pFL-Bench: A Comprehensive Benchmark for Personalized Federated Learning

TL;DR

pFL-Bench is a comprehensive benchmark designed to standardize and facilitate the evaluation of personalized federated learning methods across diverse datasets, settings, and practical scenarios, promoting reproducibility and further research.

Contribution

It introduces the first unified benchmark for pFL, including diverse datasets, a modular codebase with multiple methods, and systematic evaluation protocols.

Findings

Highlights the benefits of state-of-the-art pFL methods.

Demonstrates the robustness of pFL methods in various scenarios.

Provides insights into generalization, fairness, and system overhead.

Abstract

Personalized Federated Learning (pFL), which utilizes and deploys distinct local models, has gained increasing attention in recent years due to its success in handling the statistical heterogeneity of FL clients. However, standardized evaluation and systematical analysis of diverse pFL methods remain a challenge. Firstly, the highly varied datasets, FL simulation settings and pFL implementations prevent easy and fair comparisons of pFL methods. Secondly, the current pFL literature diverges in the adopted evaluation and ablation protocols. Finally, the effectiveness and robustness of pFL methods are under-explored in various practical scenarios, such as the generalization to new clients and the participation of resource-limited clients. To tackle these challenges, we propose the first comprehensive pFL benchmark, pFL-Bench, for facilitating rapid, reproducible, standardized and thorough pFL evaluation. The proposed benchmark contains more than 10 dataset variants in various application domains with a unified data partition and realistic heterogeneous settings; a modularized and easy-to-extend pFL codebase with more than 20 competitive pFL method implementations; and systematic evaluations under containerized environments in terms of generalization, fairness, system overhead, and convergence. We highlight the benefits and potential of state-of-the-art pFL methods and hope the pFL-Bench enables further pFL research and broad applications that would otherwise be difficult owing to the absence of a dedicated benchmark. The code is released at https://github.com/alibaba/FederatedScope/tree/master/benchmark/pFL-Bench.