Multi-Objective Maximum a posteriori Policy Optimization (MO-MPO)

Demo

This repository implements MO-MPO algorithm, introduced in (Abdolmaleki, Huang et al., 2020), by using PyTorch.

First, download this repo to your local machine,

git clone https://github.com/ToooooooT/MOMPO.git

Next, change to the repo directory, and you need to install all the required Python packages,
(In case you would not like to build a local virtual environment, you can just skip the second line.)
```
cd MOMPO/
python -m venv .venv
pip install -r requirements.txt
```
To run MO-MPO in the "Deep Sea Treasure" environment,
```
python run_cat_mompo.py
```
To run MO-MPO in the "Humanoid-run" environment,
```
python run_cont_mompo.py
```

Name		Name	Last commit message	Last commit date
Latest commit History 181 Commits
agents		agents
envs		envs
models		models
static		static
utils		utils
.gitignore		.gitignore
README.md		README.md
requirements.txt		requirements.txt
run_cate_mompo.py		run_cate_mompo.py
run_cont_mompo.py		run_cont_mompo.py