Cannot install NISQA fastapi because of old librosa output

lukaszliniewicz / Pandrator

Pandrator aspires to be a user-friendly app with a graphical interface and a one-click installer that creates high-quality speech from text in multiple languages (audiobooks, speech synchronised with subtitles and more) using local models (XTTS, Silero or VoiceCraft), plus voice cloning, LLM pre-processing, RVC enhancement, and automatic evaluation

GNU Affero General Public License v3.0

112 stars 10 forks source link

Cannot install NISQA fastapi because of old librosa output #11

Open yodatak opened 2 months ago

yodatak commented 2 months ago

Hi thanks again NISQA is unmaintend for long time and this PR could be interesting to help to install on modern python https://github.com/gabrielmittag/NISQA/pull/47

Like https://librosa.org/doc/latest/changelog.html

Librosa output is removed since 2020

i got this error https://github.com/librosa/librosa/issues/774

Did we want to make container of this fastapi ?

yodatak commented 2 months ago

Maybe using something like LocalAI https://github.com/mudler/LocalAI?tab=readme-ov-file

lukaszliniewicz commented 2 months ago

I made a very simple FastAPI implementation of NISQA: https://github.com/lukaszliniewicz/NISQA-API. It should work if you follow the conda instructions. I'm still not sure how useful it is to improve results, though.