Hugging Face's logo

Spaces:
Quitterie
/
Xenoblade-2-Style-Bert-VITS2-Editor
Running

App Files Community
Quitterie commited on
Commit
8ce7d5f
•
1 Parent(s): b28b05e

Upload 29 files

Browse files
Files changed (29) hide show
  1. model_assets/hikari/config.json +122 -0
  2. model_assets/hikari/config.json.bak +118 -0
  3. model_assets/hikari/hikari_e100_s6800.safetensors +3 -0
  4. model_assets/hikari/hikari_e74_s5000.safetensors +3 -0
  5. model_assets/hikari/style_vectors.npy +3 -0
  6. model_assets/hikari/style_vectors.npy.bak +0 -0
  7. model_assets/homura/config.json +121 -0
  8. model_assets/homura/config.json.bak +121 -0
  9. model_assets/homura/homura_e100_s7100.safetensors +3 -0
  10. model_assets/homura/homura_e71_s5000.safetensors +3 -0
  11. model_assets/homura/style_vectors.npy +3 -0
  12. model_assets/homura/style_vectors.npy.bak +0 -0
  13. model_assets/nia/config.json +122 -0
  14. model_assets/nia/config.json.bak +118 -0
  15. model_assets/nia/nia_e100_s17400.safetensors +3 -0
  16. model_assets/nia/nia_e29_s5000.safetensors +3 -0
  17. model_assets/nia/nia_e58_s10000.safetensors +3 -0
  18. model_assets/nia/nia_e87_s15000.safetensors +3 -0
  19. model_assets/nia/style_vectors.npy +3 -0
  20. model_assets/nia/style_vectors.npy.bak +0 -0
  21. model_assets/rex/config.json +122 -0
  22. model_assets/rex/config.json.bak +118 -0
  23. model_assets/rex/rex_e100_s21900.safetensors +3 -0
  24. model_assets/rex/rex_e23_s5000.safetensors +3 -0
  25. model_assets/rex/rex_e46_s10000.safetensors +3 -0
  26. model_assets/rex/rex_e69_s15000.safetensors +3 -0
  27. model_assets/rex/rex_e92_s20000.safetensors +3 -0
  28. model_assets/rex/style_vectors.npy +3 -0
  29. model_assets/rex/style_vectors.npy.bak +0 -0
model_assets/hikari/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "hikari",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\hikari\\train.list",
35
+ "validation_files": "Data\\hikari\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "hikari": 0
49
+ },
50
+ "num_styles": 5,
51
+ "style2id": {
52
+ "Neutral": 0,
53
+ "Angry": 1,
54
+ "Arrogant": 2,
55
+ "Calm": 3,
56
+ "Shout": 4
57
+ }
58
+ },
59
+ "model": {
60
+ "use_spk_conditioned_encoder": true,
61
+ "use_noise_scaled_mas": true,
62
+ "use_mel_posterior_encoder": false,
63
+ "use_duration_discriminator": false,
64
+ "use_wavlm_discriminator": true,
65
+ "inter_channels": 192,
66
+ "hidden_channels": 192,
67
+ "filter_channels": 768,
68
+ "n_heads": 2,
69
+ "n_layers": 6,
70
+ "kernel_size": 3,
71
+ "p_dropout": 0.1,
72
+ "resblock": "1",
73
+ "resblock_kernel_sizes": [
74
+ 3,
75
+ 7,
76
+ 11
77
+ ],
78
+ "resblock_dilation_sizes": [
79
+ [
80
+ 1,
81
+ 3,
82
+ 5
83
+ ],
84
+ [
85
+ 1,
86
+ 3,
87
+ 5
88
+ ],
89
+ [
90
+ 1,
91
+ 3,
92
+ 5
93
+ ]
94
+ ],
95
+ "upsample_rates": [
96
+ 8,
97
+ 8,
98
+ 2,
99
+ 2,
100
+ 2
101
+ ],
102
+ "upsample_initial_channel": 512,
103
+ "upsample_kernel_sizes": [
104
+ 16,
105
+ 16,
106
+ 8,
107
+ 2,
108
