开源工具箱，助你快速测试联邦学习抗攻击能力
ByzFL: Research Framework for Robust Federated Learning
- 统一框架集成主流鲁棒聚合算法
- 支持多种攻击与异构数据场景模拟
- 单配置文件实现实验复现与可视化
我们提出ByzFL，一个用于开发和基准测试鲁棒联邦学习（FL）算法的开源Python库。ByzFL提供了一个统一且可扩展的框架，包含最先进的鲁棒聚合器实现、可配置攻击套件，以及模拟各种联邦学习场景的工具，包括异构数据分布、多种训练算法和对抗威胁模型。该库通过单一JSON配置文件实现系统化实验，并内置结果可视化工具。兼容PyTorch张量和NumPy数组，旨在促进可复现研究与鲁棒FL解决方案的快速原型设计。ByzFL可在https://byzfl.epfl.ch/获取，源代码托管于GitHub：https://github.com/LPD-EPFL/byzfl。
原文摘要 · Abstract (English)
We present ByzFL, an open-source Python library for developing and benchmarking robust federated learning (FL) algorithms. ByzFL provides a unified and extensible framework that includes implementations of state-of-the-art robust aggregators, a suite of configurable attacks, and tools for simulating a variety of FL scenarios, including heterogeneous data distributions, multiple training algorithms, and adversarial threat models. The library enables systematic experimentation via a single JSON-based configuration file and includes built-in utilities for result visualization. Compatible with PyTorch tensors and NumPy arrays, ByzFL is designed to facilitate reproducible research and rapid prototyping of robust FL solutions. ByzFL is available at https://byzfl.epfl.ch/, with source code hosted on GitHub: https://github.com/LPD-EPFL/byzfl.
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。