现代信息科技2025，Vol.9Issue(6)：71-74,4.DOI:10.19850/j.cnki.2096-4706.2025.06.014
轻量级联邦学习平台LightFL的设计与实现
Design and Implementation of Lightweight Federated Learning Platform LightFL
摘要
Abstract
The existing Federated Learning platforms are generally complex and require in-depth learning to master their functions for users.To address this issue,a lightweight Federated Learning platform,LightFL,is developed with the aim of lowering the entry barrier of usage for Federated Learning frameworks.The platform supports simulations of both Independent and Identically Distributed and Non-Independent and Identically Distributed data,and implements classic Federated Learning algorithms such as FedAvg,FedProx,and MOON.Through a modular architecture and lightweight code design,users can easily engage in Federated Learning and experiments,and the platform is easy to deploy and maintain.Compared to other platforms,the platform can configure the Federated Learning environment with only a small amount of code to write for users.It simplifies the workflow,reduces the entry barrier of usage,and facilitates the further development of new Federated Learning algorithms.关键词
轻量级平台/联邦学习/独立同分布/非独立同分布Key words
lightweight platform/Federated Learning/Independent and Identically Distributed/Non-Independent and Identically Distributed分类
信息技术与安全科学引用本文复制引用
刘若轩,高凌航,谢国云,张宇帆,李宏恩,武文媗,王灿..轻量级联邦学习平台LightFL的设计与实现[J].现代信息科技,2025,9(6):71-74,4.基金项目
北京信息科技大学2024年大学生创新创业训练计划项目(S202411232607) （S202411232607）
北京信息科技大学"青年骨干教师"支持计划(YBT202450) （YBT202450）
