too long; didn't listen, generate voice message captions with whisper in telegram and discord
Clone the repository and download the latest version of Whisper-Faster-XXL, extract it into the bot directory so the whisper-faster-xxl executable is in ${BOT_DIR}/Whisper-Faster-XXL.
Run npm i in bot directory, copy .env.example to .env and fill out the required environment variables.
Start the bot with npm start
The Bot will create an audio transcript for every voice message it sees, this can be in a DM, Group or Server.
Voice messages are temporarily downloaded into /tmp/tldl/, this is also where transcripts are stored until they are read by the bot, the files are subsequently deleted.
see LICENSE
Test the Bot by inviting it on Discord or messaging it on Telegram