Skip to content

Commit d6bcada

Browse files
committed
[workflows]: add MAPPO/HAPPO
1 parent 4f089a5 commit d6bcada

File tree

2 files changed

+31
-13
lines changed

2 files changed

+31
-13
lines changed

.github/CHANGELOG.md

Lines changed: 8 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,14 @@
11
## 版本变更
22
### 新功能
3-
- 🚀 为 MATD3 添加了训练好的模型,可无需训练直接下载使用.
3+
- 🚀 为 MATD3 添加了训练好的模型,可无需训练直接下载使用
4+
- 🌟 新增 HAPPO-MAPPO_Continous_Heterogeneous 算法,支持异质智能体训练
5+
- ✨ 新增 MAPPO_Continous_Homogeneous 算法,优化同质智能体训练效率
6+
- 📦 将新算法添加到自动化打包发布流程中
47

58
### 修复
6-
- 🐛 无.
9+
- 🐛 无
710

811
### 文档
9-
- 📝 添加了 MATD3 追逃环境效果: gif.
10-
- 🌐 更新了文档 README.md, README_en.md.
12+
- 📝 添加了 MATD3 追逃环境效果: gif
13+
- 📖 添加了新算法的使用文档和示例
14+
- 🌐 更新了文档 README.md, README_en.md 以包含新算法介绍

.github/workflows/pack-folders.yml

Lines changed: 23 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -38,16 +38,22 @@ jobs:
3838
# 多模块独立发布包
3939
4040
## 模块介绍
41-
- **MADDPG_Continous**
41+
- **MADDPG_Continous**
4242
多智能体深度确定性策略梯度算法(连续动作空间版本),适用于连续控制场景的多智能体协同训练。
4343
44-
- ⚡️ **MATD3_Continous**
44+
- **MATD3_Continous**
4545
多智能体双延迟深度确定性策略梯度算法,在MADDPG基础上增加了延迟更新和策略平滑机制。
4646
47-
- 📚 **RL_Learning-main**
47+
- **HAPPO-MAPPO_Continous_Heterogeneous**
48+
异质智能体近端策略优化算法,支持不同类型智能体的混合训练场景。
49+
50+
- **MAPPO_Continous_Homogeneous**
51+
同质智能体近端策略优化算法,针对同类型智能体的高效训练优化。
52+
53+
- **RL_Learning-main**
4854
赵世钰老师强化学习基础教程合集,包含经典算法实现和示例代码。
4955
50-
- **hands_on_RL**
56+
- **hands_on_RL**
5157
动手学强化学习实践项目,通过Jupyter Notebook提供互动式学习体验。
5258
5359
## 使用说明
@@ -67,13 +73,19 @@ jobs:
6773
- **MADDPG_Continous**
6874
Multi-Agent Deep Deterministic Policy Gradient (continuous action space version) for cooperative multi-agent control.
6975
70-
- **MATD3_Continous**
76+
- **MATD3_Continous**
7177
Multi-Agent Twin Delayed DDPG, featuring delayed updates and policy smoothing.
7278
73-
- 📚 **RL_Learning-main**
79+
- **HAPPO-MAPPO_Continous_Heterogeneous**
80+
Heterogeneous Agent Proximal Policy Optimization for mixed-type agent scenarios.
81+
82+
- **MAPPO_Continous_Homogeneous**
83+
Homogeneous Agent Proximal Policy Optimization for efficient training of same-type agents.
84+
85+
- **RL_Learning-main**
7486
Fundamental RL tutorials with classic algorithm implementations.
7587
76-
- 🧠 **hands_on_RL**
88+
- **hands_on_RL**
7789
Interactive reinforcement learning projects via Jupyter Notebooks.
7890
7991
## Quick Start
@@ -131,6 +143,8 @@ jobs:
131143
TARGET_FOLDERS=(
132144
"MADDPG_Continous"
133145
"MATD3_Continous"
146+
"HAPPO-MAPPO_Continous_Heterogeneous"
147+
"MAPPO_Continous_Homogeneous"
134148
"RL_Learning-main"
135149
"动手学强化学习" # 保留原始中文名称
136150
)
@@ -155,9 +169,9 @@ jobs:
155169
156170
# 打包为英文名称的ZIP
157171
(cd "temp_$output_name" && zip -r "../zips/$output_name.zip" .)
158-
echo "Packaged: $folder as $output_name.zip"
172+
echo "Packaged: $folder as $output_name.zip"
159173
else
160-
echo "⚠️ Folder not found: $folder"
174+
echo "Folder not found: $folder"
161175
fi
162176
done
163177

0 commit comments

Comments
 (0)