Policy Gradient (PG) An implementation of a vanilla policy gradient algorithm for TensorFlow and PyTorch. Detailed Documentation Implementation