GUI implementation for faster-whisper

[JuergenFleiss]

Hi,

we have been using faster-whisper at our university to transcribe interviews and finally ended up building a GUI to enable researchers to transcribe interviews with speaker diarization locally on their computer.

https://github.com/BANDAS-Center/aTrain

Feel free to have a look, any feedback is welcome. Faster-whisper was a real asset, as it allowed us to reduce transcription time on CPU, so thanks for all the great work.

Best,
Jürgen

[dgoryeo]

Nicely done @Juergen-J-F (and BANDAS)!!!
I'm just trying to check it out --I used windows app installer.
Quick question: is there any utility/package that I could use to convert the json file to an srt or vtt file?
cheers

[JuergenFleiss]

Not yet, @dgoryeo, but it's on the list; there is already an issue open on our github repository. Best, Jürgen

[dgoryeo]

Thanks @Juergen-J-F . Am I right to think that the json file has the same format as WhisperX? If, so can one use WhisperV write_output to produce srt, vtt?

[JuergenFleiss]

@dgoryeo we now added srt as output format in the latest release.
https://github.com/BANDAS-Center/aTrain/releases/tag/v1.1.0

[dgoryeo]

Thanks @Juergen-J-F !

I hope someone can continue to take the helm on faster-whisper since @guillaumekln is now tied up with his new job.

[JuergenFleiss]

Hi you all,

we just released v1.2 of our standalone transcription app integrating distilled models and v3-turbo support as well as GUI model downloader and manager. Feedback is always welcome.

https://github.com/JuergenFleiss/aTrain/releases/tag/v1.2.0