Based on https://github.com/ckaytev/tgisper
Using Google Speech Recognition and OpenAI Whisper model.
- `BOT_ID`: Your Telegram bot token
- `ALLOWED_USERS`: Comma-separated list of allowed user IDs
- `PRECISION` (optional): Precision level for the ASR model (default: `int8`)
- `ASR_MODEL` (optional): Name of the ASR model to use (default: `small`)
- `MODE`: ASR mode to use (`whisper` or `google`)
Update variables in docker-compose.yml and run it