xhb commited on
Commit
ef129fe
1 Parent(s): 8219fd0

feat(模型): 首次提交vits模型

Browse files
README.md CHANGED
@@ -1,3 +1,15 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+
5
+ ## vits声线模仿,中文tts
6
+
7
+ 数据集中包含4个角色的声音:
8
+
9
+ * 派蒙(paimon) 女声
10
+ * 八重神子(miko) 女声
11
+ * 钟离(zhongli) 男声
12
+ * 云飞(yunfeimix) 男声
13
+
14
+
15
+
miko.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 1000,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 2e-4,
8
+ "betas": [0.8, 0.99],
9
+ "eps": 1e-9,
10
+ "batch_size": 48,
11
+ "fp16_run": true,
12
+ "lr_decay": 0.999875,
13
+ "segment_size": 8192,
14
+ "init_lr_ratio": 1,
15
+ "warmup_epochs": 0,
16
+ "c_mel": 45,
17
+ "c_kl": 1.0
18
+ },
19
+ "data": {
20
+ "training_files":"filelists/miko_train_chs.txt.cleaned",
21
+ "validation_files":"filelists/miko_val_chs.txt.cleaned",
22
+ "text_cleaners":["chinese_cleaners"],
23
+ "max_wav_value": 32768.0,
24
+ "sampling_rate": 48000,
25
+ "filter_length": 1024,
26
+ "hop_length": 256,
27
+ "win_length": 1024,
28
+ "n_mel_channels": 80,
29
+ "mel_fmin": 0.0,
30
+ "mel_fmax": null,
31
+ "add_blank": true,
32
+ "n_speakers": 0,
33
+ "cleaned_text": true
34
+ },
35
+ "model": {
36
+ "inter_channels": 192,
37
+ "hidden_channels": 192,
38
+ "filter_channels": 768,
39
+ "n_heads": 2,
40
+ "n_layers": 6,
41
+ "kernel_size": 3,
42
+ "p_dropout": 0.1,
43
+ "resblock": "1",
44
+ "resblock_kernel_sizes": [3,7,11],
45
+ "resblock_dilation_sizes": [[1,3,5], [1,3,5], [1,3,5]],
46
+ "upsample_rates": [8,8,2,2],
47
+ "upsample_initial_channel": 512,
48
+ "upsample_kernel_sizes": [16,16,4,4],
49
+ "n_layers_q": 3,
50
+ "use_spectral_norm": false,
51
+ "gin_channels": 256
52
+ },
53
+ "speakers": ["miko"],
54
+ "symbols": ["_", "\uff0c", "\u3002", "\uff01", "\uff1f", "\u2014", "\u2026", "\u3105", "\u3106", "\u3107", "\u3108", "\u3109", "\u310a", "\u310b", "\u310c", "\u310d", "\u310e", "\u310f", "\u3110", "\u3111", "\u3112", "\u3113", "\u3114", "\u3115", "\u3116", "\u3117", "\u3118", "\u3119", "\u311a", "\u311b", "\u311c", "\u311d", "\u311e", "\u311f", "\u3120", "\u3121", "\u3122", "\u3123", "\u3124", "\u3125", "\u3126", "\u3127", "\u3128", "\u3129", "\u02c9", "\u02ca", "\u02c7", "\u02cb", "\u02d9", " "]
55
+ }
miko_139k.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f71962de5a05a61fc51d9f12af981d3d4912701130eff6d832a9d41df8f92885
3
+ size 449983387
miko_90k.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9cd73817c59344a24426d0d4a9c2804f1e7d28bb3f462b3417127d10222cb66
3
+ size 449979566
paimon6k.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 1000,
5
+ "seed": 1234,
6
+ "epochs": 40000,
7
+ "learning_rate": 2e-4,
8
+ "betas": [0.8, 0.99],
9
+ "eps": 1e-9,
10
+ "batch_size": 46,
11
+ "fp16_run": true,
12
+ "lr_decay": 0.999875,
13
+ "segment_size": 8192,
14
+ "init_lr_ratio": 1,
15
+ "warmup_epochs": 0,
16
+ "c_mel": 45,
17
+ "c_kl": 1.0
18
+ },
19
+ "data": {
20
+ "training_files":"filelists/paimon_6k_train_chs.txt.cleaned",
21
