QEFT: Quantization for Efficient Fine-Tuning of LLMs

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lee, Changhun, Jin, Jun-gyu, Cho, Younghyun, Park, Eunhyeok
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916433626136576
author Lee, Changhun
Jin, Jun-gyu
Cho, Younghyun
Park, Eunhyeok
author_facet Lee, Changhun
Jin, Jun-gyu
Cho, Younghyun
Park, Eunhyeok
contents With the rapid growth in the use of fine-tuning for large language models (LLMs), optimizing fine-tuning while keeping inference efficient has become highly important. However, this is a challenging task as it requires improvements in all aspects, including inference speed, fine-tuning speed, memory consumption, and, most importantly, model quality. Previous studies have attempted to achieve this by combining quantization with fine-tuning, but they have failed to enhance all four aspects simultaneously. In this study, we propose a new lightweight technique called Quantization for Efficient Fine-Tuning (QEFT). QEFT accelerates both inference and fine-tuning, is supported by robust theoretical foundations, offers high flexibility, and maintains good hardware compatibility. Our extensive experiments demonstrate that QEFT matches the quality and versatility of full-precision parameter-efficient fine-tuning, while using fewer resources. Our code is available at https://github.com/xvyaward/qeft.
format Preprint
id arxiv_https___arxiv_org_abs_2410_08661
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle QEFT: Quantization for Efficient Fine-Tuning of LLMs
Lee, Changhun
Jin, Jun-gyu
Cho, Younghyun
Park, Eunhyeok
Computation and Language
Machine Learning
With the rapid growth in the use of fine-tuning for large language models (LLMs), optimizing fine-tuning while keeping inference efficient has become highly important. However, this is a challenging task as it requires improvements in all aspects, including inference speed, fine-tuning speed, memory consumption, and, most importantly, model quality. Previous studies have attempted to achieve this by combining quantization with fine-tuning, but they have failed to enhance all four aspects simultaneously. In this study, we propose a new lightweight technique called Quantization for Efficient Fine-Tuning (QEFT). QEFT accelerates both inference and fine-tuning, is supported by robust theoretical foundations, offers high flexibility, and maintains good hardware compatibility. Our extensive experiments demonstrate that QEFT matches the quality and versatility of full-precision parameter-efficient fine-tuning, while using fewer resources. Our code is available at https://github.com/xvyaward/qeft.
title QEFT: Quantization for Efficient Fine-Tuning of LLMs
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2410.08661