CI Report:
https://ci-tests.linuxserver.io/linuxserver/faster-whisper/v3.3.0-ls145/index.html
LinuxServer Changes:
Full Changelog: v3.2.0-ls144...v3.3.0-ls145
Remote Changes:
- Add FunASR speech-to-text backend (
--stt-library funasr) defaulting toFunAudioLLM/SenseVoiceSmall(@LauraGPT)- Non-autoregressive and notably faster than Whisper; supports English, Chinese, Cantonese, Japanese, and Korean well
- Install with the
funasrextra (pip install '.[funasr]')