发表机构
University of Turin（都灵大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对现有联邦学习框架缺乏可定制性和性能的问题，提出开源框架FFL，通过C/C++实现支持任意通信图，在不同平台上获得2.5至3.69倍加速，并计划提供Python接口和动态联邦支持。
AI 中文摘要
联邦学习（FL）正经历着大量的研究兴趣，许多框架被开发出来，使从业者能够轻松快速地构建联邦。这些努力中的大多数没有考虑机器学习（ML）软件的两个关键方面：可定制性和性能。本研究通过实现一个名为FastFederatedLearning（FFL）的开源FL框架来解决这些问题。FFL使用C/C++实现，专注于代码性能，并允许用户指定联邦中涉及的客户端和服务器之间的任何通信图，确保可定制性。FFL与Intel OpenFL进行了测试，在不同计算平台（x86-64、ARM-v8、RISC-V）上实现了一致的加速，范围从2.5倍到3.69倍。我们旨在用Python接口包装FFL以简化其使用，并为不同的通信后端实现中间件。我们旨在构建动态联邦，其中客户端和服务器之间的关系不是静态的，从而营造一个联邦可以被视为长期演进结构并作为服务加以利用的环境。
英文摘要
Federated Learning (FL) is experiencing a substantial research interest, with many frameworks being developed to allow practitioners to build federations easily and quickly. Most of these efforts do not consider two main aspects that are key to Machine Learning (ML) software: customizability and performance. This research addresses these issues by implementing an open-source FL framework named FastFederatedLearning (FFL). FFL is implemented in C/C++, focusing on code performance, and allows the user to specify any communication graph between clients and servers involved in the federation, ensuring customizability. FFL is tested against Intel OpenFL, achieving consistent speedups over different computational platforms (x86-64, ARM-v8, RISC-V), ranging from 2.5x and 3.69x. We aim to wrap FFL with a Python interface to ease its use and implement a middleware for different communication backends to be used. We aim to build dynamic federations in which relations between clients and servers are not static, giving life to an environment where federations can be seen as long-time evolving structures and exploited as services.
Journal refEuro-Par 2023: Parallel Processing Workshops - Euro-Par 2023 International Workshops, Limassol, Cyprus, August 28 - September 1, 2023, Revised Selected Papers, Part II
DOI：10.1007/978-3-031-48803-0_40
