Closed jurriaan closed 8 months ago
This is why I didn't include those models. Something is broken with faster-whisper or the conversation process, it seems.
English versions of the models are now available in 2.0.0: https://github.com/rhasspy/wyoming?tab=readme-ov-file#speech-to-text
I saw that the english-only versions of the models were available here: https://github.com/rhasspy/models/releases/tag/v1.0
But this add on does not support them (limited by the values in the FasterWhisperModel enum. I tried adding support for these models manually but don't get any working output.
I get these log messages with the
base.en
model, it seems that it does not correctly transcribe the audio: