A transcription service (and GUI client) to use Faster-Whisper-XXL as a service

[fwaris]

Wrote a small utility service to transcribe and diarize video (.mp4) files via Faster-Whisper-XXL.

The service is meant to run on a GPU-enabled box. A small GPU box is enough (e.g. a T1 instance on Azure).

Transcription 'jobs' can be submitted via a GUI client app. The app uploads local .mp4 files; triggers transcription; and downloads resulting text (.vtt) files containing the transcriptions. SSH connectivity is used for secure data exchange.

The app enables non-technical users to easily and securely use Faster-Whisper-XXL as-a-service.

Note: GPU infrastructure is meant to be shareable by multiple users. The jobs are executed serially so only one job runs at a time to keep the compute requirements low.

https://github.com/fwaris/TranscriptionAndDiarization