https://github.com/bstee615/rllab-adv

Code to train RL agents along with Adversarial distrubance agents

https://github.com/bstee615/rllab-adv

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
  • .zenodo.json file
  • DOI references
  • Academic publication links
    Links to: arxiv.org
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (6.9%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

Code to train RL agents along with Adversarial distrubance agents

Basic Info
Statistics
  • Stars: 0
  • Watchers: 1
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Fork of lerrel/rllab-adv
Created over 5 years ago · Last pushed over 9 years ago

https://github.com/bstee615/rllab-adv/blob/master/

>Under Development
# Robust Adversarial Reinforcement Learning

This repo contains code for training RL agents with adversarial disturbance agents in our work on Robust Adversarial Reinforcement Learning ([RARL](https://arxiv.org/abs/1703.02702)). We build heavily build on the OpenAI rllab repo.

## Installation instructions

Since we build upon the [rllab](https://github.com/openai/rllab) package for the optimizers, the installation process is similar to `rllab's` manual installation. Most of the packages are virtually installated in the anaconda `rllab3-adv` enivronment.

- Dependencies for scipy:

```
sudo apt-get build-dep python-scipy
```

- Install python modules:

```
conda env create -f environment.yml
```

- [Install MuJoCo](https://github.com/openai/mujoco-py)

- Add `rllab-adv` to your `PYTHONPATH`.

```
export PYTHONPATH=:$PYTHONPATH
```

## Example

```python
# Enter the anaconda virtual environment
source activate rllab3-adv
# Train on InvertedPendulum
python adversarial/scripts/train_adversary.py --env InvertedPendulumAdv-v1 --folder ~/rllab-adv/results
```

## Contact
Lerrel Pinto -- lerrelpATcsDOTcmuDOTedu.

Owner

  • Name: Benjamin Steenhoek
  • Login: bstee615
  • Kind: user

3rd year PhD student @ ISU. Interests and research: deep learning, program analysis

GitHub Events

Total
Last Year