LiveVTT is a tool for live transcription of streaming audio/video content, providing real-time subtitles in WebVTT format.

LiveVTT allows you to transcribe live audio/video streams and generate WebVTT subtitles. It supports various features such as model selection, CUDA utilization, silence filtering, and more.

livevtt -u <URL> [-s] [-l <BIND_ADDRESS>] [-p <BIND_PORT>] [-m <MODEL>] [-b <BEAM_SIZE>] [-c <USE_CUDA>] [-t <TRANSLATE>] [-vf <VAD_FILTER>] [-la <LANGUAGE>] [-ua <USER_AGENT>]

• -u, --url: [Required] URL of the live audio/video stream.
• -s, --hard-subs: Set if you want the subtitles to be baked into the stream itself.
• -l, --bind-address: The IP address to bind to (defaults to 127.0.0.1).
• -p, --bind-port: The port to bind to (defaults to 8000).
• -m, --model: Whisper model to use (defaults to large).
• -b, --beam-size: Beam size to use (defaults to 5).
• -c, --use-cuda: Use CUDA where available. Defaults to true.
• -t, --transcribe: If set, transcribes rather than translates the given stream.
• -vf, --vad-filter: Whether to utilize the Silero VAD model to try and filter out silences. Defaults to false.
• -la, --language: The original language of the stream, if known/not multilingual. Can be left unset.
• -ua, --user-agent: User agent to use to retrieve playlists/stream chunks (defaults to 'VLC/3.0.18 LibVLC/3.0.18').

Once the program is running, you can access the transcribed and/or translated stream at the following URL:

http://127.0.0.1:8000/playlist.m3u8

This URL may vary based on the bind address and port provided via the command-line options.

Note that the minimum target Python version for this script is Python 3.10 at present. You will also need to ensure that you have the ffmpeg package installed on your system.

1. Clone the repository:

   git clone https://github.com/Psychotropos/livevtt.git

2. Navigate to the directory:

   cd livevtt

3. Install dependencies:

   • For general installation:

   pip install -r requirements.txt

1. Transcribe a live audio/video stream with default settings:

   livevtt -u <URL>

2. Transcribe a live audio/video stream and embed subtitles:

   livevtt -u <URL> -s

Contributions are welcome! Please fork the repository and submit a pull request.

This project is licensed under the MIT License - see the LICENSE file for details.