FineWeb Filtered

FineWeb Filtered A filtered subset of HuggingFaceFW/fineweb.Only high-quality, clean English web documents are kept. No text is modified — filtering is purely keep/drop. What’s in this dataset Each record is a web document with these columns (inherited from FineWeb): Column Description text Full document text id Unique document ID url Source URL dump Common Crawl dump (e.g. CC-MAIN-2024-10) date Crawl date language Detected language code… See the full description on the dataset page:

Source: Hugging Face Hub (raj2708/fineweb-filtered). Metadata imported from the dataset’s Hub tags.