SQFT: Low-cost Model Adaptation in Low-precision Sparse Foundation Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Muñoz, Juan Pablo, Yuan, Jinjie, Jain, Nilesh
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916424439562240
author Muñoz, Juan Pablo
Yuan, Jinjie
Jain, Nilesh
author_facet Muñoz, Juan Pablo
Yuan, Jinjie
Jain, Nilesh
contents Large pre-trained models (LPMs), such as large language models, have become ubiquitous and are employed in many applications. These models are often adapted to a desired domain or downstream task through a fine-tuning stage. This paper proposes SQFT, an end-to-end solution for low-precision sparse parameter-efficient fine-tuning of LPMs, allowing for effective model manipulation in resource-constrained environments. Additionally, an innovative strategy enables the merging of sparse weights with low-rank adapters without losing sparsity and accuracy, overcoming the limitations of previous approaches. SQFT also addresses the challenge of having quantized weights and adapters with different numerical precisions, enabling merging in the desired numerical format without sacrificing accuracy. Multiple adaptation scenarios, models, and comprehensive sparsity levels demonstrate the effectiveness of SQFT. Models and code are available at https://github.com/IntelLabs/Hardware-Aware-Automated-Machine-Learning.
format Preprint
id arxiv_https___arxiv_org_abs_2410_03750
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SQFT: Low-cost Model Adaptation in Low-precision Sparse Foundation Models
Muñoz, Juan Pablo
Yuan, Jinjie
Jain, Nilesh
Machine Learning
Artificial Intelligence
Computation and Language
Large pre-trained models (LPMs), such as large language models, have become ubiquitous and are employed in many applications. These models are often adapted to a desired domain or downstream task through a fine-tuning stage. This paper proposes SQFT, an end-to-end solution for low-precision sparse parameter-efficient fine-tuning of LPMs, allowing for effective model manipulation in resource-constrained environments. Additionally, an innovative strategy enables the merging of sparse weights with low-rank adapters without losing sparsity and accuracy, overcoming the limitations of previous approaches. SQFT also addresses the challenge of having quantized weights and adapters with different numerical precisions, enabling merging in the desired numerical format without sacrificing accuracy. Multiple adaptation scenarios, models, and comprehensive sparsity levels demonstrate the effectiveness of SQFT. Models and code are available at https://github.com/IntelLabs/Hardware-Aware-Automated-Machine-Learning.
title SQFT: Low-cost Model Adaptation in Low-precision Sparse Foundation Models
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2410.03750