Related papers: CCNet: Extracting High Quality Monolingual Dataset…

In recent years, the field of document understanding has progressed a lot. A significant part of this progress has been possible thanks to the use of language models pretrained on large amounts of documents. However, pretraining corpora…

Pre-training state-of-the-art large language models (LLMs) requires vast amounts of clean and diverse text data. While the open development of large high-quality English pre-training datasets has seen substantial recent progress, training…

We present the HPLT (High Performance Language Technologies) language resources, a new massive multilingual dataset including both monolingual and bilingual corpora extracted from CommonCrawl and previously unused web crawls from the…

The need for raw large raw corpora has dramatically increased in recent years with the introduction of transfer learning and semi-supervised learning methods to Natural Language Processing. And while there have been some recent attempts to…

Large language models have led to remarkable progress on many NLP tasks, and researchers are turning to ever-larger text corpora to train them. Some of the largest corpora available are made by scraping significant portions of the internet,…

With the success of large-scale pre-training and multilingual modeling in Natural Language Processing (NLP), recent years have seen a proliferation of large, web-mined text datasets covering hundreds of languages. We manually audit the…

The performance of a large language model (LLM) depends heavily on the quality and size of its pretraining dataset. However, the pretraining datasets for state-of-the-art open LLMs like Llama 3 and Mixtral are not publicly available and…

The performance of large language models (LLMs) and large multimodal models (LMMs) depends heavily on the quality and scale of their pre-training datasets. Recent research shows that large multimodal models trained on natural documents…

Cross-lingual document alignment aims to identify pairs of documents in two distinct languages that are of comparable content or translations of each other. In this paper, we exploit the signals embedded in URLs to label web documents at…

Large language models are commonly trained on a mixture of filtered web data and curated high-quality corpora, such as social media conversations, books, or technical papers. This curation process is believed to be necessary to produce…

Parallel texts are a relatively rare language resource, however, they constitute a very useful research material with a wide range of applications. This study presents and analyses new methodologies we developed for obtaining such data from…

There is growing evidence that pretraining on high quality, carefully thought-out tokens such as code or mathematics plays an important role in improving the reasoning abilities of large language models. For example, Minerva, a PaLM model…

We introduce WordScape, a novel pipeline for the creation of cross-disciplinary, multilingual corpora comprising millions of pages with annotations for document layout detection. Relating visual and textual items on document pages has…

One of the first pre-processing steps for constructing web-scale LLM pretraining datasets involves extracting text from HTML. Despite the immense diversity of web content, existing open-source datasets predominantly apply a single fixed…

Pretraining large language models (LLMs) on high-quality, structured data such as mathematics and code substantially enhances reasoning capabilities. However, existing math-focused datasets built from Common Crawl suffer from degraded…

Massive web-crawled image-text datasets lay the foundation for recent progress in multimodal learning. These datasets are designed with the goal of training a model to do well on standard computer vision benchmarks, many of which, however,…

We are presenting a text analysis tool set that allows analysts in various fields to sieve through large collections of multilingual news items quickly and to find information that is of relevance to them. For a given document collection,…

Large language models (LLMs) rely heavily on web-scale datasets like Common Crawl, which provides over 80\% of training data for some modern models. However, the indiscriminate nature of web crawling raises challenges in data quality,…

This paper presents WanJuan-CC, a safe and high-quality open-sourced English webtext dataset derived from Common Crawl data. The study addresses the challenges of constructing large-scale pre-training datasets for language models, which…

We are presenting a set of multilingual text analysis tools that can help analysts in any field to explore large document collections quickly in order to determine whether the documents contain information of interest, and to find the…