An autoencoder train on the Flickr-Faces-HQ Dataset with 16 latent dimensions for 1000 epochs. Note: The images trained on were 128x128.
