[Request/Assistance] Amendments to utilise a local network based faster whisper server

[kputera27]

Hi, thanks for the awesome work.

I was wondering if you could help me out with my use case as I am fairly new to the coding/GitHub world.

I am currently running a Faster Whisper Server on my secondary PC using Docker. I have made dictation work successfully using SpeechPulse on my primary PC that points to the aforementioned Docker container on my secondary PC. I was wondering how I could amend any of the .py files in your project so that I can point it to using my faster whisper server instead of Open AI servers?

Not sure if this is possible. Any help would be greatly appreciated!

Thanks!

[dhruvyad]

Hey! If you already know how to access your remote server, then you can create a new transcriber class that uses your server. It should be quite small ~15-20 lines of code. Essentially you just have to create a new class that overrides the AudioTranscriber class and override the transcribe_audio function that accepts audio in the form of an io.BytesIO object and returns its transcription as a string.

Once you're done, simply change the two lines - transcriber used in main.py and its corresponding import.

Let me know if you have any questions.

[gsoul]

I just created a PR to support this: https://github.com/dhruvyad/uttertype/pull/6

Please take a look