Tviskaron / pogema-baselines

PPO and PyMARL baseline for Pogema environment
20 stars 5 forks source link
marl pathfinding ppo qmix reinforcement-learning

Pogema Baselines


Important Note:

The Baselines repository will not be receiving any future updates. The main baselines have been transferred to the POGEMA-Benchmark repository, which includes more algorithms and tools.


This repository contains a set of baselines for the Pogema environment.

Multi-Agent RL Baselines

Implementations based on the oxwhirl/pymarl

Large-Scale Experiments

Installation

Just install all dependencies using:

pip install -r docker/requirements.txt

Training Example

Run main.py with one of the configs from the configs folder:

python main.py --config configs/8x8.yaml

Detailed instructions are available in the APPO and PyMARL readme files.

We thank PyMARL and SampleFactory contributors for their implementations.