+ "validation_files":"filelists/paimon_6k_val_chs.txt.cleaned",
22
+ "text_cleaners":["chinese_cleaners"],
23
+ "max_wav_value": 32768.0,
24
+ "sampling_rate": 48000,
25
+ "filter_length": 1024,
26
+ "hop_length": 256,
27
+ "win_length": 1024,
28
+ "n_mel_channels": 80,
29
+ "mel_fmin": 0.0,
30
+ "mel_fmax": null,
31
+ "add_blank": true,
32
+ "n_speakers": 0,
33
+ "cleaned_text": true
34
+ },
35
+ "model": {
36
+ "inter_channels": 192,
37
+ "hidden_channels": 192,
38
+ "filter_channels": 768,
39
+ "n_heads": 2,
40
+ "n_layers": 6,
41
+ "kernel_size": 3,
42
+ "p_dropout": 0.1,
43
+ "resblock": "1",
44
+ "resblock_kernel_sizes": [3,7,11],
45
+ "resblock_dilation_sizes": [[1,3,5], [1,3,5], [1,3,5]],
46
+ "upsample_rates": [8,8,2,2],
47
+ "upsample_initial_channel": 512,
48
+ "upsample_kernel_sizes": [16,16,4,4],
49
+ "n_layers_q": 3,
50
+ "use_spectral_norm": false,
51
+ "gin_channels": 256
52
+ },
53
+ "speakers": ["paimon"],
54
+ "symbols": ["_", "\uff0c", "\u3002", "\uff01", "\uff1f", "\u2014", "\u2026", "\u3105", "\u3106", "\u3107", "\u3108", "\u3109", "\u310a", "\u310b", "\u310c", "\u310d", "\u310e", "\u310f", "\u3110", "\u3111", "\u3112", "\u3113", "\u3114", "\u3115", "\u3116", "\u3117", "\u3118", "\u3119", "\u311a", "\u311b", "\u311c", "\u311d", "\u311e", "\u311f", "\u3120", "\u3121", "\u3122", "\u3123", "\u3124", "\u3125", "\u3126", "\u3127", "\u3128", "\u3129", "\u02c9", "\u02ca", "\u02c7", "\u02cb", "\u02d9", " "]
55
+ }
paimon6k_390k.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b282f7fed04b92186ae8ed1956496e4492d1ee07d84870143c6e7fd3b82d5cb6
3
+ size 449983387
put_vits_model_here ADDED
File without changes
yunfeimix2.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 50,
4
+ "eval_interval": 1000,
5
+ "seed": 1234,
6
+ "epochs": 40000,
7
+ "learning_rate": 2e-4,
8
+ "betas": [0.8, 0.99],
9
+ "eps": 1e-9,
10
+ "batch_size": 48,
11
+ "fp16_run": false,
12
+ "lr_decay": 0.999875,
13
+ "segment_size": 8192,
14
+ "init_lr_ratio": 1,
15
+ "warmup_epochs": 0,
16
+ "c_mel": 45,
17
+ "c_kl": 1.0
18
+ },
19
+ "data": {
20
+ "training_files":"filelists/yunfeimix2_train.txt.cleaned",
21
+ "validation_files":"filelists/yunfeimix2_test.txt.cleaned",
22
+ "text_cleaners":["chinese_cleaners"],
23
+ "max_wav_value": 32768.0,
24
+ "sampling_rate": 48000,
25
+ "filter_length": 1024,
26
+ "hop_length": 256,
27
+ "win_length": 1024,
28
+ "n_mel_channels": 80,
29
+ "mel_fmin": 0.0,
30
+ "mel_fmax": null,
31
+ "add_blank": true,
32
+ "n_speakers": 0,
33
+ "cleaned_text": true
34
+ },
35
+ "model": {
36
+ "inter_channels": 192,
37
+ "hidden_channels": 192,
38
+ "filter_channels": 768,
39
+ "n_heads": 2,
40
+ "n_layers": 6,
41
+ "kernel_size": 3,
42
+ "p_dropout": 0.1,
43
+ "resblock": "1",
44
+ "resblock_kernel_sizes": [3,7,11],
45
+ "resblock_dilation_sizes": [[1,3,5], [1,3,5], [1,3,5]],
46
+ "upsample_rates": [8,8,2,2],
47
+ "upsample_initial_channel": 512,
48
+ "upsample_kernel_sizes": [16,16,4,4],
49
+ "n_layers_q": 3,
50
+ "use_spectral_norm": false,
