Taming LLMs by Scaling Learning Rates with Gradient Grouping

Siyuan Li; Juanxi Tian; Zedong Wang; Xin Jin; Zicheng Liu; Wentao Zhang; Dan Xu

返回文献探索

Paper · arXiv 2506.01049

Taming LLMs by Scaling Learning Rates with Gradient Grouping

Siyuan Li, Juanxi Tian, Zedong Wang, Xin Jin, Zicheng Liu, Wentao Zhang, Dan Xu

40 upvotesJune 1, 2025arXiv 预印本

AI 摘要

SGG, an optimizer wrapper, enhances adaptive learning rates for large language models by grouping gradients and applying cluster-specific scaling, improving convergence and stability.

large language modelsadaptive optimizersAdamWparameter-wise learning rate estimationtraining instabilityparameter-efficient fine-tuningScaling with Gradient Groupinggradient groupingcluster-specific scalingLLM benchmarksrobust choice for LLM optimization

Abstract

Training large language models (LLMs) poses challenges due to their massive scale and heterogeneous architectures. While adaptive optimizers like AdamW help address gradient variations, they still struggle with efficient and effective parameter-wise learning rate estimation, resulting in training instability, slow convergence, and poor compatibility with parameter-efficient fine-tuning (PEFT) techniques. This work introduces Scaling with Gradient Grouping (SGG), an optimizer wrapper that improves adaptive learning rate estimation by dynamic grouping and group-specific scaling. SGG first groups gradient statistics in each layer into clusters and then applies cluster-specific scaling to calibrate learning rates for each parameter, thus imposing collective group-wise constraints while maintaining precise per-parameter adaptation. Experiments on diverse (M)LLM benchmarks show that SGG integrates seamlessly with existing optimizers, and offers consistent gains and faster convergence over baselines, with various model sizes. Its stability across varying batch sizes and learning rates establishes SGG as a robust choice for LLM optimization.