Implements of Wasserstein Divergence for GANs

eriklindernoren / PyTorch-GAN

PyTorch implementations of Generative Adversarial Networks.

MIT License

16.41k stars 4.07k forks source link

Implements of Wasserstein Divergence for GANs #28

Closed Pinnh closed 5 years ago

Pinnh commented 5 years ago

Hi， I implemented the Wasserstein Divergence for GANs(wgan_div) and tested . the performance of wgan_div on face super-resolution and mnist is as same as wgan_gp .

eriklindernoren commented 5 years ago

That's awesome. Do you have some image samples we can add to the README?

Pinnh commented 5 years ago

Sure. In order to be consistent with previous samples, mnist image samples have committed to the assets.

eriklindernoren commented 5 years ago

If you have samples from super-resolution that would be even better. Would like to have more diverse (not only MNIST) samples in the README.

eriklindernoren commented 5 years ago

Cool, thanks! Merging