Overview

Overview

rftd

rftd is a training dataset for training, testing, and assessing legal machine learning models, including large language models (LLMs). It contains metadata for over 20,000 judgments that have been tagged and classified by human legal researchers according to a machine-readable taxonomy of issues and outcomes. The dataset is available in JSON for easy ingestion by machine learning software.

Dataset features

rftd contains human-applied semantic tags representing hundreds of thousands of …

Request sample

A specimen set of case records is available for testing and review governed by a …

Use cases

Use rftd to test your model’s accuracy against the real-world human …