17:43:36,398 44k INFO ====> Epoch: 196, cost 16.54 s
286
+ 2023-03-25 17:43:53,393 44k INFO ====> Epoch: 197, cost 16.99 s
287
+ 2023-03-25 17:44:10,373 44k INFO ====> Epoch: 198, cost 16.98 s
288
+ 2023-03-25 17:44:27,149 44k INFO ====> Epoch: 199, cost 16.78 s
289
+ 2023-03-25 17:44:43,969 44k INFO ====> Epoch: 200, cost 16.82 s
290
+ 2023-03-25 17:44:47,513 44k INFO Train Epoch: 201 [0%]
291
+ 2023-03-25 17:44:47,514 44k INFO Losses: [2.6873631477355957, 2.1986629962921143, 6.34331750869751, 10.858728408813477, 0.6867925524711609], step: 6600, lr: 9.753083879807726e-05
292
+ 2023-03-25 17:45:01,125 44k INFO ====> Epoch: 201, cost 17.16 s
293
+ 2023-03-25 17:45:17,892 44k INFO ====> Epoch: 202, cost 16.77 s
294
+ 2023-03-25 17:45:34,752 44k INFO ====> Epoch: 203, cost 16.86 s
295
+ 2023-03-25 17:45:51,161 44k INFO ====> Epoch: 204, cost 16.41 s
296
+ 2023-03-25 17:46:08,008 44k INFO ====> Epoch: 205, cost 16.85 s
297
+ 2023-03-25 17:46:25,001 44k INFO ====> Epoch: 206, cost 16.99 s
298
+ 2023-03-25 17:46:29,316 44k INFO Train Epoch: 207 [6%]
299
+ 2023-03-25 17:46:29,317 44k INFO Losses: [2.3735992908477783, 2.3111236095428467, 6.5680999755859375, 11.549568176269531, 0.12051205337047577], step: 6800, lr: 9.745771352395957e-05
300
+ 2023-03-25 17:46:42,105 44k INFO ====> Epoch: 207, cost 17.10 s
301
+ 2023-03-25 17:46:58,695 44k INFO ====> Epoch: 208, cost 16.59 s
302
+ 2023-03-25 17:47:15,241 44k INFO ====> Epoch: 209, cost 16.55 s
303
+ 2023-03-25 17:47:31,924 44k INFO ====> Epoch: 210, cost 16.68 s
304
+ 2023-03-25 17:47:48,389 44k INFO ====> Epoch: 211, cost 16.47 s
305
+ 2023-03-25 17:48:05,112 44k INFO ====> Epoch: 212, cost 16.72 s
306
+ 2023-03-25 17:48:10,310 44k INFO Train Epoch: 213 [12%]
307
+ 2023-03-25 17:48:10,311 44k INFO Losses: [2.338576316833496, 2.1111223697662354, 9.061708450317383, 16.787063598632812, 0.529293954372406], step: 7000, lr: 9.73846430766616e-05
308
+ 2023-03-25 17:48:22,335 44k INFO ====> Epoch: 213, cost 17.22 s
309
+ 2023-03-25 17:48:38,741 44k INFO ====> Epoch: 214, cost 16.41 s
310
+ 2023-03-25 17:48:55,306 44k INFO ====> Epoch: 215, cost 16.56 s
311
+ 2023-03-25 17:49:12,085 44k INFO ====> Epoch: 216, cost 16.78 s
312
+ 2023-03-25 17:49:28,552 44k INFO ====> Epoch: 217, cost 16.47 s
313
+ 2023-03-25 17:49:44,519 44k INFO ====> Epoch: 218, cost 15.97 s
314
+ 2023-03-25 17:49:50,073 44k INFO Train Epoch: 219 [18%]
315
+ 2023-03-25 17:49:50,074 44k INFO Losses: [2.538870334625244, 1.9039925336837769, 8.920069694519043, 13.447583198547363, 0.2965475022792816], step: 7200, lr: 9.731162741507607e-05
316
+ 2023-03-25 17:49:54,374 44k INFO Saving model and optimizer state at iteration 219 to ./logs/44k/G_7200.pth
317
+ 2023-03-25 17:49:55,499 44k INFO Saving model and optimizer state at iteration 219 to ./logs/44k/D_7200.pth
318
+ 2023-03-25 17:50:06,295 44k INFO ====> Epoch: 219, cost 21.78 s
319
+ 2023-03-25 17:50:21,994 44k INFO ====> Epoch: 220, cost 15.70 s
320
+ 2023-03-25 17:50:37,662 44k INFO ====> Epoch: 221, cost 15.67 s
321
+ 2023-03-25 17:50:54,290 44k INFO ====> Epoch: 222, cost 16.63 s
322
+ 2023-03-25 17:51:10,823 44k INFO ====> Epoch: 223, cost 16.53 s
323
+ 2023-03-25 17:51:27,627 44k INFO ====> Epoch: 224, cost 16.80 s
324
+ 2023-03-25 17:51:33,887 44k INFO Train Epoch: 225 [24%]
325
