PaddlePaddle / PARL

A high-performance distributed training framework for Reinforcement Learning
https://parl.readthedocs.io/
Apache License 2.0
3.28k stars 818 forks source link

PARL能否增加D3QN 和discretePPO算法? #721

Closed yglpyn8888 closed 2 years ago

yglpyn8888 commented 3 years ago

目前的场景动作空间为离散,所以需要上述两算法来跑,,看了PARL好像还没实现这两个? 能不能安排版本实现呢?谢谢!

TomorrowIsAnOtherDay commented 3 years ago

谢谢你的反馈,我们近期的计划是提供离线强化学习相关的算法,关于你的建议,我们内部讨论下再决定的:)