51
+ "gin_channels": 256
52
+ },
53
+ "speakers": ["yunfei"],
54
+ "symbols": ["_", "\uff0c", "\u3002", "\uff01", "\uff1f", "\u2014", "\u2026", "\u3105", "\u3106", "\u3107", "\u3108", "\u3109", "\u310a", "\u310b", "\u310c", "\u310d", "\u310e", "\u310f", "\u3110", "\u3111", "\u3112", "\u3113", "\u3114", "\u3115", "\u3116", "\u3117", "\u3118", "\u3119", "\u311a", "\u311b", "\u311c", "\u311d", "\u311e", "\u311f", "\u3120", "\u3121", "\u3122", "\u3123", "\u3124", "\u3125", "\u3126", "\u3127", "\u3128", "\u3129", "\u02c9", "\u02ca", "\u02c7", "\u02cb", "\u02d9", " "]
55
+ }
yunfeimix2_53k.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b2a080ba1d12c3acfe1741a1bf94458fc29f109bfd9432ad3a324a761460871
3
+ size 449979566
zhongli.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train": {
3
+ "log_interval": 200,
4
+ "eval_interval": 1000,
5
+ "seed": 1234,
6
+ "epochs": 10000,
7
+ "learning_rate": 2e-4,
8
+ "betas": [0.8, 0.99],
9
+ "eps": 1e-9,
10
+ "batch_size": 48,
11
+ "fp16_run": true,
12
+ "lr_decay": 0.999875,
13
+ "segment_size": 8192,
14
+ "init_lr_ratio": 1,
15
+ "warmup_epochs": 0,
16
+ "c_mel": 45,
17
+ "c_kl": 1.0
18
+ },
19
+ "data": {
20
+ "training_files":"filelists/zhongli_train.txt.cleaned",
21
+ "validation_files":"filelists/zhongli_val.txt.cleaned",
22
+ "text_cleaners":["chinese_cleaners"],
23
+ "max_wav_value": 32768.0,
24
+ "sampling_rate": 48000,
25
+ "filter_length": 1024,
26
+ "hop_length": 256,
27
+ "win_length": 1024,
28
+ "n_mel_channels": 80,
29
+ "mel_fmin": 0.0,
30
+ "mel_fmax": null,
31
+ "add_blank": true,
32
+ "n_speakers": 0,
33
+ "cleaned_text": true
34
+ },
35
+ "model": {
36
+ "inter_channels": 192,
37
+ "hidden_channels": 192,
38
+ "filter_channels": 768,
39
+ "n_heads": 2,
40
+ "n_layers": 6,
41
+ "kernel_size": 3,
42
+ "p_dropout": 0.1,
43
+ "resblock": "1",
44
+ "resblock_kernel_sizes": [3,7,11],
45
+ "resblock_dilation_sizes": [[1,3,5], [1,3,5], [1,3,5]],
46
+ "upsample_rates": [8,8,2,2],
47
+ "upsample_initial_channel": 512,
48
+ "upsample_kernel_sizes": [16,16,4,4],
49
+ "n_layers_q": 3,
50
+ "use_spectral_norm": false,
51
+ "gin_channels": 256
52
+ },
53
+ "speakers": ["miko"],
54
+ "symbols": ["_", "\uff0c", "\u3002", "\uff01", "\uff1f", "\u2014", "\u2026", "\u3105", "\u3106", "\u3107", "\u3108", "\u3109", "\u310a", "\u310b", "\u310c", "\u310d", "\u310e", "\u310f", "\u3110", "\u3111", "\u3112", "\u3113", "\u3114", "\u3115", "\u3116", "\u3117", "\u3118", "\u3119", "\u311a", "\u311b", "\u311c", "\u311d", "\u311e", "\u311f", "\u3120", "\u3121", "\u3122", "\u3123", "\u3124", "\u3125", "\u3126", "\u3127", "\u3128", "\u3129", "\u02c9", "\u02ca", "\u02c7", "\u02cb", "\u02d9", " "]
55
+ }
zhongli_29k.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c27773c1626506245dc0479e1b2973d7373ccddfce035690e79482317a35334b
3
+ size 449979566
zhongli_44k.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a941b4d273e8fb6e4327fc7b7978908cacb2d91ced11d72e80be763806507c6
3
+ size 449979566