paper-with-me

홈 › Papers

A Rank Stabilization Scaling Factor for Fine-Tuning with LoRA

2023-11-28 · Damjan Kalajdzievski

As large language models (LLMs) have become increasingly compute and memory intensive, parameter-efficient fine-tuning (PEFT) methods are now a common strategy to fine-tune LLMs. A popular PEFT method is Low-Rank Adapters (LoRA), which adds trainable low-rank "adapters" to selected layers. Each adapter consists of a low-rank matrix product, multiplicatively scaled by a rank-dependent factor. This scaling factor, which divides adapters by a factor of the rank, results in slowed learning and stunted performance for LoRA with higher-rank adapters. Consequently, the use of LoRA in practice has generally been limited to very low ranks. In this work, we study the impact of the scaling factor on the learning process and prove that LoRA adapters should be divided by a factor of the square root of the rank. Modifying LoRA with the appropriate scaling factor, which we call the rank-stabilized LoRA (rsLoRA) method, easily provides for a fine-tuning compute/performance trade-off, where larger ranks can be used to trade off increased computational resources during training for better fine-tuning performance, with no change in inference computing cost.

📄 PDF Abstract BibTeX arXiv:2312.03732

Code (0)

등록된 구현이 없습니다.

Tasks

parameter-efficient fine-tuning

Methods 이 논문이 사용한 방법론

Adapter 설명 없음

Similar Papers 제목 키워드 기반

RoRA: Efficient Fine-Tuning of LLM with Reliability Optimization for Rank Adaptation

2025-01-08 · Jun Liu, Zhenglun Kong, Peiyan Dong, Changdi Yang 외

Fine-tuning helps large language models (LLM) recover degraded information and enhance task performance. Although Low-Rank Adaptation (LoRA) is widely used and effective for fine-tuning, we have observed that its scaling…

Stabilized Fine-Tuning with LoRA in Federated Learning: Mitigating the Side Effect of Client Size and Rank via the Scaling Factor

2026-03-09 · Jiayu Huang, Xiaohu Wu, Tiantian He, Qicheng Lao arxiv

Large Language Models (LLMs) are pivotal in natural language processing. The impracticality of full fine-tuning has prompted Parameter-Efficient Fine-Tuning (PEFT) methods like Low-Rank Adaptation (LoRA), optimizing low-…

parameter-efficient fine-tuningFederated Learning

Learning Rate Scaling across LoRA Ranks and Transfer to Full Finetuning

2026-02-05 · Nan Chen, Soledad Villar, Soufiane Hayou arxiv

Low-Rank Adaptation (LoRA) is a standard tool for parameter-efficient finetuning of large models. While it induces a small memory footprint, its training dynamics can be surprisingly complex as they depend on several hyp…

Reinforcement LearningImage Generation

LoRA-Muon: Spectral Steepest Descent on the Low-Rank Manifold

2026-06-11 · Franz Louis Cesista, Katherine Crowson, Cédric Simal, Stella Biderman arxiv

Low-Rank Adaptation (LoRA) significantly reduces compute and memory costs for finetuning Deep Learning models but is often harder to tune than dense training: when using factor-wise optimizers such as AdamW, it is sensit…

LoFT: Low-Rank Adaptation That Behaves Like Full Fine-Tuning

2025-05-27 · Nurbek Tastan, Stefanos Laskaridis, Martin Takac, Karthik Nandakumar 외

Large pre-trained models are commonly adapted to downstream tasks using parameter-efficient fine-tuning methods such as Low-Rank Adaptation (LoRA), which injects small trainable low-rank matrices instead of updating all …

parameter-efficient fine-tuning