This repository contains the following:
- A Vector-quantized variational auto-encoder implementation, fully type-hinted, flake8-checked.
- A scraper module to get images from FAFA
- A config file being its own Kubernetes manifest
- A helm chart containing the Kubernetes operator for the VQ-VAE implementation
- A test suite to validate various parts of the implementation
- A GitHub Actions workflow to run the test on GitHub
- A full lab journal containing descriptions of almost every run I did and what my thoughts and doubts were
Oof, that's a lot. It's because I spent a lot of my free time on this. Let's start with the really complicated part.
This subject deserves its own README. An elaborate description is in models/README.md