search

shivammehta25 / Matcha-TTS

[ICASSP 2024] 🍵 Matcha-TTS: A fast TTS architecture with conditional flow matching
https://shivammehta25.github.io/Matcha-TTS/
MIT License
747 stars 96 forks source link

diffusers dependency #88

Closed frostyplanet closed 3 months ago

frostyplanet commented 3 months ago

Please do not bind specified diffusers version, lead to pip dep resolve conflict with other library that requires newer version of diffusers

xinference[audio,embedding,image,rerank,transformers,vllm] 0+unknown depends on diffusers>=0.29.0; extra == "image"
deepcache 0.1.0 depends on diffusers 
matcha-tts 0.0.6.0 depends on diffusers==0.25.0
xinference[audio,embedding,image,rerank,transformers,vllm] 0+unknown depends on diffusers>=0.29.0; extra == "image"
deepcache 0.1.0 depends on diffusers
matcha-tts 0.0.5.1 depends on diffusers==0.25.0
shivammehta25 commented 3 months ago

Hello, I have bound the version because of reproducibility, in case someone wants to reproduce, they can go to the previous tag and use the source code from there. Feel free to remove it as long as it works for you.

shivammehta25 commented 3 months ago

Done! I uploaded it to pypy as well future: pip install -U matcha-tts should not include the fixed diffusion version of matcha tts.

PyPi: https://pypi.org/project/matcha-tts Binaries: https://github.com/shivammehta25/Matcha-TTS/releases/tag/v0.0.7 Commit: https://github.com/shivammehta25/Matcha-TTS/commit/77804265f877b0c42f13cfdece6541dde7838090