This repository contains some experimental implementations of the Variational Autoencoder.
conda env create -f environment.yml# Train a VAE on mnist using fully connected neural networks as encoder/decoder
python vae.py
# Train a VAE on mnist using convolutional neural networks as encoder/decoder
python vae.py with cnn
# Train a conditional VAE
python vae.py with conditional
# Train a convolutional VAE on mnist with batch size 128
python vae.py with cnn "batch_size=128"