+ 2023-03-25 17:51:33,888 44k INFO Losses: [2.269322395324707, 2.707637310028076, 11.994312286376953, 19.770584106445312, 0.7525858283042908], step: 7400, lr: 9.723866649812655e-05
326
+ 2023-03-25 17:51:44,009 44k INFO ====> Epoch: 225, cost 16.38 s
327
+ 2023-03-25 17:52:00,202 44k INFO ====> Epoch: 226, cost 16.19 s
328
+ 2023-03-25 17:52:16,282 44k INFO ====> Epoch: 227, cost 16.08 s
329
+ 2023-03-25 17:52:32,042 44k INFO ====> Epoch: 228, cost 15.76 s
330
+ 2023-03-25 17:52:47,741 44k INFO ====> Epoch: 229, cost 15.70 s
331
+ 2023-03-25 17:53:03,442 44k INFO ====> Epoch: 230, cost 15.70 s
332
+ 2023-03-25 17:53:10,435 44k INFO Train Epoch: 231 [30%]
333
+ 2023-03-25 17:53:10,436 44k INFO Losses: [2.1911444664001465, 2.909498691558838, 12.862323760986328, 18.82267951965332, 0.7391169667243958], step: 7600, lr: 9.716576028476738e-05
334
+ 2023-03-25 17:53:19,687 44k INFO ====> Epoch: 231, cost 16.25 s
335
+ 2023-03-25 17:53:37,561 44k INFO ====> Epoch: 232, cost 17.87 s
336
+ 2023-03-25 17:53:53,361 44k INFO ====> Epoch: 233, cost 15.80 s
337
+ 2023-03-25 17:54:11,201 44k INFO ====> Epoch: 234, cost 17.84 s
338
+ 2023-03-25 17:54:26,804 44k INFO ====> Epoch: 235, cost 15.60 s
339
+ 2023-03-25 17:54:42,860 44k INFO ====> Epoch: 236, cost 16.06 s
340
+ 2023-03-25 17:54:50,779 44k INFO Train Epoch: 237 [36%]
341
+ 2023-03-25 17:54:50,780 44k INFO Losses: [2.5400569438934326, 2.370675802230835, 7.975701332092285, 14.911909103393555, 0.5070781111717224], step: 7800, lr: 9.709290873398365e-05
342
+ 2023-03-25 17:54:59,921 44k INFO ====> Epoch: 237, cost 17.06 s
343
+ 2023-03-25 17:55:15,649 44k INFO ====> Epoch: 238, cost 15.73 s
344
+ 2023-03-25 17:55:31,290 44k INFO ====> Epoch: 239, cost 15.64 s
345
+ 2023-03-25 17:55:47,042 44k INFO ====> Epoch: 240, cost 15.75 s
346
+ 2023-03-25 17:56:02,659 44k INFO ====> Epoch: 241, cost 15.62 s
347
+ 2023-03-25 17:56:18,393 44k INFO ====> Epoch: 242, cost 15.73 s
348
+ 2023-03-25 17:56:26,769 44k INFO Train Epoch: 243 [42%]
349
+ 2023-03-25 17:56:26,770 44k INFO Losses: [2.5333597660064697, 2.1053004264831543, 8.916841506958008, 15.705123901367188, 0.7188340425491333], step: 8000, lr: 9.702011180479129e-05
350
+ 2023-03-25 17:56:30,936 44k INFO Saving model and optimizer state at iteration 243 to ./logs/44k/G_8000.pth
351
+ 2023-03-25 17:56:32,159 44k INFO Saving model and optimizer state at iteration 243 to ./logs/44k/D_8000.pth
352
+ 2023-03-25 17:56:39,973 44k INFO ====> Epoch: 243, cost 21.58 s
353
+ 2023-03-25 17:56:55,890 44k INFO ====> Epoch: 244, cost 15.92 s
354
+ 2023-03-25 17:57:11,491 44k INFO ====> Epoch: 245, cost 15.60 s
355
+ 2023-03-25 17:57:27,164 44k INFO ====> Epoch: 246, cost 15.67 s
356
+ 2023-03-25 17:57:43,077 44k INFO ====> Epoch: 247, cost 15.91 s
357
+ 2023-03-25 17:57:59,349 44k INFO ====> Epoch: 248, cost 16.27 s
358
+ 2023-03-25 17:58:08,985 44k INFO Train Epoch: 249 [48%]
359
+ 2023-03-25 17:58:08,987 44k INFO Losses: [2.553239345550537, 1.8566803932189941, 6.609829902648926, 16.236770629882812, 0.5468897819519043], step: 8200, lr: 9.694736945623688e-05
360
+ 2023-03-25 17:58:16,350 44k INFO ====> Epoch: 249, cost 17.00 s
361
+ 2023-03-25 17:58:32,729 44k INFO ====> Epoch: 250, cost 16.38 s
362
+ 2023-03-25 17:58:49,065 44k INFO ====> Epoch: 251, cost 16.34 s
363
