Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond

Abstract

In distributed learning, local SGD (also known as federated averaging) and its simple baseline minibatch SGD are widely studied optimization methods. Most existing analyses of these methods assume independent and unbiased gradient estimates obtained via with-replacement sampling. In contrast, we study shuffling-based variants: minibatch and local Random Reshuffling, which draw stochastic gradients without replacement and are thus closer to practice. For smooth functions satisfying the Polyak-Łojasiewicz condition, we obtain convergence bounds (in the large epoch regime) which show that these shuffling-based variants converge faster than their with-replacement counterparts. Moreover, we prove matching lower bounds showing that our convergence analysis is tight. Finally, we propose an algorithmic modification called synchronized shuffling that leads to convergence rates faster than our lower bounds in near-homogeneous settings.

ICLR 2022 camera-ready (selected for an oral presentation); 76 pages, 3 figures

References in corpus (6)

• On the Convergence of Local Descent Methods in Federated Learning
• Local SGD With a Communication Overhead Depending Only on the Number of Workers
• The Min-Max Complexity of Distributed Stochastic Convex Optimization with Intermittent Communication
• Proximal and Federated Random Reshuffling
• Random Shuffling Beats SGD Only After Many Epochs on Ill-Conditioned Problems
• Permutation-Based SGD: Is Random Optimal?