Whisper API for inference

This codebase provides code to generate transcript and subtitles from audio input from a FastAPI API

WARNING: This codebase assumes that you have access to an NVIDIA GPU (CUDA 12.1) as well a DeepL API key

Requirements

This project requires Python 3.11 to run as well as Docker.

You will also need to the python packages in requirements.txt using pip.

pip install -r requirements.txt

You will need to have the following as environnment variables:

You can set them in the compose.yaml file.

docker-compose -f compose.yaml up

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
static		static
.dockerignore		.dockerignore
Dockerfile		Dockerfile
README.md		README.md
app.py		app.py
app_utils.py		app_utils.py
compose.yaml		compose.yaml
requirements.txt		requirements.txt
subtitle_utils.py		subtitle_utils.py
transcript.py		transcript.py