VITS2 Backbone with multilingual bert
For a quick guide, please refer to webui_preprocess.py.
For a simple tutorial, see webui_preprocess.py.
The new autoregressive TTS under FishAudio, Fish-Speech, is now available. It achieves the current open-source SOTA level and is continuously maintained. It is recommended to use this project as a replacement for BV2/GSV. This project will not be maintained in the short term.
Demo Video: https://www.bilibili.com/video/BV18E421371Q
Tech slides Video: https://www.bilibili.com/video/BV1zJ4m1K7cj
Please note, the core idea of this project comes from anyvoiceai/MassTTS, a very good tts project
MassTTS's demo is AI version of Feng Ge's sharp review of himself, and finding back his lost kidney in the Golden Triangle
//: # (## This project has no relation to PlayVoice/vits_chinese)
Experienced travelers/pioneers/captains/doctors/senseis/demon hunters/meow meows/V should refer to the code to learn how to train themselves.
- anyvoiceai/MassTTS
- jaywalnut310/vits
- p0p4k/vits2_pytorch
- svc-develop-team/so-vits-svc
- PaddlePaddle/PaddleSpeech
- emotional-vits
- fish-speech
- Bert-VITS2-UI
//: # (# All code references in this project are stated, the idea of the bert part comes from AI Feng Ge, and has no relation to vits_chinese. Everyone is welcome to check the code. At the same time, we also strongly condemn the behavior of clashing, even opening the box developers.)