Skip to content

Dataset · isaacus

Open Australian legal corpus

isaacus/open-australian-legal-corpus

229,122 Australian legal documents in English: legislation and judicial acts.

What's inside

The corpus combines 229,122 texts totaling over 60 million rows and 1.4 billion tokens. It includes current statutes and regulations of the Commonwealth, New South Wales, Queensland, Western Australia, South Australia, Tasmania, and Norfolk Island, as well as judicial documents. This is the first open multi-jurisdictional Australian legal corpus in English.

What you can do with it

  • Fine-tuning legal LLMs and LoRA for Australian law
  • Search and retrieval over legislation and court decisions
  • Evaluating models on legal texts
Tasks
Text generation, Fill mask, Text retrieval
Languages
English
Size
100K–1M rows
Downloads this month
7.3K
Likes
109

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("isaacus/open-australian-legal-corpus")