Play OpenAI's Lunar Lander with Policy Gradients

Code trains a neural network to play lunar lander iwth reinforcement learning
Post explaining code at https://codeodysseys.com/posts/lunar-lander-pg/