A simple dashboard for sending translations to Whisper (Open source speech recognition model from OpenAI).

You can use this dashboard to create user accounts, upload audio files, and download the results of the speech recognition model.

This dashboard uses Node.js and Express for the backend with a React frontend. Whisper's Python API is used to run the speech recognition model.

Login screen	Transcriptions list
Transcription result	Admin panel

You can install this dashboard with Docker or by running the backend and frontend separately.

The first time the dashboard is started, an admin user will be created, with username "admin" and password "admin". You should change this user's password on the profile page.

As admin, you can create or delete users, manage user roles (user or admin), and view and manage transcriptions sent by all users.

You can configure the dashboard with the following environment variables:

The variable CORS_ORIGIN is useful for development with the frontend running on http://localhost:5173 while the backend runs on http://localhost:3000.

You can build the Docker image with:

docker build -t whisper-dashboard .

Or you can pull the image from GitHub Container Registry:

docker pull ghcr.io/gauthier-th/whisper-dashboard:latest

Then run the container with:

docker run -d \
  -p 3000:3000 \
  -e JWT_SECRET=secret \
  -e MAX_PARALLEL_TRANSCRIPTIONS=1 \
  -e WHISPER_MODEL=tiny \
  -v /path/to/data:/config \
  ghcr.io/gauthier-th/whisper-dashboard:latest

You can also use this Docker Compose to run the dashboard:

version: "3.8"
services:
    whisper-dashboard:
        ports:
            - '3000:3000'
        environment:
            - JWT_SECRET=secret
            - MAX_PARALLEL_TRANSCRIPTIONS=1
            - WHISPER_MODEL=tiny
        volumes:
            - '/path/to/data:/config'
        image: 'ghcr.io/gauthier-th/whisper-dashboard:latest'

• Node.js 18+
• Python 3.8+

1. Clone the repository
2. Install the dependencies for the backend:

   cd api
   npm install
   pip install -U openai-whisper

3. Copy the .env.example file to .env and fill in the values from the table above. You can leave the default values.
4. Install the dependencies for the frontend:

   cd app
   npm install

5. Copy the .env.example file to .env and fill in this value:

• VITE_API_URL: The URL of the backend API (e.g. http://localhost:3000 for development)

You should run the backend and frontend in separate terminals. The backend will run on port 3000 by default, and the frontend will run on port 5173 by default. You can use the default values of the .env.example files to get started.

1. Start the backend:

   cd api
   npm run dev

2. Start the frontend:

   cd app
   npm run dev

1. Build the frontend:

   cd app
   npm run build

2. Start the backend:

   cd api
   npm start

3. The frontend will be served from the backend at the / path.

Pull requests are welcome! This project is still in early development, so there are many features that could be added. If you have any questions, feel free to open an issue or email me at [email protected].

This project is licensed under the MIT License.