+ 2023-03-25 17:59:05,556 44k INFO ====> Epoch: 252, cost 16.49 s
364
+ 2023-03-25 17:59:21,096 44k INFO ====> Epoch: 253, cost 15.54 s
365
+ 2023-03-25 17:59:36,732 44k INFO ====> Epoch: 254, cost 15.64 s
366
+ 2023-03-25 17:59:46,629 44k INFO Train Epoch: 255 [55%]
367
+ 2023-03-25 17:59:46,630 44k INFO Losses: [2.565453290939331, 1.9224674701690674, 7.469606876373291, 16.19024658203125, 0.5049387812614441], step: 8400, lr: 9.687468164739773e-05
368
+ 2023-03-25 17:59:53,023 44k INFO ====> Epoch: 255, cost 16.29 s
369
+ 2023-03-25 18:00:08,717 44k INFO ====> Epoch: 256, cost 15.69 s
370
+ 2023-03-25 18:00:24,316 44k INFO ====> Epoch: 257, cost 15.60 s
371
+ 2023-03-25 18:00:40,441 44k INFO ====> Epoch: 258, cost 16.13 s
372
+ 2023-03-25 18:00:57,101 44k INFO ====> Epoch: 259, cost 16.66 s
373
+ 2023-03-25 18:01:13,862 44k INFO ====> Epoch: 260, cost 16.76 s
374
+ 2023-03-25 18:01:25,107 44k INFO Train Epoch: 261 [61%]
375
+ 2023-03-25 18:01:25,109 44k INFO Losses: [2.4458837509155273, 2.323704719543457, 10.821569442749023, 17.358577728271484, 0.8134444952011108], step: 8600, lr: 9.680204833738185e-05
376
+ 2023-03-25 18:01:30,953 44k INFO ====> Epoch: 261, cost 17.09 s
377
+ 2023-03-25 18:01:47,414 44k INFO ====> Epoch: 262, cost 16.46 s
378
+ 2023-03-25 18:02:03,338 44k INFO ====> Epoch: 263, cost 15.92 s
379
+ 2023-03-25 18:02:21,478 44k INFO ====> Epoch: 264, cost 18.14 s
380
+ 2023-03-25 18:02:37,738 44k INFO ====> Epoch: 265, cost 16.26 s
381
+ 2023-03-25 18:02:54,410 44k INFO ====> Epoch: 266, cost 16.67 s
382
+ 2023-03-25 18:03:06,556 44k INFO Train Epoch: 267 [67%]
383
+ 2023-03-25 18:03:06,557 44k INFO Losses: [2.6201024055480957, 2.173233985900879, 5.443298816680908, 13.764752388000488, 0.6760918498039246], step: 8800, lr: 9.67294694853279e-05
384
+ 2023-03-25 18:03:11,333 44k INFO Saving model and optimizer state at iteration 267 to ./logs/44k/G_8800.pth
385
+ 2023-03-25 18:03:12,666 44k INFO Saving model and optimizer state at iteration 267 to ./logs/44k/D_8800.pth
386
+ 2023-03-25 18:03:17,845 44k INFO ====> Epoch: 267, cost 23.43 s
387
+ 2023-03-25 18:03:34,712 44k INFO ====> Epoch: 268, cost 16.87 s
388
+ 2023-03-25 18:03:51,312 44k INFO ====> Epoch: 269, cost 16.60 s
389
+ 2023-03-25 18:04:07,990 44k INFO ====> Epoch: 270, cost 16.68 s
390
+ 2023-03-25 18:04:24,400 44k INFO ====> Epoch: 271, cost 16.41 s
391
+ 2023-03-25 18:04:41,096 44k INFO ====> Epoch: 272, cost 16.70 s
392
+ 2023-03-25 18:04:54,136 44k INFO Train Epoch: 273 [73%]
393
+ 2023-03-25 18:04:54,136 44k INFO Losses: [2.3441848754882812, 2.227895736694336, 10.538599967956543, 14.83042049407959, 0.6608119010925293], step: 9000, lr: 9.665694505040515e-05
394
+ 2023-03-25 18:04:58,449 44k INFO ====> Epoch: 273, cost 17.35 s
395
+ 2023-03-25 18:05:14,970 44k INFO ====> Epoch: 274, cost 16.52 s
396
+ 2023-03-25 18:05:31,373 44k INFO ====> Epoch: 275, cost 16.40 s
397
+ 2023-03-25 18:05:47,843 44k INFO ====> Epoch: 276, cost 16.47 s
398
+ 2023-03-25 18:06:04,185 44k INFO ====> Epoch: 277, cost 16.34 s
399
+ 2023-03-25 18:06:20,638 44k INFO ====> Epoch: 278, cost 16.45 s
400
+ 2023-03-25 18:06:34,148 44k INFO Train Epoch: 279 [79%]
401
+ 2023-03-25 18:06:34,149 44k INFO Losses: [2.3495030403137207, 2.2592837810516357, 14.851377487182617, 18.4000244140625, 0.6196873188018799], step: 9200, lr: 9.658447499181352e-05
