Closing the convergence gap of SGD without replacement

Abstract

Stochastic gradient descent without replacement sampling is widely used in practice for model training. However, the vast majority of SGD analyses assumes data is sampled with replacement, and when the function minimized is strongly convex, an $\mathcal{O}\left(\frac{1}{T}\right)$ rate can be established when SGD is run for $T$ iterations. A recent line of breakthrough works on SGD without replacement (SGDo) established an $\mathcal{O}\left(\frac{n}{T^2}\right)$ convergence rate when the function minimized is strongly convex and is a sum of $n$ smooth functions, and an $\mathcal{O}\left(\frac{1}{T^2}+\frac{n^3}{T^3}\right)$ rate for sums of quadratics. On the other hand, the tightest known lower bound postulates an $\Omega \left(\frac{1}{T^2}+\frac{n^2}{T^3}\right)$ rate, leaving open the possibility of better SGDo convergence rates in the general case. In this paper, we close this gap and show that SGD without replacement achieves a rate of $\mathcal{O}\left(\frac{1}{T^2}+\frac{n^2}{T^3}\right)$ when the sum of the functions is a quadratic, and offer a new lower bound of $\Omega \left(\frac{n}{T^2}\right)$ for strongly convex functions that are sums of smooth functions.

arXiv commentSimplified some proofs and fixed typos

Cited by in corpus (8)

• A Unified Convergence Analysis for Shuffling-Type Gradient Methods
• Fast Distributionally Robust Learning with Variance Reduced Min-Max Optimization
• Incremental Without Replacement Sampling in Nonconvex Optimization
• Can Single-Shuffle SGD be Better than Reshuffling SGD and GD?
• SMG: A Shuffling Gradient-Based Method with Momentum
• Minibatch vs Local SGD with Shuffling: Tight Convergence Bounds and Beyond
• Permutation-Based SGD: Is Random Optimal?
• Optimal Rates for Random Order Online Optimization