FLamby

FLamby is a suite of 7 healthcare datasets designed to evaluate cross-silo federated learning strategies, featuring natural splits and covering multiple tasks, modalities, and data volumes.

Papers using FLamby (3)

Auto-FL-Research: Agentic Search for Federated Learning Algorithms2026FLamby: Datasets and Benchmarks for Cross-Silo Federated Learning in Realistic Healthcare Settings2022 · 53 citesA Comprehensive View of Personalized Federated Learning on Heterogeneous Clinical Datasets2023