ahk-d / transcribe-video-audio

An OpenAI's Whisper-based full-stack project to transcribe audio and video files using React & Django.
36 stars 13 forks source link
django django-channels openai openai-whisper react tailwindcss typescript

Video/Audio Transcriber

An OpenAI's Whisper-based full-stack project to transcribe audio and video files using React & Django.

Demo

ezgif com-gif-maker

Installation

You need the latest versions of pipenv and node to setup this project.

1- To install the server's dependencies, in the terminal run the following:

  cd api
  pipenv install
  pipenv shell 
  pipenv run python manage.py makemigrations
  pipenv run python manage.py migrate
  pipenv run server

2- To run the client, execute the following commands:

  cd client
  yarn install
  touch .env.local
  yarn dev

You need to have the following env variables in your .env.local file:

NEXT_PUBLIC_ROOT_URL=http://127.0.0.1:8000
NEXT_PUBLIC_ALL_FILES_URL=http://127.0.0.1:8000/api/transcription/files/
NEXT_PUBLIC_FILE_UPLOAD_URL=http://127.0.0.1:8000/api/transcription/files/
NEXT_PUBLIC_TRANSCRIPTION_URL=ws://127.0.0.1:8000/ws/transcribe

Adjust these if your going to deploy to any remote server.

3- Select a video or audio file to upload The transcripts will be extracted and displayed on the page.

4- You can also view all your uploaded video and audio files and see their transcriptions. Their is an additional export button that allows you to download a JSON file with all their information.

License

MIT

Contributing

Contributions are always welcome!

Open a PR or an issue for extra features and bug fixes.

Authors