datablations

AI & ML interests

Scaling Data-Constrained Language Models

Recent Activity

thomwolf  authored a paper 20 days ago

FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language

craffel  authored a paper 20 days ago

FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language

craffel  authored a paper about 1 month ago

The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text

View all activity

Team members 9