Implementation of my proposed hybrid algorithm for deep reinforcement learning. Combines the gradient-based PPO algorithm with a gradient-free genetic algorithm