+ 2
109
+ ],
110
+ "n_layers_q": 3,
111
+ "use_spectral_norm": false,
112
+ "gin_channels": 512,
113
+ "slm": {
114
+ "model": "./slm/wavlm-base-plus",
115
+ "sr": 16000,
116
+ "hidden": 768,
117
+ "nlayers": 13,
118
+ "initial_channel": 64
119
+ }
120
+ },
121
+ "version": "2.4.1-JP-Extra"
122
+ }
model_assets/hikari/config.json.bak ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "hikari",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\hikari\\train.list",
35
+ "validation_files": "Data\\hikari\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "hikari": 0
49
+ },
50
+ "num_styles": 1,
51
+ "style2id": {
52
+ "Neutral": 0
53
+ }
54
+ },
55
+ "model": {
56
+ "use_spk_conditioned_encoder": true,
57
+ "use_noise_scaled_mas": true,
58
+ "use_mel_posterior_encoder": false,
59
+ "use_duration_discriminator": false,
60
+ "use_wavlm_discriminator": true,
61
+ "inter_channels": 192,
62
+ "hidden_channels": 192,
63
+ "filter_channels": 768,
64
+ "n_heads": 2,
65
+ "n_layers": 6,
66
+ "kernel_size": 3,
67
+ "p_dropout": 0.1,
68
+ "resblock": "1",
69
+ "resblock_kernel_sizes": [
70
+ 3,
71
+ 7,
72
+ 11
73
+ ],
74
+ "resblock_dilation_sizes": [
75
+ [
76
+ 1,
77
+ 3,
78
+ 5
79
+ ],
80
+ [
81
+ 1,
82
+ 3,
83
+ 5
84
+ ],
85
+ [
86
+ 1,
87
+ 3,
88
+ 5
89
+ ]
90
+ ],
91
+ "upsample_rates": [
92
+ 8,
93
+ 8,
94
+ 2,
95
+ 2,
96
+ 2
97
+ ],
98
+ "upsample_initial_channel": 512,
99
+ "upsample_kernel_sizes": [
100
+ 16,
101
+ 16,
102
+ 8,
103
+ 2,
104
+ 2
105
+ ],
106
+ "n_layers_q": 3,
107
+ "use_spectral_norm": false,
108
+ "gin_channels": 512,
109
+ "slm": {
110
+ "model": "./slm/wavlm-base-plus",
111
+ "sr": 16000,
112
+ "hidden": 768,
113
+ "nlayers": 13,
114
+ "initial_channel": 64
115
+ }
116
+ },
117
+ "version": "2.4.1-JP-Extra"
118
+ }
model_assets/hikari/hikari_e100_s6800.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:177e76de4280bd83ac89f009ef5534e20193468651ea47bacf8838f13501e6d9
3
+ size 251150980
model_assets/hikari/hikari_e74_s5000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6359acd165ac1925ff39d356d08b20f87e76b9f02e3146c51c1768144eceec2d
3
+ size 251150980
model_assets/hikari/style_vectors.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64633f7f0dbeee5672ecb3b46ed280a6941b060cbe3adf36a854eec61ccb4075
3
+ size 5248
model_assets/hikari/style_vectors.npy.bak ADDED
Binary file (1.15 kB). View file
 
model_assets/homura/config.json ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "homura",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\homura\\train.list",
35
+ "validation_files": "Data\\homura\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "homura": 0
49
+ },
50
+ "num_styles": 4,
51
+ "style2id": {
52
+ "Neutral": 0,
53
+ "Aggressive": 1,
54
+ "Calm": 2,
55
+ "Happy": 3
56
+ }
57
+ },
58
+ "model": {
59
+ "use_spk_conditioned_encoder": true,
60
+ "use_noise_scaled_mas": true,
61
+ "use_mel_posterior_encoder": false,
62
+ "use_duration_discriminator": false,
63
