Purfview
Purfview
> whisper-v3 is worse than whisper-v2 when using faster-whisper 1) There is no such thing as "whisper-v3" nor "whisper-v2". 2) This doesn't make sense -> "whisper-v3 is worse than whisper-v2"...
Oh that. Make sure that you are using the same settings. Post what parameters you run.
Did you run same for Whisper?
And what results do you get from Whisper? Can you share the audio sample with the issue?
> Today, I did test again- for whisper-v3 with fast-whisper interface. For the same audio, it was recognized ten times and the correct result was obtained every time. Then-v3 maybe...
> Unfortunately, even OpenAI incorrectly call it [whisper v3](https://openai.com/blog/new-models-and-developer-products-announced-at-devday#whisper-v3-and-consistency-decoder) in their news reports. Ha, that's where it's coming from, I was thinking what is going on... I guess it's some...
Use smaller model or use `device="cpu"`. > can't you download the model locally and use it without the need for it all to be loaded into memory? You can't.
Distil models are English only, you need to use a multilanguage model.
Try to test that file locally.
Where is the audio sample to reproduce the issue?