402
+ 2023-03-25 18:06:37,606 44k INFO ====> Epoch: 279, cost 16.97 s
403
+ 2023-03-25 18:06:54,350 44k INFO ====> Epoch: 280, cost 16.74 s
404
+ 2023-03-25 18:07:11,093 44k INFO ====> Epoch: 281, cost 16.74 s
405
+ 2023-03-25 18:07:27,706 44k INFO ====> Epoch: 282, cost 16.61 s
406
+ 2023-03-25 18:07:44,093 44k INFO ====> Epoch: 283, cost 16.39 s
407
+ 2023-03-25 18:08:00,560 44k INFO ====> Epoch: 284, cost 16.47 s
408
+ 2023-03-25 18:08:14,883 44k INFO Train Epoch: 285 [85%]
409
+ 2023-03-25 18:08:14,884 44k INFO Losses: [2.4534430503845215, 2.458757162094116, 7.7290263175964355, 14.926005363464355, 0.39300334453582764], step: 9400, lr: 9.651205926878348e-05
410
+ 2023-03-25 18:08:17,638 44k INFO ====> Epoch: 285, cost 17.08 s
411
+ 2023-03-25 18:08:34,163 44k INFO ====> Epoch: 286, cost 16.52 s
412
+ 2023-03-25 18:08:50,591 44k INFO ====> Epoch: 287, cost 16.43 s
413
+ 2023-03-25 18:09:07,061 44k INFO ====> Epoch: 288, cost 16.47 s
414
+ 2023-03-25 18:09:23,434 44k INFO ====> Epoch: 289, cost 16.37 s
415
+ 2023-03-25 18:09:39,824 44k INFO ====> Epoch: 290, cost 16.39 s
416
+ 2023-03-25 18:09:54,922 44k INFO Train Epoch: 291 [91%]
417
+ 2023-03-25 18:09:54,923 44k INFO Losses: [2.393284797668457, 2.443169116973877, 8.165253639221191, 16.708036422729492, 0.9939846396446228], step: 9600, lr: 9.643969784057613e-05
418
+ 2023-03-25 18:09:59,308 44k INFO Saving model and optimizer state at iteration 291 to ./logs/44k/G_9600.pth
419
+ 2023-03-25 18:10:00,733 44k INFO Saving model and optimizer state at iteration 291 to ./logs/44k/D_9600.pth
420
+ 2023-03-25 18:10:02,769 44k INFO ====> Epoch: 291, cost 22.95 s
421
+ 2023-03-25 18:10:19,195 44k INFO ====> Epoch: 292, cost 16.43 s
422
+ 2023-03-25 18:10:35,513 44k INFO ====> Epoch: 293, cost 16.32 s
423
+ 2023-03-25 18:10:52,152 44k INFO ====> Epoch: 294, cost 16.64 s
424
+ 2023-03-25 18:11:08,493 44k INFO ====> Epoch: 295, cost 16.34 s
425
+ 2023-03-25 18:11:24,789 44k INFO ====> Epoch: 296, cost 16.30 s
426
+ 2023-03-25 18:11:41,201 44k INFO Train Epoch: 297 [97%]
427
+ 2023-03-25 18:11:41,203 44k INFO Losses: [2.896554708480835, 2.316603660583496, 5.762091159820557, 10.353036880493164, 0.30002638697624207], step: 9800, lr: 9.636739066648303e-05
428
+ 2023-03-25 18:11:42,365 44k INFO ====> Epoch: 297, cost 17.58 s
429
+ 2023-03-25 18:11:58,961 44k INFO ====> Epoch: 298, cost 16.60 s
430
+ 2023-03-25 18:12:15,138 44k INFO ====> Epoch: 299, cost 16.18 s
431
+ 2023-03-25 18:12:31,489 44k INFO ====> Epoch: 300, cost 16.35 s
432
+ 2023-03-25 18:12:47,799 44k INFO ====> Epoch: 301, cost 16.31 s
433
+ 2023-03-25 18:13:03,973 44k INFO ====> Epoch: 302, cost 16.17 s
434
+ 2023-03-25 18:13:20,685 44k INFO ====> Epoch: 303, cost 16.71 s
435
+ 2023-03-25 18:13:24,919 44k INFO Train Epoch: 304 [3%]
436
+ 2023-03-25 18:13:24,920 44k INFO Losses: [2.515075922012329, 2.2078559398651123, 9.70125675201416, 15.418092727661133, 0.8266330361366272], step: 10000, lr: 9.628310081361311e-05
437
+ 2023-03-25 18:13:38,068 44k INFO ====> Epoch: 304, cost 17.38 s
438
+ 2023-03-25 18:13:54,461 44k INFO ====> Epoch: 305, cost 16.39 s
439
+ 2023-03-25 18:14:11,062 44k INFO ====> Epoch: 306, cost 16.60 s
440
+ 2023-03-25 18:14:27,536 44k INFO ====> Epoch: 307, cost 16.47 s
441
