/basic_training

A barebones VAE training loop as a jumping off point for other experiments

Primary LanguagePython

A Toy Problem

A simple repo containing a barebones implementation of a VAE (modified from AntixK's) and a training/logging/validation loop for it. My intention is that this repo will make it so I always have a ready-to-go implementation of toy problem on hand so I can test out weird training procedures, optimization ideas, or architecture tweaks.