CLAILGOct 12, 2023

LoftQ: LoRA-Fine-Tuning-Aware Quantization for Large Language Models

Georgia TechMicrosoft
arXiv:2310.08659v4226 citationsh-index: 35Has Code
Originality Incremental advance
AI Analysis

This addresses the problem of efficient fine-tuning for LLM deployment, offering a domain-specific improvement for scenarios combining quantization and LoRA.

The paper tackles the performance gap between full fine-tuning and quantization plus LoRA fine-tuning for large language models by proposing LoftQ, a framework that simultaneously quantizes the model and finds a low-rank initialization for LoRA, improving generalization on downstream tasks with effectiveness shown in 2-bit and mixed-precision regimes.

Quantization is an indispensable technique for serving Large Language Models (LLMs) and has recently found its way into LoRA fine-tuning. In this work we focus on the scenario where quantization and LoRA fine-tuning are applied together on a pre-trained model. In such cases it is common to observe a consistent gap in the performance on downstream tasks between full fine-tuning and quantization plus LoRA fine-tuning approach. In response, we propose LoftQ (LoRA-Fine-Tuning-aware Quantization), a novel quantization framework that simultaneously quantizes an LLM and finds a proper low-rank initialization for LoRA fine-tuning. Such an initialization alleviates the discrepancy between the quantized and full-precision model and significantly improves generalization in downstream tasks. We evaluate our method on natural language understanding, question answering, summarization, and natural language generation tasks. Experiments show that our method is highly effective and outperforms existing quantization methods, especially in the challenging 2-bit and 2/4-bit mixed precision regimes. The code is available on https://github.com/yxli2123/LoftQ.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes