arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.12729cs.LGcs.ROstat.ML
收藏

深度策略梯度中的执行问题:PPO和TRPO的案例研究

Implementation Matters in Deep Policy Gradients: A Case Study on PPO and TRPO

MIT(麻省理工学院);Two Sigma(双西格玛)

Logan Engstrom, Andrew Ilyas, Shibani Santurkar, Dimitris Tsipras, Firdaus Janoos, Larry Rudolph, Aleksander Madry

摘要:

通过对两种常用的策略优化算法:近端策略优化(PPO)和信任域策略优化(TRPO)的案例研究,分析了深度策略梯度算法取得进步的根源。具体来说,我们调查的后果“代码级优化:”算法增强只在实现或描述为辅助细节的核心算法。这种优化似乎是次要的,但事实证明它对代理行为有重大影响。我们的研究结果表明,它们(a)是PPO在TRPO上获得累积奖励的主要原因,(b)从根本上改变了RL方法的功能。这些见解表明了在深度强化学习中归因于性能增益的困难和重要性。复制我们结果的代码可在https://github.com/MadryLab/implementation-matters上获得。

英文摘要:

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 .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