An end to end model that can synthesize images from a given text using GAN (Generative Adversarial Networks). There are currently two GAN architectures in this repository, and both work on the MNSIT dataset.

GAN_fc.py, This is a conditional GAN that can generate the desired hand-written digit from 0 to 9.

GAN_vanilla.py, This is a vanilla GAN that can a distribution of hand-written digits almost identical to that of the MNSIT dataset