Q-adaptive: Multi-Agent Reinforcement Learning Routing on Dragonfly Network

Multi-agent reinforcement learning improves routing efficiency in Dragonfly networks.