Mime: Mimicking Centralized Stochastic Algorithms in Federated Learning
This addresses a major open problem in federated learning by enabling more efficient and stable optimization across heterogeneous devices, representing a significant advancement rather than an incremental improvement.
The paper tackles the problem of client drift in federated learning due to data heterogeneity, proposing Mime, a framework that adapts centralized algorithms like momentum and Adam to federated settings, and proves it can be faster than centralized methods with momentum-based variance reduction.
Federated learning (FL) is a challenging setting for optimization due to the heterogeneity of the data across different clients which gives rise to the client drift phenomenon. In fact, obtaining an algorithm for FL which is uniformly better than simple centralized training has been a major open problem thus far. In this work, we propose a general algorithmic framework, Mime, which i) mitigates client drift and ii) adapts arbitrary centralized optimization algorithms such as momentum and Adam to the cross-device federated learning setting. Mime uses a combination of control-variates and server-level statistics (e.g. momentum) at every client-update step to ensure that each local update mimics that of the centralized method run on iid data. We prove a reduction result showing that Mime can translate the convergence of a generic algorithm in the centralized setting into convergence in the federated setting. Further, we show that when combined with momentum based variance reduction, Mime is provably faster than any centralized method--the first such result. We also perform a thorough experimental exploration of Mime's performance on real world datasets.