/off-policy

PyTorch implementations of popular off-policy multi-agent reinforcement learning algorithms, including QMix, VDN, MADDPG, and MATD3.

Primary LanguagePythonMIT LicenseMIT

Off-Policy Multi-Agent Reinforcement Learning (MARL) Algorithms

This repository contains implementations of various off-policy multi-agent reinforcement learning (MARL) algorithms.

Authors: Akash Velu and Chao Yu

Algorithms supported:

  • MADDPG (MLP and RNN)
  • MATD3 (MLP and RNN)
  • QMIX (MLP and RNN)
  • VDN (MLP and RNN)

Environments supported:

1. Usage

WARNING #1: by default all experiments assume a shared policy by all agents i.e. there is one neural network shared by all agents

WARNING #2: only QMIX and MADDPG are thoroughly tested; however,our VDN and MATD3 implementations make small modifications to QMIX and MADDPG, respectively. We display results using our implementation here.

All core code is located within the offpolicy folder. The algorithms/ subfolder contains algorithm-specific code for all methods. RMADDPG and RMATD3 refer to RNN implementationso of MADDPG and MATD3, and mQMIX and mVDN refer to MLP implementations of QMIX and VDN. We additionally support prioritized experience replay (PER).

  • The envs/ subfolder contains environment wrapper implementations for the MPEs and SMAC.

  • Code to perform training rollouts and policy updates are contained within the runner/ folder - there is a runner for each environment.

  • Executable scripts for training with default hyperparameters can be found in the scripts/ folder. The files are named in the following manner: train_algo_environment.sh. Within each file, the map name (in the case of SMAC and the MPEs) can be altered.

  • Python training scripts for each environment can be found in the scripts/train/ folder.

  • The config.py file contains relevant hyperparameter and env settings. Most hyperparameters are defaulted to the ones used in the paper; however, please refer to the appendix for a full list of hyperparameters used.

2. Installation

Here we give an example installation on CUDA == 10.1. For non-GPU & other CUDA version installation, please refer to the PyTorch website.

# create conda environment
conda create -n marl python==3.6.1
conda activate marl
pip install torch==1.5.1+cu101 torchvision==0.6.1+cu101 -f https://download.pytorch.org/whl/torch_stable.html
# install on-policy package
cd on-policy
pip install -e .

Even though we provide requirement.txt, it may have redundancy. We recommend that the user try to install other required packages by running the code and finding which required package hasn't installed yet.

2.1 Install StarCraftII 4.10

unzip SC2.4.10.zip
# password is iagreetotheeula
echo "export SC2PATH=~/StarCraftII/" > ~/.bashrc

2.2 Install MPE

# install this package first
pip install seaborn

There are 3 Cooperative scenarios in MPE:

  • simple_spread
  • simple_speaker_listener, which is 'Comm' scenario in paper
  • simple_reference

3.Train

Here we use train_mpe_maddpg.sh as an example:

cd offpolicy/scripts
chmod +x ./train_mpe_maddpg.sh
./train_mpe_maddpg.sh

Local results are stored in subfold scripts/results. Note that we use Weights & Bias as the default visualization platform; to use Weights & Bias, please register and login to the platform first. More instructions for using Weights&Bias can be found in the official documentation. Adding the --use_wandb in command line or in the .sh file will use Tensorboard instead of Weights & Biases.

4. Results

Results for the performance of RMADDPG and QMIX on the Particle Envs and QMIX in SMAC are depicted here. These results are obtained using a normal (not prioitized) replay buffer.