Talk to GPT, and GPT responds with a very lifelike voice. Uses GPT output streaming Elevenlabs input + output streaming for fast response time.
- Make python virtual environment and install python dependencies
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
-
Try running
python3 main.py
. There may be some additional audio dependencies.brew install
them. -
Create a .env file, and put in your ElevenLabs and OpenAI API keys. See .env.example.
-
create stop condition (when conversation is over. otherwise bot will keep listening)
-
clean dependencies
-
optimize whisper streaming thru non-local model (api?)