Training Generative Adversarial Networks
     by Solving Ordinary Differential Equations

   Chongli Qin*, Yan Wu*, Jost Tobias Springenberg, Andrew Brock, Jeff Donahue,
             Timothy P. Lillicrap, Pushmeet Kohli
                   DeepMind
             chongliqin, yanwu@google.com

                       Abstract
     The instability of Generative Adversarial Network (GAN) training has frequently
     been attributed to gradient descent. Consequently, recent methods have aimed
     to ta ...