Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ankur-deka/emergent-multiagent-strategies
Emergence of complex strategies through multiagent competition
https://github.com/ankur-deka/emergent-multiagent-strategies
competitive marl multiagent-reinforcement-learning multiagent-systems robotics
Last synced: 2 months ago
JSON representation
Emergence of complex strategies through multiagent competition
- Host: GitHub
- URL: https://github.com/ankur-deka/emergent-multiagent-strategies
- Owner: Ankur-Deka
- License: mit
- Created: 2020-07-05T00:27:14.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2022-12-08T06:21:31.000Z (about 2 years ago)
- Last Synced: 2024-01-28T21:40:53.659Z (12 months ago)
- Topics: competitive, marl, multiagent-reinforcement-learning, multiagent-systems, robotics
- Language: Python
- Homepage:
- Size: 30.6 MB
- Stars: 37
- Watchers: 5
- Forks: 13
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Emergent Multiagent Strategies
[Ankur Deka](https://www.linkedin.com/in/ankur-deka-120392135) and [Katia Sycara](http://www.cs.cmu.edu/~sycara/)
Robotics Institute, Carnegie Mellon University
This is the official repository of the paper [Natural Emergence of Heterogeneous Strategies in Artificially Intelligent Competitive Teams](https://arxiv.org/abs/2007.03102), presented in a [spotlight talk](https://youtu.be/ltHgKYc0F-E) in Robotics Science and Systems (RSS) 2020 workshop on Heterogeneous Multi-Robot Task Allocation and Coordination. Increasingly complex multiagent strategies can emerge naturally through competition in a mixed cooperative-competitive environment.
## Installation
I have tested the code on Ubuntu 16.04 and 18.04. See `requirements.txt` file for the list of dependencies. Create an anaconda environment or virtualenv with python 3.6 and setup everything by executing `pip install -r requirements.txt`.
```
conda create -n fortattack python=3.6 pip
conda activate fortattack
pip install -r requirements.txt
```
## Training
### Joint Training
`python train_fortattack.py --save-dir tmp_3`
See `arguments.py` file for the list of various command line arguments one can set while running the scripts.
### Ensemble Training of Guards
`python train_fortattack_v2.py --train-guards-only --pretrained-guard --guard-load-dir tmp_1 --guard-ckpt 2520 --attacker-load-dir tmp_1 --attacker-ckpts 220 650 1240 1600 2520 --save-dir tmp_4 --render`
## Testing
### Joint training results
`python test_fortattack.py --test --load-dir tmp_1 --ckpt 220`
### Ensemble training results
`python test_fortattack_v2.py --test --train-guards-only --num-eval-episodes 10 --load-dir tmp_2 --ckpt 5050 --attacker-load-dir tmp_1 --attacker-ckpts 220 --render`
### Trained models
I have provided some pretrained models in `marlsave` directory
1. `tmp_1`: normal training results. Multiagent strategies corresponding the checkpoints are as follows. Please refer to the paper for details of these strategies.
1. `ep220`: Flash laser strategy of guards
1. `ep650`: Sneaking strategy of attackers
1. `ep1240`: Spreading and flashing strategy of attackers
1. `ep1600`: Sneaking strategy of attackers
1. `ep2520`: Smartly spreading strategy of guards
1. `tmp_2/ep5050.pt`: Guards' policy after ensemble training
## Citing
If you use our code in your research, please cite our paper:
`@article{deka2020natural,
title={Natural Emergence of Heterogeneous Strategies in Artificially Intelligent Competitive Teams},
author={Deka, Ankur and Sycara, Katia},
journal={arXiv preprint arXiv:2007.03102},
year={2020}
}
`
## Contact
For any queries, feel free to raise an issue or contact me at [email protected] or [email protected].
## Acknowledgment
The environment is built up on [MAPE](https://github.com/openai/multiagent-particle-envs)
The algorithm for this repo is built up on [marl_transfer](https://github.com/sumitsk/marl_transfer)
## License
This project is licensed under the MIT License.