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

TL;DR

SQFT introduces a low-cost, sparse, and low-precision fine-tuning method for large pre-trained models, enabling efficient adaptation in resource-limited settings while maintaining accuracy.

Contribution

The paper presents SQFT, a novel end-to-end approach for sparse, low-precision fine-tuning and merging of weights and adapters without accuracy loss.

Findings

Effective across multiple models and sparsity levels

Maintains accuracy with low-precision, sparse fine-tuning

Enables resource-efficient model adaptation

Abstract

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.