VolleyBots

该数据集为一个竞争性的多代理环境，旨在让智能体在3对3的无人机排球游戏中学习策略。该数据集是通过包括低级别技能训练和策略自我对弈在内的多个训练阶段生成的。规模上，该数据集涉及多个智能体（3对3）通过自我对弈的方式进行训练。任务方面，它专注于基于回合的竞争环境下的多代理强化学习。

This dataset is a competitive multi-agent environment dedicated to training agents to learn optimal strategies in a 3-on-3 drone volleyball game. It is generated through multiple training phases, including low-level skill training and policy self-play. In terms of scale, the dataset involves multiple agents (3 vs 3) undergoing training via self-play. As for task orientation, it focuses on multi-agent reinforcement learning in turn-based competitive scenarios.