Automatic Grouping for Efficient Cooperative Multi-Agent Reinforcement Learning

Authors: Yifan Zang, Jinmin He, Kai Li, Haobo Fu, Qiang Fu, Junliang Xing, Jian Cheng

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on the Star Craft II micromanagement tasks and Google Research Football scenarios verify our method s effectiveness. Extensive component studies show how grouping works and enhances performance.
Researcher Affiliation Collaboration Yifan Zang1,2, Jinmin He1,2, Kai Li1,2 , Haobo Fu3, Qiang Fu3, Junliang Xing4, Jian Cheng1,2 1Institute of Automation, Chinese Academy of Sciences 2School of Artificial Intelligence, University of Chinese Academy of Sciences 3Tencent AI Lab, 4Tsinghua University
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code Yes The detailed setting of Go MARL s hyperparameters is shown in our source code1. 1https://github.com/zyfsjycc/Go MARL
Open Datasets Yes We compare all the methods in six Hard and Super Hard Star Craft II micromanagement tasks (SMAC) [29] and three challenging Google Research Football (GRF) [16] scenarios.
Dataset Splits No The paper mentions training and evaluation episodes, but does not provide specific training/validation/test dataset splits in terms of percentages or counts, which is typical for reinforcement learning environments where data is generated dynamically.
Hardware Specification Yes We use one NVIDIA Titan V GPU for training.
Software Dependencies No The paper mentions using the Py MARL2 framework but does not specify version numbers for PyMARL or any other software dependencies.
Experiment Setup Yes All the methods are trained with 8 parallel runners for 10M steps and are evaluated every 10K steps with 32 episodes.