forked from TheCoffeMaker/WhisperBot
XMPP bot that transliterates audio messages using OpenAI's Whisper libraries
- Python 93.5%
- Dockerfile 6.5%
| .gitignore |
revert e231997864
|
|
| .pre-commit-config.yaml | Added pre-commit | |
| about.txt | WhisperBot kickoff | |
| bot.env | Added WHISPER_MODEL to bot.env | |
| docker-compose.yml | fix docker image name | |
| Dockerfile | Bump | |
| LICENSE | Initial commit | |
| quotes.txt | WhisperBot kickoff | |
| README.md | Other ways to run it | |
| requirements.txt | WhisperBot kickoff | |
| whisperbot.py | Add an event handler of group chat message | |
WhisperBot
XMPP bot that transliterates audio messages using OpenAI's Whisper libraries
Running Whisperbot
As KISS as it can get
git clone https://codeberg.org/TheCoffeMaker/WhisperBot.git
cd WhisperBot
docker build -t whisperbot .
docker docker run -e JID="bot_acct@your.xmpp.server" \
-e PASS="a_very_secure_password" \
-e ROOM="that_room_with_audio_messages@muc.at.some.server" \
-e NICK="Bot_Nick" \
-v /host/volume/path/:/root/.cache/ \
--name whisperbot \
whisperbot:latest
Or with docker-compose:
git clone https://codeberg.org/TheCoffeMaker/WhisperBot.git
cd Whisperbot
vi bot.env # Edit bot parameters
docker-compose up
Or directly downloading from DockerHub:
docker docker run -e JID="bot_acct@your.xmpp.server" \
-e PASS="a_very_secure_password" \
-e ROOM="that_room_with_audio_messages@muc.at.some.server" \
-e NICK="Bot_Nick" \
-v /host/volume/path/:/root/.cache/ \
--name whisperbot \
thecoffemaker/whisperbot:latest
Container's volume can be omited, but will download the Whisper's model each time it runs. By using volume we avoid to download the model each time the container get restarted.
You can see it working or reach us at Cyberdelia's Lost in Cyberspace channel.