Containerized Distributed Value-Based Multi-Agent Reinforcement Learning
This work addresses system-level bottlenecks in distributed multi-agent reinforcement learning for researchers and practitioners, offering a scalable solution with significant performance gains, though it is incremental in its technical approach.
The paper tackles the high sample demand and system challenges of distributed multi-agent reinforcement learning by proposing a containerized framework that packs environment instances, a local learner, buffer, and multi-queue manager into containers to enhance scalability and diversity. It achieves state-of-the-art results, including solving the Google Research Football full game 5_v_5 for the first time and getting 4-18x better results on the StarCraft II benchmark.
Multi-agent reinforcement learning tasks put a high demand on the volume of training samples. Different from its single-agent counterpart, distributed value-based multi-agent reinforcement learning faces the unique challenges of demanding data transfer, inter-process communication management, and high requirement of exploration. We propose a containerized learning framework to solve these problems. We pack several environment instances, a local learner and buffer, and a carefully designed multi-queue manager which avoids blocking into a container. Local policies of each container are encouraged to be as diverse as possible, and only trajectories with highest priority are sent to a global learner. In this way, we achieve a scalable, time-efficient, and diverse distributed MARL learning framework with high system throughput. To own knowledge, our method is the first to solve the challenging Google Research Football full game $5\_v\_5$. On the StarCraft II micromanagement benchmark, our method gets $4$-$18\times$ better results compared to state-of-the-art non-distributed MARL algorithms.