Reproducibility

Toward Reproducible Benchmarks in Federated Learning Research

Abstract Federated learning papers frequently report results on incompatible experimental setups — differing client …