poke-vae implements a Variational Autoencoder (VAE) in PyTorch to encode and decode Pokémon statistics. The VAE is structured to learn a compressed representation of Pokémon data which includes base stats and types, enabling the generation of new Pokémon-like data.
python machine-learning pokemon deep-learning pytorch vae representation-learning unsupervised-learning variational-autoencoder content-generation latent-space generative-ai creative-ml
-
Updated
Feb 11, 2026 - Python