Arche151

Whisper itself can't stream, so I dont think so, unfortunately

Esapanso now works for me! I had to install it via snap (the 10th of july version) by running `sudo snap install espanso --edge --classic` Then I had to run...

@trungkienbkhn okay, thanks a lot for the information and required commands :) I will give it a go! Do you maybe also know whether I need a lot of compute...

@trungkienbkhn So, I converted the model and used float16 quantization and the quality of the transcription compared to the original large-v3 is really bad :( A lot of words are...

@trungkienbkhn I did that already, so I don't think, that's the issue, unfortunately.

> @Arche151 , could you try again with compute_type="default" (or remove this command when initializing whisper model) ? Thanks for the quick reply and suggestion! I'll try that and report...

> > Large-v3 model hallucinates, large-v2 doesn't > > It's known that large-v2 hallucinates much more that large-v2, read there: [Whisper-v3 Hallucinations on Real World Data](https://deepgram.com/learn/whisper-v3-results) Damn, that sucks hard....

> > > Then I guess I'll stay with large-v3. > > > > > > Did you meant "large-v2"? > > On my Standalone Faster-Whisper I've added auto-offsets to...

@trungkienbkhn Thank you so much for the info and the comparisons! Now, I only have to wait for distil-whisper to support large-v3 haha

Same issue here. Seems like the plugin isn't maintained anymore though, so idk, if it'll be fixed :(