+ 2023-03-25 18:14:43,967 44k INFO ====> Epoch: 308, cost 16.43 s
442
+ 2023-03-25 18:15:00,505 44k INFO ====> Epoch: 309, cost 16.54 s
443
+ 2023-03-25 18:15:05,371 44k INFO Train Epoch: 310 [9%]
444
+ 2023-03-25 18:15:05,372 44k INFO Losses: [2.3045690059661865, 2.395820140838623, 7.718233108520508, 16.340299606323242, 0.6959915161132812], step: 10200, lr: 9.621091105059392e-05
445
+ 2023-03-25 18:15:17,643 44k INFO ====> Epoch: 310, cost 17.14 s
446
+ 2023-03-25 18:15:34,607 44k INFO ====> Epoch: 311, cost 16.96 s
447
+ 2023-03-25 18:15:50,909 44k INFO ====> Epoch: 312, cost 16.30 s
448
+ 2023-03-25 18:16:07,296 44k INFO ====> Epoch: 313, cost 16.39 s
449
+ 2023-03-25 18:16:23,498 44k INFO ====> Epoch: 314, cost 16.20 s
450
+ 2023-03-25 18:16:39,983 44k INFO ====> Epoch: 315, cost 16.48 s
451
+ 2023-03-25 18:16:45,429 44k INFO Train Epoch: 316 [15%]
452
+ 2023-03-25 18:16:45,430 44k INFO Losses: [2.462087392807007, 2.2239341735839844, 9.908827781677246, 15.508355140686035, 0.837641716003418], step: 10400, lr: 9.613877541298036e-05
453
+ 2023-03-25 18:16:49,915 44k INFO Saving model and optimizer state at iteration 316 to ./logs/44k/G_10400.pth
454
+ 2023-03-25 18:16:51,219 44k INFO Saving model and optimizer state at iteration 316 to ./logs/44k/D_10400.pth
455
+ 2023-03-25 18:17:02,702 44k INFO ====> Epoch: 316, cost 22.72 s
456
+ 2023-03-25 18:17:18,983 44k INFO ====> Epoch: 317, cost 16.28 s
457
+ 2023-03-25 18:17:35,311 44k INFO ====> Epoch: 318, cost 16.33 s
458
+ 2023-03-25 18:17:51,528 44k INFO ====> Epoch: 319, cost 16.22 s
459
+ 2023-03-25 18:18:07,950 44k INFO ====> Epoch: 320, cost 16.42 s
460
+ 2023-03-25 18:18:24,626 44k INFO ====> Epoch: 321, cost 16.68 s
461
+ 2023-03-25 18:18:30,769 44k INFO Train Epoch: 322 [21%]
462
+ 2023-03-25 18:18:30,770 44k INFO Losses: [2.3425683975219727, 2.628889322280884, 12.140241622924805, 20.59270477294922, 0.054636288434267044], step: 10600, lr: 9.606669386019102e-05
463
+ 2023-03-25 18:18:41,415 44k INFO ====> Epoch: 322, cost 16.79 s
464
+ 2023-03-25 18:18:57,759 44k INFO ====> Epoch: 323, cost 16.34 s
465
+ 2023-03-25 18:19:14,078 44k INFO ====> Epoch: 324, cost 16.32 s
466
+ 2023-03-25 18:19:30,347 44k INFO ====> Epoch: 325, cost 16.27 s
467
+ 2023-03-25 18:19:47,250 44k INFO ====> Epoch: 326, cost 16.90 s
468
+ 2023-03-25 18:20:03,686 44k INFO ====> Epoch: 327, cost 16.44 s
469
+ 2023-03-25 18:20:10,686 44k INFO Train Epoch: 328 [27%]
470
+ 2023-03-25 18:20:10,687 44k INFO Losses: [2.4458670616149902, 2.2881720066070557, 10.13757610321045, 14.627115249633789, 0.47115644812583923], step: 10800, lr: 9.599466635167497e-05
471
+ 2023-03-25 18:20:20,765 44k INFO ====> Epoch: 328, cost 17.08 s
472
+ 2023-03-25 18:20:37,407 44k INFO ====> Epoch: 329, cost 16.64 s
473
+ 2023-03-25 18:20:53,950 44k INFO ====> Epoch: 330, cost 16.54 s
474
+ 2023-03-25 18:21:10,322 44k INFO ====> Epoch: 331, cost 16.37 s
475
+ 2023-03-25 18:21:26,633 44k INFO ====> Epoch: 332, cost 16.31 s
476
+ 2023-03-25 18:21:42,887 44k INFO ====> Epoch: 333, cost 16.25 s
477
+ 2023-03-25 18:21:50,532 44k INFO Train Epoch: 334 [33%]
478
+ 2023-03-25 18:21:50,533 44k INFO Losses: [2.60351824760437, 2.2693228721618652, 7.4690985679626465, 13.879813194274902, 0.5584508776664734], step: 11000, lr: 9.592269284691169e-05
479
+ 2023-03-25 18:22:00,284 44k INFO ====> Epoch: 334, cost 17.40 s
