metadata
license: mit
base_model:
- Akjava/matcha-tts_ja_single-speaker_ja005
This model has single speaker model
家具商人のフィシェルは、荷車と仔馬を貸してくれた。
spk0:Annie
こんにちは、アニーだよ
spk0:Annie
license
My training data is created by MIT Licensed model output. https://huggingface.co/ResembleAI/chatterbox
Matcha-TTS is MIT https://github.com/shivammehta25/Matcha-TTS
Training
need checkpoint from there https://huggingface.co/Akjava/matcha-tts_ja_single-speaker_ja005
Use this. https://github.com/akjava/Matcha-TTS-Japanese
Some dataset file name is Japanese,on some env, you need to change file name.
Demo
https://ai-game-bu.itch.io/ai-gaming-voice
similar model
- this one is faild model.trying to train few audios,but faild. https://huggingface.co/Akjava/matcha-tts_ja_single-speaker_ja001