Support for whisper.cpp stream capabilities for realtime transcription.

[blundercode]

What is the feasibility of integrating whisper.cpp or more importantly their real-time capabilities?

It says they have docker support and was wondering if this is something you could implement. If your docker supported real-time it would be the best all-around whisper implementation I have seen by far.

https://github.com/ggerganov/whisper.cpp
https://github.com/ggerganov/whisper.cpp/tree/master/examples/stream

It does look to use CPU mode only which is fine. Thoughts on the feasibility of implementation? If it's possible I would be willing to donate some money for your time and effort as my developer skills don't translate super well to this type of thing.

Love the repo and all your hard work!