Skip to content

Dataset · free-law

U.S. caselaw corpus

free-law/Caselaw_Access_Project

Digitized U.S. court decisions in English for legal text training and analysis.

What's inside

The Caselaw Access Project digitized U.S. court decisions: over 40 million decisions per card, including 6.7 million cases over the past 360 years. Data provided by Harvard Law Library together with Ravel Law and available via API and bulk download. Language is English; metadata volume is listed as 1M–10M.

What you can do with it

  • Fine-tuning a legal LLM on legal texts
  • SFT for legal document generation
  • Analysis and summarization of case law
Tasks
Text generation
Languages
English
Size
1M–10M rows
Downloads this month
1.1K
Likes
105
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("free-law/Caselaw_Access_Project")