LoRA-Pro: Are Low-Rank Adapters Properly Optimized?

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Zhengbo, Liang, Jian, He, Ran, Wang, Zilei, Tan, Tieniu
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912287977111552
author Wang, Zhengbo
Liang, Jian
He, Ran
Wang, Zilei
Tan, Tieniu
author_facet Wang, Zhengbo
Liang, Jian
He, Ran
Wang, Zilei
Tan, Tieniu
contents Low-rank adaptation, also known as LoRA, has emerged as a prominent method for parameter-efficient fine-tuning of foundation models. Despite its computational efficiency, LoRA still yields inferior performance compared to full fine-tuning. In this paper, we first uncover a fundamental connection between the optimization processes of LoRA and full fine-tuning: using LoRA for optimization is mathematically equivalent to full fine-tuning using a low-rank gradient for parameter updates. And this low-rank gradient can be expressed in terms of the gradients of the two low-rank matrices in LoRA. Leveraging this insight, we introduce LoRA-Pro, a method that enhances LoRA's performance by strategically adjusting the gradients of these low-rank matrices. This adjustment allows the low-rank gradient to more accurately approximate the full fine-tuning gradient, thereby narrowing the performance gap between LoRA and full fine-tuning. Furthermore, we theoretically derive the optimal solutions for adjusting the gradients of the low-rank matrices, applying them during fine-tuning in LoRA-Pro. We conduct extensive experiments across natural language understanding, dialogue generation, mathematical reasoning, code generation, and image classification tasks, demonstrating that LoRA-Pro substantially improves LoRA's performance, effectively narrowing the gap with full fine-tuning. Code is publicly available at https://github.com/mrflogs/LoRA-Pro.
format Preprint
id arxiv_https___arxiv_org_abs_2407_18242
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LoRA-Pro: Are Low-Rank Adapters Properly Optimized?
Wang, Zhengbo
Liang, Jian
He, Ran
Wang, Zilei
Tan, Tieniu
Machine Learning
Artificial Intelligence
Computation and Language
Low-rank adaptation, also known as LoRA, has emerged as a prominent method for parameter-efficient fine-tuning of foundation models. Despite its computational efficiency, LoRA still yields inferior performance compared to full fine-tuning. In this paper, we first uncover a fundamental connection between the optimization processes of LoRA and full fine-tuning: using LoRA for optimization is mathematically equivalent to full fine-tuning using a low-rank gradient for parameter updates. And this low-rank gradient can be expressed in terms of the gradients of the two low-rank matrices in LoRA. Leveraging this insight, we introduce LoRA-Pro, a method that enhances LoRA's performance by strategically adjusting the gradients of these low-rank matrices. This adjustment allows the low-rank gradient to more accurately approximate the full fine-tuning gradient, thereby narrowing the performance gap between LoRA and full fine-tuning. Furthermore, we theoretically derive the optimal solutions for adjusting the gradients of the low-rank matrices, applying them during fine-tuning in LoRA-Pro. We conduct extensive experiments across natural language understanding, dialogue generation, mathematical reasoning, code generation, and image classification tasks, demonstrating that LoRA-Pro substantially improves LoRA's performance, effectively narrowing the gap with full fine-tuning. Code is publicly available at https://github.com/mrflogs/LoRA-Pro.
title LoRA-Pro: Are Low-Rank Adapters Properly Optimized?
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2407.18242