480
+ 2023-03-25 18:22:16,678 44k INFO ====> Epoch: 335, cost 16.39 s
481
+ 2023-03-25 18:22:33,068 44k INFO ====> Epoch: 336, cost 16.39 s
482
+ 2023-03-25 18:22:49,695 44k INFO ====> Epoch: 337, cost 16.63 s
483
+ 2023-03-25 18:23:06,120 44k INFO ====> Epoch: 338, cost 16.42 s
484
+ 2023-03-25 18:23:22,651 44k INFO ====> Epoch: 339, cost 16.53 s
485
+ 2023-03-25 18:23:31,178 44k INFO Train Epoch: 340 [39%]
486
+ 2023-03-25 18:23:31,179 44k INFO Losses: [2.6228575706481934, 2.4648935794830322, 8.234376907348633, 15.624024391174316, 0.521338939666748], step: 11200, lr: 9.5850773305411e-05
487
+ 2023-03-25 18:23:35,801 44k INFO Saving model and optimizer state at iteration 340 to ./logs/44k/G_11200.pth
488
+ 2023-03-25 18:23:37,093 44k INFO Saving model and optimizer state at iteration 340 to ./logs/44k/D_11200.pth
489
+ 2023-03-25 18:23:45,697 44k INFO ====> Epoch: 340, cost 23.05 s
490
+ 2023-03-25 18:24:01,807 44k INFO ====> Epoch: 341, cost 16.11 s
491
+ 2023-03-25 18:24:18,320 44k INFO ====> Epoch: 342, cost 16.51 s
492
+ 2023-03-25 18:24:34,861 44k INFO ====> Epoch: 343, cost 16.54 s
493
+ 2023-03-25 18:24:51,178 44k INFO ====> Epoch: 344, cost 16.32 s
494
+ 2023-03-25 18:25:07,875 44k INFO ====> Epoch: 345, cost 16.70 s
495
+ 2023-03-25 18:25:16,986 44k INFO Train Epoch: 346 [45%]
496
+ 2023-03-25 18:25:16,987 44k INFO Losses: [2.8093314170837402, 2.031580686569214, 7.040884971618652, 13.01242446899414, 0.9223231673240662], step: 11400, lr: 9.577890768671308e-05
497
+ 2023-03-25 18:25:25,186 44k INFO ====> Epoch: 346, cost 17.31 s
498
+ 2023-03-25 18:25:41,721 44k INFO ====> Epoch: 347, cost 16.53 s
499
+ 2023-03-25 18:25:57,898 44k INFO ====> Epoch: 348, cost 16.18 s
500
+ 2023-03-25 18:26:14,619 44k INFO ====> Epoch: 349, cost 16.72 s
501
+ 2023-03-25 18:26:31,532 44k INFO ====> Epoch: 350, cost 16.91 s
502
+ 2023-03-25 18:26:47,928 44k INFO ====> Epoch: 351, cost 16.40 s
503
+ 2023-03-25 18:26:57,859 44k INFO Train Epoch: 352 [52%]
504
+ 2023-03-25 18:26:57,861 44k INFO Losses: [2.3367810249328613, 2.234591484069824, 11.762563705444336, 19.299331665039062, 0.5701344013214111], step: 11600, lr: 9.570709595038851e-05
505
+ 2023-03-25 18:27:04,732 44k INFO ====> Epoch: 352, cost 16.80 s
506
+ 2023-03-25 18:27:21,403 44k INFO ====> Epoch: 353, cost 16.67 s
507
+ 2023-03-25 18:27:37,737 44k INFO ====> Epoch: 354, cost 16.33 s
508
+ 2023-03-25 18:27:54,743 44k INFO ====> Epoch: 355, cost 17.01 s
509
+ 2023-03-25 18:28:11,397 44k INFO ====> Epoch: 356, cost 16.65 s
510
+ 2023-03-25 18:28:27,983 44k INFO ====> Epoch: 357, cost 16.59 s
511
+ 2023-03-25 18:28:38,827 44k INFO Train Epoch: 358 [58%]
512
+ 2023-03-25 18:28:38,828 44k INFO Losses: [2.4830808639526367, 2.3789222240448, 9.961544036865234, 20.941343307495117, 0.4926968812942505], step: 11800, lr: 9.56353380560381e-05
513
+ 2023-03-25 18:28:45,245 44k INFO ====> Epoch: 358, cost 17.26 s
514
+ 2023-03-25 18:29:01,674 44k INFO ====> Epoch: 359, cost 16.43 s
515
+ 2023-03-25 18:29:18,150 44k INFO ====> Epoch: 360, cost 16.48 s
516
+ 2023-03-25 18:29:34,541 44k INFO ====> Epoch: 361, cost 16.39 s
517
+ 2023-03-25 18:29:51,314 44k INFO ====> Epoch: 362, cost 16.77 s
518
+ 2023-03-25 18:30:07,952 44k INFO ====> Epoch: 363, cost 16.64 s
519
