Simple API for WhisperX, long-attention solution to the original model.
https://github.com/m-bain/whisperX
pip install -r requirements.txt
Make sure you have GPU acceleration configured and turned on.
docker build --build-arg MODEL=large-v2 -t whisperx-api .
docker run -p 8000:8000 -t whisperx-api
/GET /status
Find out service is on and GPU is available
/POST /transcribe
:file: File to upload