AutoGD: Automatic Learning Rate Selection for Gradient Descent

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Surjanovic, Nikola, Bouchard-Côté, Alexandre, Campbell, Trevor
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917004793872384
author Surjanovic, Nikola
Bouchard-Côté, Alexandre
Campbell, Trevor
author_facet Surjanovic, Nikola
Bouchard-Côté, Alexandre
Campbell, Trevor
contents The performance of gradient-based optimization methods, such as standard gradient descent (GD), greatly depends on the choice of learning rate. However, it can require a non-trivial amount of user tuning effort to select an appropriate learning rate schedule. When such methods appear as inner loops of other algorithms, expecting the user to tune the learning rates may be impractical. To address this, we introduce AutoGD: a gradient descent method that automatically determines whether to increase or decrease the learning rate at a given iteration. We establish the convergence of AutoGD, and show that we can recover the optimal rate of GD (up to a constant) for a broad class of functions without knowledge of smoothness constants. Experiments on a variety of traditional problems and variational inference optimization tasks demonstrate strong performance of the method, along with its extensions to AutoBFGS and AutoLBFGS.
format Preprint
id arxiv_https___arxiv_org_abs_2510_09923
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AutoGD: Automatic Learning Rate Selection for Gradient Descent
Surjanovic, Nikola
Bouchard-Côté, Alexandre
Campbell, Trevor
Machine Learning
Optimization and Control
Computation
The performance of gradient-based optimization methods, such as standard gradient descent (GD), greatly depends on the choice of learning rate. However, it can require a non-trivial amount of user tuning effort to select an appropriate learning rate schedule. When such methods appear as inner loops of other algorithms, expecting the user to tune the learning rates may be impractical. To address this, we introduce AutoGD: a gradient descent method that automatically determines whether to increase or decrease the learning rate at a given iteration. We establish the convergence of AutoGD, and show that we can recover the optimal rate of GD (up to a constant) for a broad class of functions without knowledge of smoothness constants. Experiments on a variety of traditional problems and variational inference optimization tasks demonstrate strong performance of the method, along with its extensions to AutoBFGS and AutoLBFGS.
title AutoGD: Automatic Learning Rate Selection for Gradient Descent
topic Machine Learning
Optimization and Control
Computation
url https://arxiv.org/abs/2510.09923