paper-with-me

Papers

Adaptive Batch Size and Learning Rate Scheduler for Stochastic Gradient Descent Based on Minimization of Stochastic First-order Oracle Complexity

2025-08-07 · Hikaru Umeda, Hideaki Iiduka arxiv

The convergence behavior of mini-batch stochastic gradient descent (SGD) is highly sensitive to the batch size and learning rate settings. Recent theoretical studies have identified the existence of a critical batch size that minimizes stochastic first-order oracle (SFO) complexity, defined as the expected number of gradient evaluations required to reach a stationary point of the empirical loss function in a deep neural network. An adaptive scheduling strategy is introduced to accelerate SGD that leverages theoretical findings on the critical batch size. The batch size and learning rate are adjusted on the basis of the observed decay in the full gradient norm during training. Experiments using an adaptive joint scheduler based on this strategy demonstrated improved convergence speed compared with that of existing schedulers.

📄 PDF Abstract BibTeX arXiv:2508.05302

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Increasing Both Batch Size and Learning Rate Accelerates Stochastic Gradient Descent

2024-09-13 · Hikaru Umeda, Hideaki Iiduka

The performance of mini-batch stochastic gradient descent (SGD) strongly depends on setting the batch size and learning rate to minimize the empirical loss in training the deep neural network. In this paper, we present t…

Automated Learning Rate Scheduler for Large-batch Training

2021-07-13 · ICML Workshop AutoML 2021 7 · Chiheon Kim, Saehoon Kim, Jongmin Kim, Donghoon Lee 외

Large-batch training has been essential in leveraging large-scale datasets and models in deep learning. While it is computationally beneficial to use large batch sizes, it often requires a specially designed learning rat…

image-classificationImage ClassificationScheduling

Hyper-Learning for Gradient-Based Batch Size Adaptation

2022-05-17 · Calum Robert MacLellan, Feng Dong

Scheduling the batch size to increase is an effective strategy to control gradient noise when training deep neural networks. Current approaches implement scheduling heuristics that neglect structure within the optimizati…

Hyperparameter OptimizationScheduling

Adaptive Sharpness-Aware Minimization with a Polyak-type Step size: A Theory-Grounded Scheduler

2026-06-01 · Dimitris Oikonomou, Nicolas Loizou arxiv

Sharpness-Aware Minimization (SAM) has established itself as a powerful and widely adopted optimizer for training machine learning models. By explicitly minimizing the sharpness of the loss landscape, SAM often improves …

Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler

2024-08-23 · Yikang Shen, Matthew Stallone, Mayank Mishra, Gaoyuan Zhang 외

Finding the optimal learning rate for language model pretraining is a challenging task. This is not only because there is a complicated correlation between learning rate, batch size, number of training tokens, model size…