+ "use_wavlm_discriminator": true,
64
+ "inter_channels": 192,
65
+ "hidden_channels": 192,
66
+ "filter_channels": 768,
67
+ "n_heads": 2,
68
+ "n_layers": 6,
69
+ "kernel_size": 3,
70
+ "p_dropout": 0.1,
71
+ "resblock": "1",
72
+ "resblock_kernel_sizes": [
73
+ 3,
74
+ 7,
75
+ 11
76
+ ],
77
+ "resblock_dilation_sizes": [
78
+ [
79
+ 1,
80
+ 3,
81
+ 5
82
+ ],
83
+ [
84
+ 1,
85
+ 3,
86
+ 5
87
+ ],
88
+ [
89
+ 1,
90
+ 3,
91
+ 5
92
+ ]
93
+ ],
94
+ "upsample_rates": [
95
+ 8,
96
+ 8,
97
+ 2,
98
+ 2,
99
+ 2
100
+ ],
101
+ "upsample_initial_channel": 512,
102
+ "upsample_kernel_sizes": [
103
+ 16,
104
+ 16,
105
+ 8,
106
+ 2,
107
+ 2
108
+ ],
109
+ "n_layers_q": 3,
110
+ "use_spectral_norm": false,
111
+ "gin_channels": 512,
112
+ "slm": {
113
+ "model": "./slm/wavlm-base-plus",
114
+ "sr": 16000,
115
+ "hidden": 768,
116
+ "nlayers": 13,
117
+ "initial_channel": 64
118
+ }
119
+ },
120
+ "version": "2.4.1-JP-Extra"
121
+ }
model_assets/homura/config.json.bak ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "homura",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\homura\\train.list",
35
+ "validation_files": "Data\\homura\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "homura": 0
49
+ },
50
+ "num_styles": 4,
51
+ "style2id": {
52
+ "Neutral": 0,
53
+ "Cry out": 1,
54
+ "Calm": 2,
55
+ "Happy": 3
56
+ }
57
+ },
58
+ "model": {
59
+ "use_spk_conditioned_encoder": true,
60
+ "use_noise_scaled_mas": true,
61
+ "use_mel_posterior_encoder": false,
62
+ "use_duration_discriminator": false,
63
+ "use_wavlm_discriminator": true,
64
+ "inter_channels": 192,
65
+ "hidden_channels": 192,
66
+ "filter_channels": 768,
67
+ "n_heads": 2,
68
+ "n_layers": 6,
69
+ "kernel_size": 3,
70
+ "p_dropout": 0.1,
71
+ "resblock": "1",
72
+ "resblock_kernel_sizes": [
73
+ 3,
74
+ 7,
75
+ 11
76
+ ],
77
+ "resblock_dilation_sizes": [
78
+ [
79
+ 1,
80
+ 3,
81
+ 5
82
+ ],
83
+ [
84
+ 1,
85
+ 3,
86
+ 5
87
+ ],
88
+ [
89
+ 1,
90
+ 3,
91
+ 5
92
+ ]
93
+ ],
94
+ "upsample_rates": [
95
+ 8,
96
+ 8,
97
+ 2,
98
+ 2,
99
+ 2
100
+ ],
101
+ "upsample_initial_channel": 512,
102
+ "upsample_kernel_sizes": [
103
+ 16,
104
+ 16,
105
+ 8,
106
+ 2,
107
+ 2
108
+ ],
109
+ "n_layers_q": 3,
110
+ "use_spectral_norm": false,
111
+ "gin_channels": 512,
112
+ "slm": {
113
+ "model": "./slm/wavlm-base-plus",
114
+ "sr": 16000,
115
+ "hidden": 768,
116
+ "nlayers": 13,
117
+ "initial_channel": 64
118
+ }
119
+ },
120
+ "version": "2.4.1-JP-Extra"
121
+ }
model_assets/homura/homura_e100_s7100.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:231aab2af2e1f3b5bbe4699e340ef34a26ff7e649bf16625e70d1cc96bb33a11
3
+ size 251150980
model_assets/homura/homura_e71_s5000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fbc0a7ea8f02b101ff2edf69c8e4bf56d5c03e6708e14aaed700c297b75fa40
3
+ size 251150980
model_assets/homura/style_vectors.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:398d54354def517400ee18c676d57d4fc5c7ed930e9c943b22381481ad1e1444
3
+ size 4224
