martinarjovsky / WassersteinGAN

BSD 3-Clause "New" or "Revised" License
3.2k stars 725 forks source link

Wasserstein GAN

Code accompanying the paper "Wasserstein GAN"

A few notes

Prerequisites

Two main empirical claims:

Generator sample quality correlates with discriminator loss

gensample

Improved model stability

stability

Reproducing LSUN experiments

With DCGAN:

python main.py --dataset lsun --dataroot [lsun-train-folder] --cuda

With MLP:

python main.py --mlp_G --ngf 512

Generated samples will be in the samples folder.

If you plot the value -Loss_D, then you can reproduce the curves from the paper. The curves from the paper (as mentioned in the paper) have a median filter applied to them:

med_filtered_loss = scipy.signal.medfilt(-Loss_D, dtype='float64'), 101)

More improved README in the works.