Running
README
👀
AI & ML interests
We release large pre-training datasets to accelerate open LLM development. Part of the Hugging Face Science team (hf.co/science)
Recent Activity
Papers
FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language
The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale
Running
Featured
65
FinePDFs: Liberating 3T of the finest tokens from PDFs
📄
Running
11
FineWiki Viewer
🌐
Viewer to explore the finewiki dataset
Running
Featured
1.28k
FineWeb: decanting the web for the finest text data at scale
🍷
Generate high-quality text data for LLMs using FineWeb
Running
88
Scaling FineWeb to 1000+ languages: Step 1: finding signal in 100s of evaluation tasks
📝
Evaluate multilingual models using FineTasks
Build error
1
Tasks Explorer
🏢
Explore and analyze experiment results
Runtime error
4
Datasets Metrics Explorer
📊
Launch an interactive demo interface