强化学习算法：PPO and TRPO算法实现细节 —— Implementation Matters in Deep RL: A Case Study on PPO and TRPO

相关：

https://vitalab.github.io/article/2020/01/14/Implementation_Matters.html

论文地址：

https://openreview.net/pdf?id=r1etN1rtPB

Highlights

Much of the performance of PPO over TRPO comes from code-level optimization and not the original paper’s main selling points
PPO code-optimizations are significantly more important in terms of final reward achieved than the choice of general training algorithm (TRPO vs. PPO)