Craw4LLM: Efficient Web Crawling for LLM Pretraining

TL;DR

Craw4LLM introduces a web crawling method that prioritizes pages based on their influence on LLM pretraining, significantly improving data quality and efficiency while reducing web crawling waste.

Contribution

It proposes a novel influence-based priority scoring system for web crawling tailored to LLM pretraining, outperforming traditional connectivity-based methods.

Findings

Achieves comparable downstream performance with only 21% URLs crawled.

Reduces crawling waste and website burden significantly.

Demonstrates efficiency on a large web graph of 900 million pages.

Abstract

Web crawl is a main source of large language models' (LLMs) pretraining data, but the majority of crawled web pages are discarded in pretraining due to low data quality. This paper presents Craw4LLM, an efficient web crawling method that explores the web graph based on the preference of LLM pretraining. Specifically, it leverages the influence of a webpage in LLM pretraining as the priority score of the web crawler's scheduler, replacing the standard graph connectivity based priority. Our experiments on a web graph containing 900 million webpages from a commercial search engine's index demonstrate the efficiency of Craw4LLM in obtaining high-quality pretraining data. With just 21% URLs crawled, LLMs pretrained on Craw4LLM data reach the same downstream performances of previous crawls, significantly reducing the crawling waste and alleviating the burdens on websites. Our code is publicly available at https://github.com/cxcscmu/Craw4LLM.