model_assets/homura/style_vectors.npy.bak ADDED
Binary file (4.22 kB). View file
 
model_assets/nia/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "nia",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\nia\\train.list",
35
+ "validation_files": "Data\\nia\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "nia": 0
49
+ },
50
+ "num_styles": 5,
51
+ "style2id": {
52
+ "Neutral": 0,
53
+ "Happy": 1,
54
+ "Shout": 2,
55
+ "Weak": 3,
56
+ "Angry": 4
57
+ }
58
+ },
59
+ "model": {
60
+ "use_spk_conditioned_encoder": true,
61
+ "use_noise_scaled_mas": true,
62
+ "use_mel_posterior_encoder": false,
63
+ "use_duration_discriminator": false,
64
+ "use_wavlm_discriminator": true,
65
+ "inter_channels": 192,
66
+ "hidden_channels": 192,
67
+ "filter_channels": 768,
68
+ "n_heads": 2,
69
+ "n_layers": 6,
70
+ "kernel_size": 3,
71
+ "p_dropout": 0.1,
72
+ "resblock": "1",
73
+ "resblock_kernel_sizes": [
74
+ 3,
75
+ 7,
76
+ 11
77
+ ],
78
+ "resblock_dilation_sizes": [
79
+ [
80
+ 1,
81
+ 3,
82
+ 5
83
+ ],
84
+ [
85
+ 1,
86
+ 3,
87
+ 5
88
+ ],
89
+ [
90
+ 1,
91
+ 3,
92
+ 5
93
+ ]
94
+ ],
95
+ "upsample_rates": [
96
+ 8,
97
+ 8,
98
+ 2,
99
+ 2,
100
+ 2
101
+ ],
102
+ "upsample_initial_channel": 512,
103
+ "upsample_kernel_sizes": [
104
+ 16,
105
+ 16,
106
+ 8,
107
+ 2,
108
+ 2
109
+ ],
110
+ "n_layers_q": 3,
111
+ "use_spectral_norm": false,
112
+ "gin_channels": 512,
113
+ "slm": {
114
+ "model": "./slm/wavlm-base-plus",
115
+ "sr": 16000,
116
+ "hidden": 768,
117
+ "nlayers": 13,
118
+ "initial_channel": 64
119
+ }
120
+ },
121
+ "version": "2.4.1-JP-Extra"
122
+ }
model_assets/nia/config.json.bak ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "nia",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\nia\\train.list",
35
+ "validation_files": "Data\\nia\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "nia": 0
49
+ },
50
+ "num_styles": 1,
51
+ "style2id": {
52
+ "Neutral": 0
53
+ }
54
+ },
55
+ "model": {
56
+ "use_spk_conditioned_encoder": true,
57
+ "use_noise_scaled_mas": true,
58
+ "use_mel_posterior_encoder": false,
59
+ "use_duration_discriminator": false,
60
+ "use_wavlm_discriminator": true,
61
+ "inter_channels": 192,
62
+ "hidden_channels": 192,
63
+ "filter_channels": 768,
64
+ "n_heads": 2,
65
+ "n_layers": 6,
66
+ "kernel_size": 3,
67
+ "p_dropout": 0.1,
68
+ "resblock": "1",
69
+ "resblock_kernel_sizes": [
70
+ 3,
71
+ 7,
72
+ 11
73
+ ],
74
+ "resblock_dilation_sizes": [
75
+ [
76
+ 1,
77
+ 3,
78
+ 5
79
+ ],
80
+ [
81
+ 1,
82
+ 3,
83
+ 5
84
+ ],
85
+ [
86
+ 1,
87
+ 3,
88
+ 5
89
+ ]
90
+ ],
91
+ "upsample_rates": [
92
+ 8,
93
+ 8,
94
+ 2,
95
+ 2,
96
+ 2
97
+ ],
98
+ "upsample_initial_channel": 512,
99
+ "upsample_kernel_sizes": [
100
+ 16,
101
+ 16,
102
+ 8,
103
+ 2,
104
+ 2
105
+ ],
106
+ "n_layers_q": 3,
107
+ "use_spectral_norm": false,
108
+ "gin_channels": 512,
109
+ "slm": {
110
+ "model": "./slm/wavlm-base-plus",
111
+ "sr": 16000,
112
+ "hidden": 768,
113
+ "nlayers": 13,
114
+ "initial_channel": 64
