Dataset · coastalcph
LexGLUE legal benchmark
coastalcph/lex_glue
A collection of legal text tasks for evaluating language understanding models in English.
What's inside
LexGLUE is a benchmark for evaluating legal English language understanding, inspired by GLUE and SuperGLUE. It includes classification and QA tasks on legal texts. Size ranges from 100k to 1M examples; modality is text.
What you can do with it
- evaluate legal LLMs on text classification
- test QA on legal documents
- fine-tune LoRA for legal NLP
- Tasks
- Question answering, Text classification
- Languages
- English
- Size
- 100K–1M rows
- Downloads this month
- 28.7K
- Likes
- 85
Models trained on it
- StrandsAgents/strands-decider-2B-qwen3.5-v1-2610Text classification 24.5K
- jaredpalmer/kev-27bText classification 3.4K
- mradermacher/kev-27b-GGUFMultiple choice 1.6K
- mradermacher/elf-1-12b-GGUF 737
- mradermacher/elf-1-31b-GGUF 710
- mradermacher/bobcat-1.3-GGUF 263
- sanghwa-na/bobcat-1.3Zero shot classification 232
- amazon/strands-decider-2B-qwen3.5-v1-2610Text classification 190
How to load it
from datasets import load_dataset
ds = load_dataset("coastalcph/lex_glue")