The purpose of the WaveGan is to generate 1 second audio samples using a generative adversarial network. This implimentation is based on https://arxiv.org/abs/1802.04208 Adversarial Audio Synthesis paper by Chris Donahue, Julian McAuley, Miller Puckette.