# Optimal mini-batch and step sizes for SAGA

arXiv: https://arxiv.org/abs/1902.00071 · HTML: https://mlsanity.com/paper/1902.00071/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 31 Jan 2019.

## Built on by (most cited first, 9 of 9)

- [Datamodels: Predicting Predictions from Training Data](https://mlsanity.com/paper/2202.00622/) (2022)
- [Leveraging Sparse Linear Layers for Debuggable Deep Networks](https://mlsanity.com/paper/2105.04857/) (2021)
- [Small Batch Size Training for Language Models: When Vanilla SGD Works, and Why Gradient Accumulation Is Wasteful](https://mlsanity.com/paper/2507.07101/) (2025)
- [MURANA: A Generic Framework for Stochastic Variance-Reduced Optimization](https://mlsanity.com/paper/2106.03056/) (2021)
- [SketchySGD: Reliable Stochastic Optimization via Randomized Curvature Estimates](https://mlsanity.com/paper/2211.08597/) (2022)
- [Take 5: Interpretable Image Classification with a Handful of Features](https://mlsanity.com/paper/2303.13166/) (2023)
- [Memory Augmented Optimizers for Deep Learning](https://mlsanity.com/paper/2106.10708/) (2021)
- [SGD with Coordinate Sampling: Theory and Practice](https://mlsanity.com/paper/2105.11818/) (2021)
- [BLOKS: Scaling Kernel Solvers with Sketched Block Preconditioning](https://mlsanity.com/paper/2610.13382/) (2026)

## Abstract (arXiv)

Recently it has been shown that the step sizes of a family of variance reduced gradient methods called the JacSketch methods depend on the expected smoothness constant. In particular, if this expected smoothness constant could be calculated a priori, then one could safely set much larger step sizes which would result in a much faster convergence rate. We fill in this gap, and provide simple closed form expressions for the expected smoothness constant and careful numerical experiments verifying these bounds. Using these bounds, and since the SAGA algorithm is part of this JacSketch family, we suggest a new standard practice for setting the step sizes and mini-batch size for SAGA that are competitive with a numerical grid search. Furthermore, we can now show that the total complexity of the SAGA algorithm decreases linearly in the mini-batch size up to a pre-defined value: the optimal mini-batch size. This is a rare result in the stochastic variance reduced literature, only previously shown for the Katyusha algorithm. Finally we conjecture that this is the case for many other stochastic variance reduced methods and that our bounds and analysis of the expected smoothness constant is key to extending these results.