115
+ }
116
+ },
117
+ "version": "2.4.1-JP-Extra"
118
+ }
model_assets/nia/nia_e100_s17400.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad3b98a8dd04ea9bca4b97065b077a39b0ff5ae11c3e5971b5d44ff4789b2749
3
+ size 251150980
model_assets/nia/nia_e29_s5000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef5290b23bf70c623e9a6406b37c66d27ca7e520e99dc73ec5370d75b2334d48
3
+ size 251150980
model_assets/nia/nia_e58_s10000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01e4a96e95ffb7997b6537aba2d48e71e40376f13c98daf47da498c165ab891f
3
+ size 251150980
model_assets/nia/nia_e87_s15000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2298b05f9d9d8e9437f6a269098617805924794ebb2c1772aaaf992fe3f144d4
3
+ size 251150980
model_assets/nia/style_vectors.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b09f69981fad660999c83c63bbe2656b7a160ba130b8f091efa64d90650a46e0
3
+ size 5248
model_assets/nia/style_vectors.npy.bak ADDED
Binary file (1.15 kB). View file
 
model_assets/rex/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "rex",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\rex\\train.list",
35
+ "validation_files": "Data\\rex\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "rex": 0
49
+ },
50
+ "num_styles": 5,
51
+ "style2id": {
52
+ "Neutral": 0,
53
+ "Shout": 1,
54
+ "fun": 2,
55
+ "Aggressive": 3,
56
+ "Cheer": 4
57
+ }
58
+ },
59
+ "model": {
60
+ "use_spk_conditioned_encoder": true,
61
+ "use_noise_scaled_mas": true,
62
+ "use_mel_posterior_encoder": false,
63
+ "use_duration_discriminator": false,
64
+ "use_wavlm_discriminator": true,
65
+ "inter_channels": 192,
66
+ "hidden_channels": 192,
67
+ "filter_channels": 768,
68
+ "n_heads": 2,
69
+ "n_layers": 6,
70
+ "kernel_size": 3,
71
+ "p_dropout": 0.1,
72
+ "resblock": "1",
73
+ "resblock_kernel_sizes": [
74
+ 3,
75
+ 7,
76
+ 11
77
+ ],
78
+ "resblock_dilation_sizes": [
79
+ [
80
+ 1,
81
+ 3,
82
+ 5
83
+ ],
84
+ [
85
+ 1,
86
+ 3,
87
+ 5
88
+ ],
89
+ [
90
+ 1,
91
+ 3,
92
+ 5
93
+ ]
94
+ ],
95
+ "upsample_rates": [
96
+ 8,
97
+ 8,
98
+ 2,
99
+ 2,
100
+ 2
101
+ ],
102
+ "upsample_initial_channel": 512,
103
+ "upsample_kernel_sizes": [
104
+ 16,
105
+ 16,
106
+ 8,
107
+ 2,
108
+ 2
109
+ ],
110
+ "n_layers_q": 3,
111
+ "use_spectral_norm": false,
112
+ "gin_channels": 512,
113
+ "slm": {
114
+ "model": "./slm/wavlm-base-plus",
115
+ "sr": 16000,
116
+ "hidden": 768,
117
+ "nlayers": 13,
118
+ "initial_channel": 64
119
+ }
120
+ },
121
+ "version": "2.4.1-JP-Extra"
122
+ }
model_assets/rex/config.json.bak ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "rex",
3
+ "train": {
4
+ "log_interval": 200,
5
+ "eval_interval": 5000,
6
+ "seed": 42,
7
+ "epochs": 100,
8
+ "learning_rate": 0.0001,
9
+ "betas": [
10
+ 0.8,
11
+ 0.99
12
+ ],
13
+ "eps": 1e-09,
14
+ "batch_size": 2,
15
+ "bf16_run": false,
16
+ "fp16_run": false,
17
+ "lr_decay": 0.99996,
18
+ "segment_size": 16384,
19
+ "init_lr_ratio": 1,
20
+ "warmup_epochs": 0,
21
+ "c_mel": 45,
22
