Implementation Matters in Deep Policy Gradients: A Case Study on PPO and TRPO
This work highlights the difficulty and importance of attributing performance gains in deep reinforcement learning, which is crucial for researchers and practitioners in the field.
The study investigated the impact of code-level optimizations in deep policy gradient algorithms, specifically PPO and TRPO, finding that these optimizations are responsible for most of PPO's gain in cumulative reward over TRPO and fundamentally change how RL methods function.
We study the roots of algorithmic progress in deep policy gradient algorithms through a case study on two popular algorithms: Proximal Policy Optimization (PPO) and Trust Region Policy Optimization (TRPO). Specifically, we investigate the consequences of "code-level optimizations:" algorithm augmentations found only in implementations or described as auxiliary details to the core algorithm. Seemingly of secondary importance, such optimizations turn out to have a major impact on agent behavior. Our results show that they (a) are responsible for most of PPO's gain in cumulative reward over TRPO, and (b) fundamentally change how RL methods function. These insights show the difficulty and importance of attributing performance gains in deep reinforcement learning. Code for reproducing our results is available at https://github.com/MadryLab/implementation-matters .