Fairness for Cooperative Multi-Agent Learning with Equivariant Policies

This repository is the official implementation of Fairness for Cooperative Multi-Agent Learning with Equivariant Policies.

Setup

Python 3.5.4 (anaconda environment recommended)

To install requirements:

pip install -r requirements.txt

To setup multi-agent environments:

cd simple_particle_envs
pip install -e .

To verify installation, run:

python baselines/baselines.py --mode test --render

Training

To train a Fair-E model, run:

python main.py --env simple_torus --algorithm ddpg_symmetric

To train a Fair-ER model, run:

python main.py --env simple_torus --algorithm ddpg_speed_fair

• The control parameter of fairness can be adjusted in configs.py.

To resume training from a checkpoint, run:

python main.py --env simple_torus --algorithm ddpg_symmetric --checkpoint_path /path/to/model/checkpoints

Evaluation

To collect trajectories from a trained model, run eval/collect_actions.py. Here are a few examples:

• Greedy pursuers against random-moving evader:

python eval/collect_actions.py --env simple_torus --pred_policy greedy --prey_policy random --seed 75 

• CD-DDPG pursuers against sophisticated evader:

python eval/collect_actions.py --env simple_torus --pred_policy ddpg --prey_policy cosine --seed 75 --checkpoint_path /path/to/model/checkpoints

To create fairness vs. utility plots, run:

python eval/fairness_vs_utility.py --fp /path/to/folder/of/results