I'll be adding a new pipeline for tts which include
TTS -> Voice Conversion -> audio.wav
The idea is to make users switch between any tts available but can customize how they sound just by training voice conversion model
Which will be a lot easier than training a TTS model that includes more complex data with this we open a room for multilingual TTS
In this case, anything other than English and Japanese that was implemented in the first place while laid foundation
for customizable TTS for more advanced user
Referring to THIS #27 PR we now add new pipeline to the repo
The docs of it are in progress but currently it is now possible to add your own TTS/Voice Conversion model easily
I'll be adding a new pipeline for tts which include TTS -> Voice Conversion -> audio.wav The idea is to make users switch between any tts available but can customize how they sound just by training voice conversion model Which will be a lot easier than training a TTS model that includes more complex data with this we open a room for multilingual TTS In this case, anything other than English and Japanese that was implemented in the first place while laid foundation for customizable TTS for more advanced user
Pipeline: https://github.com/HRNPH/AIwaifu-Voifu