StarCraft II unit micromanagement benchmark

The dataset used in the paper is a multi-agent reinforcement learning dataset, where agents learn to coordinate their actions to achieve a common goal.

BibTex: