Return to Article Details GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium Download Download PDF