用多智能体强化学习高效映射神经网络，显著提升硬件部署性能。
Multi-Agent Reinforcement Learning for Sample-Efficient Deep Neural Network Mapping
- 多个智能体分布式并行搜索映射空间，加速探索
- 相比单智能体方法，采样效率提升30至300倍
- 适合芯片设计与深度学习加速器研发人员
将深度神经网络（DNN）映射到硬件是优化延迟、能耗和资源利用率的关键，也是高性能加速器设计的核心。由于映射空间巨大且复杂，强化学习（RL）虽具潜力，但常受限于采样效率低下。本文提出一种去中心化的多智能体强化学习（MARL）框架，通过多智能体分布搜索加速探索。为避免并行训练多个智能体带来的冗余，我们引入基于相关性分析的智能体聚类算法，将相似映射参数分配给同一智能体，实现高效去中心化并行学习。实验表明，在相同采样量条件下，该方法相较标准单智能体RL样本效率提升30–300倍，可实现最高32.61倍延迟降低和16.45倍能量-延迟积（EDP）降低。
原文摘要 · Abstract (English)
Mapping deep neural networks (DNNs) to hardware is critical for optimizing latency, energy consumption, and resource utilization, making it a cornerstone of high-performance accelerator design. Due to the vast and complex mapping space, reinforcement learning (RL) has emerged as a promising approach-but its effectiveness is often limited by sample inefficiency. We present a decentralized multi-agent reinforcement learning (MARL) framework designed to overcome this challenge. By distributing the search across multiple agents, our framework accelerates exploration. To avoid inefficiencies from training multiple agents in parallel, we introduce an agent clustering algorithm that assigns similar mapping parameters to the same agents based on correlation analysis. This enables a decentralized, parallelized learning process that significantly improves sample efficiency. Experimental results show our MARL approach improves sample efficiency by 30-300x over standard single-agent RL, achieving up to 32.61x latency reduction and 16.45x energy-delay product (EDP) reduction under iso-sample conditions.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。