+ 2023-03-25 18:30:19,472 44k INFO Train Epoch: 364 [64%]
520
+ 2023-03-25 18:30:19,473 44k INFO Losses: [2.204318046569824, 2.510396957397461, 13.009056091308594, 19.046606063842773, 0.6128262281417847], step: 12000, lr: 9.556363396329299e-05
521
+ 2023-03-25 18:30:23,964 44k INFO Saving model and optimizer state at iteration 364 to ./logs/44k/G_12000.pth
522
+ 2023-03-25 18:30:25,298 44k INFO Saving model and optimizer state at iteration 364 to ./logs/44k/D_12000.pth
523
+ 2023-03-25 18:30:30,855 44k INFO ====> Epoch: 364, cost 22.90 s
524
+ 2023-03-25 18:30:47,333 44k INFO ====> Epoch: 365, cost 16.48 s
525
+ 2023-03-25 18:31:03,833 44k INFO ====> Epoch: 366, cost 16.50 s
526
+ 2023-03-25 18:31:20,318 44k INFO ====> Epoch: 367, cost 16.49 s
527
+ 2023-03-25 18:31:36,481 44k INFO ====> Epoch: 368, cost 16.16 s
528
+ 2023-03-25 18:31:52,959 44k INFO ====> Epoch: 369, cost 16.48 s
529
+ 2023-03-25 18:32:05,768 44k INFO Train Epoch: 370 [70%]
530
+ 2023-03-25 18:32:05,769 44k INFO Losses: [2.8103220462799072, 1.9135441780090332, 4.667817115783691, 12.864496231079102, 0.5750285983085632], step: 12200, lr: 9.54919836318146e-05
531
+ 2023-03-25 18:32:10,376 44k INFO ====> Epoch: 370, cost 17.42 s
532
+ 2023-03-25 18:32:26,649 44k INFO ====> Epoch: 371, cost 16.27 s
533
+ 2023-03-25 18:32:43,415 44k INFO ====> Epoch: 372, cost 16.77 s
534
+ 2023-03-25 18:32:59,729 44k INFO ====> Epoch: 373, cost 16.31 s
535
+ 2023-03-25 18:33:16,345 44k INFO ====> Epoch: 374, cost 16.62 s
536
+ 2023-03-25 18:33:32,600 44k INFO ====> Epoch: 375, cost 16.26 s
537
+ 2023-03-25 18:33:45,666 44k INFO Train Epoch: 376 [76%]
538
+ 2023-03-25 18:33:45,667 44k INFO Losses: [2.4472713470458984, 2.2148959636688232, 10.443349838256836, 16.083932876586914, 0.6222093105316162], step: 12400, lr: 9.542038702129457e-05
539
+ 2023-03-25 18:33:49,459 44k INFO ====> Epoch: 376, cost 16.86 s
540
+ 2023-03-25 18:34:05,825 44k INFO ====> Epoch: 377, cost 16.37 s
541
+ 2023-03-25 18:34:22,051 44k INFO ====> Epoch: 378, cost 16.23 s
542
+ 2023-03-25 18:34:38,473 44k INFO ====> Epoch: 379, cost 16.42 s
543
+ 2023-03-25 18:34:54,719 44k INFO ====> Epoch: 380, cost 16.25 s
544
+ 2023-03-25 18:35:10,879 44k INFO ====> Epoch: 381, cost 16.16 s
545
+ 2023-03-25 18:35:24,611 44k INFO Train Epoch: 382 [82%]
546
+ 2023-03-25 18:35:24,612 44k INFO Losses: [2.351381778717041, 2.2640938758850098, 10.070740699768066, 19.411256790161133, 0.5809462666511536], step: 12600, lr: 9.534884409145477e-05
547
+ 2023-03-25 18:35:27,619 44k INFO ====> Epoch: 382, cost 16.74 s
548
+ 2023-03-25 18:35:44,312 44k INFO ====> Epoch: 383, cost 16.69 s
549
+ 2023-03-25 18:36:00,575 44k INFO ====> Epoch: 384, cost 16.26 s
550
+ 2023-03-25 18:36:17,367 44k INFO ====> Epoch: 385, cost 16.79 s
551
+ 2023-03-25 18:36:33,862 44k INFO ====> Epoch: 386, cost 16.49 s
552
+ 2023-03-25 18:36:50,147 44k INFO ====> Epoch: 387, cost 16.28 s
553
+ 2023-03-25 18:37:04,689 44k INFO Train Epoch: 388 [88%]
554
+ 2023-03-25 18:37:04,690 44k INFO Losses: [2.483032464981079, 2.0824134349823, 9.235447883605957, 15.548537254333496, 0.5608479976654053], step: 12800, lr: 9.527735480204728e-05
555
+ 2023-03-25 18:37:09,240 44k INFO Saving model and optimizer state at iteration 388 to ./logs/44k/G_12800.pth
556
+ 2023-03-25 18:37:10,581 44k INFO Saving model and optimizer state at iteration 388 to ./logs/44k/D_12800.pth
