Quick Start:
- Install requirements using
pip3 install -r requirements.txt - Edit
configuration.pyand add your API keys, etc. - Create a voice on the ElevenLabs TTS website and add the voice name to
configuration.py
Simple example of using ChatGPT API with Whisper API, and ElevenLabs TTS API.
Enter your API keys and other config in configuration.py
Run the script with python3 main.py or flask_app.py
combine_output will combine the output of the chat session from main.py into a single video file.
Every character has their own folder with animation.mp4, personality.txt and photo.jpg.
personality.txt is the system message/prompt for the chat session.
The first line of personality.txt has the name, formatted as Name: <name>.
animation.mp4 can be created by d-id automatically from photo.jpg when main.py is run or you can produce your own and place it in the folder.
Previously, d-id speaking avatar animations were used, but that's very expensive for this purpose. The code is there, but it is disabled.
This is a quick and dirty example. This code is provided as-is, and is not meant to be used in production. It is only for reference.
No support will be provided for this code, as it is in an unfinished state and the author is not actively working on it.