A Clojure port of XinJingHao’s PPO implementation using libpython-clj2, PyTorch, and Quil. PPO is a reinforcement learning method which has become popular because it addresses the problem of stability. The PPO implementation is tested using the inverted pendulum problem.