FedRDMA: Communication-Efficient Cross-Silo Federated LLM via Chunked RDMA Transmission

Autor:	Zhang, Zeling, Cai, Dongqi, Zhang, Yiran, Xu, Mengwei, Wang, Shangguang, Zhou, Ao
Rok vydání:	2024
Předmět:	Computer Science - Machine Learning Computer Science - Distributed Parallel and Cluster Computing Computer Science - Networking and Internet Architecture
Druh dokumentu:	Working Paper
Popis:	Communication overhead is a significant bottleneck in federated learning (FL), which has been exaggerated with the increasing size of AI models. In this paper, we propose FedRDMA, a communication-efficient cross-silo FL system that integrates RDMA into the FL communication protocol. To overcome the limitations of RDMA in wide-area networks (WANs), FedRDMA divides the updated model into chunks and designs a series of optimization techniques to improve the efficiency and robustness of RDMA-based communication. We implement FedRDMA atop the industrial federated learning framework and evaluate it on a real-world cross-silo FL scenario. The experimental results show that \sys can achieve up to 3.8$\times$ speedup in communication efficiency compared to traditional TCP/IP-based FL systems. Comment: under review
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2403.00881 Zobrazit plný text záznamu View this record from Arxiv