# DL_RL_Zoo

**Repository Path**: publiagent123/DL_RL_Zoo

## Basic Information

- **Project Name**: DL_RL_Zoo
- **Description**: Lightweight, stable, efficient PyTorch implement of reinforcement learning
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-04-14
- **Last Updated**: 2024-06-05

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# LightWeight_Stable_ReinfLearning
Lightweight, stable, efficient PyTorch implement of reinforcement learning

# run
`python3 DelayDDPG.py`

All code is written in ↑ this file

# GIF

You can see these gif/png ↓ in file "Result_GIF".

BipedalWalkerHardcore-V2-total:

![BipedalWalkerHardcore-v2-total](https://github.com/Yonv1943/LightWeight_Stable_ReinfLearning/blob/master/Result_GIF/BipedalWalkerHardcore-v2-total.gif)
 
LunarLanderContinuous-V2:
 
![LunarLanderTwinDelay3](https://github.com/Yonv1943/LightWeight_Stable_ReinfLearning/blob/master/Result_GIF/LunarLanderTwinDelay3.gif)
  
Plot LunarLanderContinuous-V2, TrainEpoch: 78, TimeUsed: 996s:

![DelayDDPG_plot](https://github.com/Yonv1943/LightWeight_Stable_ReinfLearning/blob/master/Result_GIF/DelayDDPG_plot.png)

# More
If you can understand Chinese, more details of DelayDDPG are described in Chinese in this website ↓
 
[Enhanced Learning DelayDDPG, LunarLander, BipedalWalker, only need to train half an hour of lightweight, stable code](https://zhuanlan.zhihu.com/p/72586697)

如果你能看得懂中文，那么我用中文写了对这个算法的详细介绍:
  
[强化学习DelayDDPG，月球着陆器，双足机器人，只需训练半个小时的轻量、稳定代码](https://zhuanlan.zhihu.com/p/72586697)