thu-ml/tianshou

Extend and fix reward/return normalizations

Open

#927 创建于 2023年8月29日

在 GitHub 查看
 (0 评论) (0 反应) (0 负责人)Python (7,121 star) (1,072 fork)batch import
enhancementgood first issue

描述

There are some ways in which reward/return/value normalization could be improved. But one most drastic thing is the following:

Currently PGPolicy instantiates self.ret_rms = RunningMeanStd(), and RunningMeanStd has a default value of clip_max=10. This cannot be adjusted by users! (except through monkey-patching, ofc)

This might work well for some standard envs, but the clipping value is arbitrary and making it non-configurable is a major hinderance for users, who are most probably not aware of this.

Generally, how to best normalize stuff in RL is an active discussion and normalization can play an important role in performance. I believe tianshou should be extended to accomodate various normalization strategies

贡献者指南

技术栈
pythonpytorch
领域
machine learningai
议题类型
feature
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
3
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
1-2 days
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
fresh
清晰度议题是否清楚说明期望改动、验收标准和下一步。
mostly clear
前置要求
understanding of RL reward normalizationfamiliarity with tianshou codebase
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
40
研究方向
Investigate the current normalization implementation in PGPolicy, specifically the `RunningMeanStd` class with default `clip max=10` in `tianshou/policy/pg.py`. The goal is to make `clip max` configurable and possibly add support for alternative normalization strategies. Review the entire codebase for similar hardcoded normalization values and consider creating a unified configuration interface.