+ "c_kl": 1.0,
23
+ "c_commit": 100,
24
+ "skip_optimizer": false,
25
+ "freeze_ZH_bert": false,
26
+ "freeze_JP_bert": false,
27
+ "freeze_EN_bert": false,
28
+ "freeze_emo": false,
29
+ "freeze_style": false,
30
+ "freeze_decoder": false
31
+ },
32
+ "data": {
33
+ "use_jp_extra": true,
34
+ "training_files": "Data\\rex\\train.list",
35
+ "validation_files": "Data\\rex\\val.list",
36
+ "max_wav_value": 32768.0,
37
+ "sampling_rate": 44100,
38
+ "filter_length": 2048,
39
+ "hop_length": 512,
40
+ "win_length": 2048,
41
+ "n_mel_channels": 128,
42
+ "mel_fmin": 0.0,
43
+ "mel_fmax": null,
44
+ "add_blank": true,
45
+ "n_speakers": 1,
46
+ "cleaned_text": true,
47
+ "spk2id": {
48
+ "rex": 0
49
+ },
50
+ "num_styles": 1,
51
+ "style2id": {
52
+ "Neutral": 0
53
+ }
54
+ },
55
+ "model": {
56
+ "use_spk_conditioned_encoder": true,
57
+ "use_noise_scaled_mas": true,
58
+ "use_mel_posterior_encoder": false,
59
+ "use_duration_discriminator": false,
60
+ "use_wavlm_discriminator": true,
61
+ "inter_channels": 192,
62
+ "hidden_channels": 192,
63
+ "filter_channels": 768,
64
+ "n_heads": 2,
65
+ "n_layers": 6,
66
+ "kernel_size": 3,
67
+ "p_dropout": 0.1,
68
+ "resblock": "1",
69
+ "resblock_kernel_sizes": [
70
+ 3,
71
+ 7,
72
+ 11
73
+ ],
74
+ "resblock_dilation_sizes": [
75
+ [
76
+ 1,
77
+ 3,
78
+ 5
79
+ ],
80
+ [
81
+ 1,
82
+ 3,
83
+ 5
84
+ ],
85
+ [
86
+ 1,
87
+ 3,
88
+ 5
89
+ ]
90
+ ],
91
+ "upsample_rates": [
92
+ 8,
93
+ 8,
94
+ 2,
95
+ 2,
96
+ 2
97
+ ],
98
+ "upsample_initial_channel": 512,
99
+ "upsample_kernel_sizes": [
100
+ 16,
101
+ 16,
102
+ 8,
103
+ 2,
104
+ 2
105
+ ],
106
+ "n_layers_q": 3,
107
+ "use_spectral_norm": false,
108
+ "gin_channels": 512,
109
+ "slm": {
110
+ "model": "./slm/wavlm-base-plus",
111
+ "sr": 16000,
112
+ "hidden": 768,
113
+ "nlayers": 13,
114
+ "initial_channel": 64
115
+ }
116
+ },
117
+ "version": "2.4.1-JP-Extra"
118
+ }
model_assets/rex/rex_e100_s21900.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:124f99739bea9ee6dd7fa1fcaf327f01afe6c2b36feca49702b0acade713987a
3
+ size 251150980
model_assets/rex/rex_e23_s5000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b765815d0cbdf3965afed7fc58b8f8c74ac5a9422c289b7f66aba24aa52d2918
3
+ size 251150980
model_assets/rex/rex_e46_s10000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7d6d58530b465a2fdf3a834bf9c654843c8c75c727993fa530649ff46bb23e7
3
+ size 251150980
model_assets/rex/rex_e69_s15000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f620517aca5b75fbd228785a22190d815238650dd3f233eda9a0daee19f6614
3
+ size 251150980
model_assets/rex/rex_e92_s20000.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49e257238d9ff1da5ecfc13234d4f34e25d1ba28c6a1509208decd5942cb63cb
3
+ size 251150980
model_assets/rex/style_vectors.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:557110e7986d6fae8c6ca2b1ca10dea606a5c41462b143417d07222210620dff
3
+ size 5248
model_assets/rex/style_vectors.npy.bak ADDED
Binary file (1.15 kB). View file