kli017/KAN-TTS

KAN-TTS is a speech-synthesis training framework, please try the demos we have posted at https://modelscope.cn/models?page=1&tasks=text-to-speech

KAN-TTS

With KAN-TTS you can train your own TTS model from zero to hero :).

Models

Temporarily we support sam-bert and hifi-GAN, other models are coming soon.

Support Languages

Language	Model Links
Mandarin	https://modelscope.cn/models?name=zhcn&page=1&tasks=text-to-speech&type=audio
English	https://modelscope.cn/models?name=enus&page=1&tasks=text-to-speech&type=audio
British	https://modelscope.cn/models?name=engb&page=1&tasks=text-to-speech&type=audio
Shanghainese	https://modelscope.cn/models?name=WuuShanghai&page=1&tasks=text-to-speech&type=audio
Sichuanese	https://modelscope.cn/models?name=Sichuan&page=1&tasks=text-to-speech&type=audio
Cantonese	https://modelscope.cn/models?name=Cantonese&page=1&tasks=text-to-speech&type=audio
Italian	https://modelscope.cn/models?name=itit&page=1&tasks=text-to-speech&type=audio
Spanish	https://modelscope.cn/models?name=eses&page=1&tasks=text-to-speech&type=audio
Russian	https://modelscope.cn/models?name=ruru&page=1&tasks=text-to-speech&type=audio
Korean	https://modelscope.cn/models?name=kokr&page=1&tasks=text-to-speech&type=audio
More languages are coming soon.

Training Tutorial

You can find the training tutorial in our wiki page KAN-TTS Wiki.

ModelScope Demo

Try our demo on ModelScope KAN-TTS Demo.

Contribute to this repo

pip install -r requirements.txt
pre-commit install

Contact us

If you have any questions, please feel free to contact us.

Scan the QR code to join our DingTalk group.