HTTPS SSH

README

This is a first attempt at building a GAN. It uses MNIST data for training. It deviates from a DCGAN by including Wasserstein loss (mse) and weight clipping to [-0.01, 0.01]