GPT Voice Assistant Notes
This repository contains notes for a Python program for using GPT from OpenAI as a voice assistant.
Plan
The plan is to create 5 different articles that tackle different sections of the program by functionality.
- Part 1 - Audio Channel
- Part 2 - AI Models and Services
- Part 3 - Web Server and Storage
- Part 4 - Flow and Conclusion
Part 1 will contain audio theory and practice for audio device, torchaudio, toml config file and audio calibration.
Part 2 will contain constructing classes for loading up local AI models, as well as connecting to remote AI services, plus loading SSML and a YAML file.
Part 3 will contain a simple SQL database, threading and a web socket server for serving HTML and updating messages in real time.
Part 4 will refine the application flow, setup commands, add extra functionality and suggest possible future improvements.