Skip to content
A library providing various existing GANs in PyTorch.
Jupyter Notebook Python
Branch: master
Clone or download
Latest commit 2e8513c Apr 10, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
examples Add the param to train Apr 9, 2019
notebooks Renamed notebooks Jan 30, 2019
vegans
.gitignore Update gitignore for iml Mar 29, 2019
CONTRIBUTING.md
LICENSE Update license in Unit8 name Jan 24, 2019
README.md Update README Apr 2, 2019
requirements.txt GPU updates and testing (#5) Mar 28, 2019
setup.py Include more information in README Jan 30, 2019

README.md

VeGANs

A library to easily train various existing GANs (Generative Adversarial Networks) in PyTorch.

This library targets mainly GAN users, who want to use existing GAN training techniques with their own generators/discriminators. However researchers may also find the GAN base class useful for quicker implementation of new GAN training techniques.

The focus is on simplicity and providing reasonable defaults.

How to install

You need python 3.5 or above. Then: pip install vegans

How to use

The basic idea is that the user provides discriminator and generator networks, and the library takes care of training them in a selected GAN setting:

from vegans import WGAN
from vegans.utils import plot_losses, plot_image_samples

netD = ### Your discriminator/critic (torch.nn.Module)
netG = ### Your generator (torch.nn.Module)
dataloader = ### Your dataloader (torch.utils.data.DataLoader)

# Build a Wasserstein GAN
gan = WGAN(netG, netD, dataloader, nr_epochs=20)

# train it
gan.train()

# vizualise results
img_list, D_losses, G_losses = gan.get_training_results()
plot_losses(G_losses, D_losses)
plot_image_samples(img_list, 50)

You can currently use the following GANs:

Slightly More Details:

All of these GAN objects inherit from a GAN base class. When building any such GAN, you must give in argument a generator and discriminator networks (some torch.nn.Module), as well as a torch.utils.data.DataLoader. In addition, you can specify some parameters supported by all GAN implementations:

  • optimizer_D and optimizer_G: some PyTorch optimizers (from torch.optim) for the discriminator and generator networks. By defaults those are set with default optimization parameters suggested in the original papers.
  • nr_epochs: the number of epochs (default: 5)
  • nz: size of the noise vector (input of the generator) - by default nz=100.
  • save_every: VeGANs will store some samples produced by the generator every save_every iteration. Default: 500
  • fixed_noise_size: The number of samples to save (from fixed noise vectors)
  • print_every: The number of iterations between printing training progress. Default: 50

Finally, when calling train() you can specify some parameters specific to each GAN. For example, for the Wasserstein GAN we can do:

gan = WGAN(netG, netD, dataloader)
gan.train(clip_value=0.1)

This will train a Wasserstein GAN with clipping values of 0.1 (instead of the default 0.01).

If you are researching new GAN training algorithms, you may find it useful to inherit from the GAN base class.

Learn more:

Currently the best way to learn more about how to use VeGANs is to have a look at the example notebooks. You can start with this simple example showing how to sample from a univariate Gaussian using a GAN. Alternatively, can run example scripts.

Contribute

PRs and suggestions are welcome. Look here for more details on the setup.

Credits

Some of the code has been inspired by some existing GAN implementations:

You can’t perform that action at this time.