V-MPO Simple code to demonstrate Deep Reinforcement Learning by using an on-policy adaptation of Maximum a Posteriori Policy Optimization (MPO) in Pyt