The Impact of Quantization on Retrieval-Augmented Generation: An Analysis of Small LLMs

arXiv (Cornell University)

Abstract

Post-training quantization reduces the computational demand of Large Language Models (LLMs) but can weaken some of their capabilities. Since LLM abilities emerge with scale, smaller LLMs are more sensitive to quantization. In this paper, we explore how quantization affects smaller LLMs' ability to perform retrieval-augmented generation (RAG), specifically in longer contexts. We chose personalization for evaluation because it is a challenging domain to perform using RAG as it requires long-context reasoning over multiple documents. We compare the original FP16 and the quantized INT4 performance of multiple 7B and 8B LLMs on two tasks while progressively increasing the number of retrieved documents to test how quantized models fare against longer contexts. To better understand the effect of retrieval, we evaluate three retrieval models in our experiments. Our findings reveal that if a 7B LLM performs the task well, quantization does not impair its performance and long-context reasoning capabilities. We conclude that it is possible to utilize RAG with quantized smaller LLMs.

Authors 3

1. Mert Yazan Aachen

   Leiden University · Amsterdam University of Applied Sciences · University of Applied Sciences Leiden

   Affiliation as printed

   Amsterdam University of Applied Sciences , Fraijlemaborg 133 , 1102 CV Amsterdam , Netherlands

   Leiden University , Einsteinweg 55 , 2333 CC Leiden , Netherlands

2. Suzan Verberne Aachen

   Leiden University · University of Applied Sciences Leiden

   Affiliation as printed

   Leiden University , Einsteinweg 55 , 2333 CC Leiden , Netherlands

3. Frederik Situmeang

   Amsterdam University of Applied Sciences

   Affiliation as printed

   Amsterdam University of Applied Sciences , Fraijlemaborg 133 , 1102 CV Amsterdam , Netherlands