#
drl-algorithms
Here are 2 public repositories matching this topic...
本项目是作者（MRL Liu）使用AI算法的强化学习方法玩迷宫游戏的一个阶段性总结，本项目的迷宫游戏是简单的方格迷宫，其状态空间和动作空间都足够简单，是作者整理的手中的第1个RL项目。该项目重构了作者之前学习时的一些基于Value的RL算法，将它们的例如经验回放池的对象等抽象出来为一个对象，便于整理知识网络。该项目的原始算法代码使用的是莫烦Python的相关实现，在此向莫烦老师表示感谢。本项目的特色是使用了统一范式的代码来定义基于Value的算法系列的实现，封装了Q-Table和ReplayBuffer对象；添加了网络模型的保存与加载功能、TensorFlow可视化功能、经验池保存和加载等。整个项目基于良好的面向对象思想，方法定义层层推进。
-
Updated
May 2, 2021 - Python
Improve this page
Add a description, image, and links to the drl-algorithms topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the drl-algorithms topic, visit your repo's landing page and select "manage topics."
Hi,
I am currently using your FinRL_PortfolioAllocation_NeurIPS_2020 code and I have some strange behavior at the beginning of training. Sometimes the first episode reward mean value is super high and then drops during the training as shown on the tensorboard plot. This high value is never reached again. Any idea why this is happening ?
Edit: I'm training PPO agent from stablebaselines3 wit