Papers
arxiv:2411.06402

Fineweb-Edu-Ar: Machine-translated Corpus to Support Arabic Small Language Models

Published on Nov 10
Authors:
,

Abstract

As large language models (LLMs) grow and develop, so do their data demands. This is especially true for multilingual LLMs, where the scarcity of high-quality and readily available data online has led to a multitude of synthetic dataset generation approaches. A key technique in this space is machine translation (MT), where high-quality English text is adapted to a target, comparatively low-resource language. This report introduces FineWeb-Edu-Ar, a machine-translated version of the exceedingly popular (deduplicated) FineWeb-Edu dataset from HuggingFace. To the best of our knowledge, FineWeb-Edu-Ar is the largest publicly available machine-translated Arabic dataset out there, with its size of 202B tokens of an Arabic-trained tokenizer.

Community

Paper author

Discussion

Hey everyone! I'm excited to share our work on FineWeb-Edu-Ar, a large-scale Arabic dataset we created by machine-translating the FineWeb-Edu corpus using NLLB-200. With 202 billion tokens of paired Arabic-English text, we hope this dataset will help advance research in Arabic small language models, particularly for researchers working with limited computational resources. We've included both the original English and translated Arabic versions to facilitate future research in cross-lingual learning. I'm particularly interested in hearing the community's thoughts on using machine translation for creating training data and potential applications for paired multilingual datasets in smaller models, especially in Arabic. Do you all have any suggestions for what could have been improved? Looking forward to discussing this with you all :)

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2411.06402 in a model README.md to link it from this page.

Datasets citing this paper 1

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2411.06402 in a Space README.md to link it from this page.

Collections including this paper 1