rsxdalv / tts-generation-webui

TTS Generation Web UI (Bark, MusicGen + AudioGen, Tortoise, RVC, Vocos, Demucs, SeamlessM4T, MAGNet, StyleTTS2, MMS)
https://rsxdalv.github.io/tts-generation-webui/
MIT License
1.45k stars 159 forks source link

voice Cloning RU #317

Open KPY7030P opened 3 weeks ago

KPY7030P commented 3 weeks ago

Hi! there are models based on HuBert, but for the Russian language, I'm not sure that this is what is needed, but if so, then it would be good to add them. https://cloud.ru/ru/datahub/rugpt3family/rubert-base https://habr.com/ru/companies/sberbank/articles/567776/

rsxdalv commented 3 weeks ago

It seems to have a different format, there also needs to be a model that translates the tokens for use by Bark, perhaps gitmylo who made the HuBert based voice cloning used in this project can weigh in on it: https://github.com/gitmylo/bark-voice-cloning-HuBERT-quantizer/issues