WeightBridge: An Efficient Weight Transfer Library for Reinforcement Learning

原文：WeightBridge: An Efficient Weight Transfer Library for Reinforcement Learning

作者：Xuanlin Jiang, Samuel Hsia, Michael Kuchnik, Zachary DeVito, Minlan Yu, Carole-Jean Wu

来源：arXiv cs.DC（分布式并行）

正文

Computer Science > Distributed, Parallel, and Cluster Computing

arXiv:2609.25442v1 (cs)

[Submitted on 21 Sep 2026]

Title:WeightBridge: An Efficient Weight Transfer Library for Reinforcement Learning

Authors:Xuanlin Jiang, Samuel Hsia, Michael Kuchnik, Zachary DeVito, Minlan Yu, Carole-Jean Wu

View a PDF of the paper titled WeightBridge: An Efficient Weight Transfer Library for Reinforcement Learning, by Xuanlin Jiang and 5 other authors

View PDF

HTML (experimental)

Abstract:Weight transfer - the propagation of updated parameters from trainers to rollout generators - is becoming an important performance bottleneck in reinforcement learning (RL) systems for LLMs. The central challenge is supporting the diverse trainer and rollout layouts and synchronization requirements of modern RL workloads without sacrificing efficiency. Existing solutions are efficient under some configurations but perform poorly or lack support under others. We present WeightBridge, a flexible, efficient weight-transfer library designed to deliver high performance across diverse RL configurations. WeightBridge first automatically extracts the correspondence between trainer and rollout weight layouts, then plans and executes redundancy-free and load-balanced weight transfer. It exposes a small, general API while coordinating workers across diverse synchronization modes. Across configurations spanning different models, parallelization layouts, and synchronization modes, WeightBridge reduces average GPU stall time by up to 42$\times$ over the state-of-the-art open-source RL framework and achieves high performance in all settings. A coding agent was able to integrate WeightBridge into two different RL frameworks without manual guidance, demonstrating the generality and ease of use of its APIs.

Comments:

25 pages, 16 figures

Subjects:

Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG); Networking and Internet Architecture (cs.NI)

Cite as:

arXiv:2609.25442 [cs.DC]

(or

arXiv:2609.25442v1 [cs.DC] for this version)

https://doi.org/10.48550/arXiv.2609.25442

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

主题

分布式计算 · 机器学习 · cs.NI

由「前沿雷达」于 2026-09-22 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。