Purfview
Purfview
>Unfortunately --compute_type auto did not make difference It should've some difference, as it frees ~1.6 GB RAM, when your error is about 429 MB allocation. Did you test exact same...
Maybe he meant why that "." is separated from "J"?
>yes, "JD Vance" is a name No, that's not what I meant.
Could you try whisper-faster.exe from https://github.com/Purfview/whisper-standalone-win on audio where you've issue with "stuck in a failure loop"?
What GPU do you have? Check cuDNN version. Try `tiny` model.
> However, when silent audio data (not just speech) was input, an issue occurred where the inferred text was output as the words entered as hotwords(진료 모드). That's expected hallucination...
>Is this an issue needed to be fixed? It's just a warning, nothing to worry about, produced because torch doesn't support flash attention on Windows. I'll hide it in the...
Should be fixed in Pro r3.256.1
>But there are several problems with it.. IMO, that's expected behaviour. > I'm not 100% what it was designed to do.. It's to cut audio to chunks. >Do we even...
>Not really, the audio is cut into I meant it as a whole, whole point of it is to get chunks. That's why you can't disable VAD in batched. >So...