|
--- |
|
license: other |
|
language: |
|
- ja |
|
tags: |
|
- Style-Bert-VITS2 |
|
--- |
|
|
|
# Style-Bert-VITS2向けの事前学習モデル |
|
|
|
[Style-Bert-VITS2](https://github.com/litagin02/Style-Bert-VITS2)で使用できる以下の学習データで学習を行ったクリーンな(*1)事前学習データになります |
|
|
|
(*1) |
|
ここでいうクリーンは事前学習に使用した学習データが明記されていることを指しています |
|
|
|
## 学習データセット |
|
* [つくよみちゃんコーパス│声優統計コーパス（JVSコーパス準拠）](https://tyc.rei-yumesaki.net/material/corpus/) |
|
* [みんなで作るJSUTコーパスbasic5000 BASIC5000_0001～BASIC5000_0600](https://tyc.rei-yumesaki.net/material/minnade-jsut/) （夢前黎担当部分を許可を得て使用） |
|
* [黄鏡博人さん](https://twitter.com/KikyoHiloto) からボイスデータをご提供いただきました |
|
* [あみたろの声素材工房](https://amitaro.net/) |
|
|
|
## 学習パラメータ |
|
* 学習ステップ数 : 600k step |
|
* bfloat16 : false |
|
|
|
**config.json** |
|
|
|
```json |
|
{ |
|
"model_name": "pretraing", |
|
"train": { |
|
"log_interval": 200, |
|
"eval_interval": 2000, |
|
"seed": 42, |
|
"epochs": 2100, |
|
"learning_rate": 0.0001, |
|
"betas": [ |
|
0.8, |
|
0.99 |
|
], |
|
"eps": 1e-09, |
|
"batch_size": 8, |
|
"bf16_run": false, |
|
"fp16_run": false, |
|
"lr_decay": 0.99996, |
|
"segment_size": 16384, |
|
"init_lr_ratio": 1, |
|
"warmup_epochs": 0, |
|
"c_mel": 45, |
|
"c_kl": 1.0, |
|
"c_commit": 100, |
|
"skip_optimizer": false, |
|
"freeze_ZH_bert": false, |
|
"freeze_JP_bert": false, |
|
"freeze_EN_bert": false, |
|
"freeze_emo": false, |
|
"freeze_style": false, |
|
"freeze_decoder": false |
|
}, |
|
"data": { |
|
"use_jp_extra": true, |
|
"training_files": "Data/pretraing/train.list", |
|
"validation_files": "Data/pretraing/val.list", |
|
"max_wav_value": 32768.0, |
|
"sampling_rate": 44100, |
|
"filter_length": 2048, |
|
"hop_length": 512, |
|
"win_length": 2048, |
|
"n_mel_channels": 128, |
|
"mel_fmin": 0.0, |
|
"mel_fmax": null, |
|
"add_blank": true, |
|
"n_speakers": 1, |
|
"cleaned_text": true, |
|
"spk2id": { |
|
"pretraing": 0 |
|
} |
|
}, |
|
"model": { |
|
"use_spk_conditioned_encoder": true, |
|
"use_noise_scaled_mas": true, |
|
"use_mel_posterior_encoder": false, |
|
"use_duration_discriminator": false, |
|
"use_wavlm_discriminator": true, |
|
"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, |
|
8, |
|
2, |
|
2 |
|
], |
|
"n_layers_q": 3, |
|
"use_spectral_norm": false, |
|
"gin_channels": 512, |
|
"slm": { |
|
"model": "./slm/wavlm-base-plus", |
|
"sr": 16000, |
|
"hidden": 768, |
|
"nlayers": 13, |
|
"initial_channel": 64 |
|
} |
|
}, |
|
"version": "2.4.1-JP-Extra" |
|
} |
|
``` |
|
|
|
## SpeechMOSによる自然性評価 |
|
|
|
mos_pretraing.csvも同封しています |
|
|
|
 |
|
|
|
# ライセンス |
|
|
|
ライセンスは、以下に準じます |
|
* [つくよみちゃんコーパス│声優統計コーパス（JVSコーパス準拠）](https://tyc.rei-yumesaki.net/material/corpus/) |
|
* [あみたろの声素材工房（https://amitaro.net/） フリー声素材ご利用規約](https://amitaro.net/voice/faq/#index_id6) |
