Skip to content

Latest commit

 

History

History
31 lines (24 loc) · 558 Bytes

README.md

File metadata and controls

31 lines (24 loc) · 558 Bytes

Introduction

This is the implementation of Wasserstein Auto-Encoders paper in PyTorch. simple and clean WAE

I rewrite the model in DC-GAN for the encoder/decoder/discriminator instead of simple MLP

Requirement

  • python 3
  • PyTorch >= 0.4
  • torchvision
  • numpy
  • tqdm

Train

  • To train a WAE-GAN:
python wae_gan.py
  • To train a WAE-MMD:
python wae_mmd.py

Randomly Generated Images (100 Epochs)

WAE-MMD

WAE_MMD

WAE-GAN

WAE_GAN