557
+ 2023-03-25 18:37:12,943 44k INFO ====> Epoch: 388, cost 22.80 s
558
+ 2023-03-25 18:37:29,278 44k INFO ====> Epoch: 389, cost 16.33 s
559
+ 2023-03-25 18:37:45,699 44k INFO ====> Epoch: 390, cost 16.42 s
560
+ 2023-03-25 18:38:02,084 44k INFO ====> Epoch: 391, cost 16.39 s
561
+ 2023-03-25 18:38:18,298 44k INFO ====> Epoch: 392, cost 16.21 s
562
+ 2023-03-25 18:38:34,621 44k INFO ====> Epoch: 393, cost 16.32 s
563
+ 2023-03-25 18:38:50,007 44k INFO Train Epoch: 394 [94%]
564
+ 2023-03-25 18:38:50,008 44k INFO Losses: [2.5371432304382324, 1.9721053838729858, 9.028430938720703, 12.427369117736816, 0.916906476020813], step: 13000, lr: 9.520591911285433e-05
565
+ 2023-03-25 18:38:51,490 44k INFO ====> Epoch: 394, cost 16.87 s
566
+ 2023-03-25 18:39:08,072 44k INFO ====> Epoch: 395, cost 16.58 s
567
+ 2023-03-25 18:39:24,382 44k INFO ====> Epoch: 396, cost 16.31 s
568
+ 2023-03-25 18:39:41,864 44k INFO ====> Epoch: 397, cost 17.48 s
569
+ 2023-03-25 18:39:58,320 44k INFO ====> Epoch: 398, cost 16.46 s
570
+ 2023-03-25 18:40:14,427 44k INFO ====> Epoch: 399, cost 16.11 s
571
+ 2023-03-25 18:40:31,773 44k INFO ====> Epoch: 400, cost 17.35 s
572
+ 2023-03-25 18:40:35,493 44k INFO Train Epoch: 401 [0%]
573
+ 2023-03-25 18:40:35,495 44k INFO Losses: [2.464968204498291, 2.3817968368530273, 5.272903919219971, 11.98283863067627, 0.8528004288673401], step: 13200, lr: 9.512264516656537e-05
574
+ 2023-03-25 18:40:48,882 44k INFO ====> Epoch: 401, cost 17.11 s
575
+ 2023-03-25 18:41:05,273 44k INFO ====> Epoch: 402, cost 16.39 s
576
+ 2023-03-25 18:41:21,778 44k INFO ====> Epoch: 403, cost 16.51 s
577
+ 2023-03-25 18:41:38,055 44k INFO ====> Epoch: 404, cost 16.28 s
578
+ 2023-03-25 18:41:54,305 44k INFO ====> Epoch: 405, cost 16.25 s
579
+ 2023-03-25 18:42:11,211 44k INFO ====> Epoch: 406, cost 16.91 s
580
+ 2023-03-25 18:42:15,794 44k INFO Train Epoch: 407 [6%]
581
+ 2023-03-25 18:42:15,795 44k INFO Losses: [2.748605728149414, 2.0154476165771484, 8.186812400817871, 10.141654014587402, -0.014795398339629173], step: 13400, lr: 9.505132547334502e-05
582
+ 2023-03-25 18:42:28,420 44k INFO ====> Epoch: 407, cost 17.21 s
583
+ 2023-03-25 18:42:45,205 44k INFO ====> Epoch: 408, cost 16.79 s
584
+ 2023-03-25 18:43:01,444 44k INFO ====> Epoch: 409, cost 16.24 s
585
+ 2023-03-25 18:43:17,926 44k INFO ====> Epoch: 410, cost 16.48 s
586
+ 2023-03-25 18:43:34,340 44k INFO ====> Epoch: 411, cost 16.41 s
587
+ 2023-03-25 18:43:50,913 44k INFO ====> Epoch: 412, cost 16.57 s
588
+ 2023-03-25 18:43:56,056 44k INFO Train Epoch: 413 [12%]
589
+ 2023-03-25 18:43:56,057 44k INFO Losses: [2.4909560680389404, 2.0936198234558105, 6.849160671234131, 12.217107772827148, 0.795183539390564], step: 13600, lr: 9.498005925318179e-05
590
+ 2023-03-25 18:44:00,475 44k INFO Saving model and optimizer state at iteration 413 to ./logs/44k/G_13600.pth
591
+ 2023-03-25 18:44:01,790 44k INFO Saving model and optimizer state at iteration 413 to ./logs/44k/D_13600.pth
592
+ 2023-03-25 18:44:13,600 44k INFO ====> Epoch: 413, cost 22.69 s
so-vits-svc-4.0/README.md ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ 推理项目地址
2
+ Inference project address
3
+
4
+ https://github.com/zwa73/so-vits-svc-fork
5
+ https://github.com/svc-develop-team/so-vits-svc