lexlms/lex_files
Dataset Card for "LexFiles" Dataset Summary The LeXFiles is a new diverse English multinational legal corpus that we created including 11 distinct sub-corpora that cover legislation and case law from 6 primarily English-speaking legal systems (EU, CoE, Canada, US, UK, India). The corpus contains approx. 19 billion tokens. In comparison, the "Pile of Law" corpus released by Hendersons et al. (2022) comprises 32 billion in total, where the majority (26/30) of… See the full description on the dataset page: https://huggingface.co/datasets/lexlms/lex_files.
| Author / organization | lexlms |
|---|---|
| License (as declared) | cc-by-nc-sa-4.0 |
| Downloads | 366 (as reported by Hugging Face) |
| Likes | 23 |
| Created | 10 May 2023 |
| Last updated at source | 24 July 2023 |
Popularity, last 3 days
366 now (-222 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
- lexlms/legal-roberta-largeModel · used to train
- lexlms/legal-roberta-baseModel · used to train
Recent changes
- 17 Oct 2026 Description: Dataset Card for "LexFiles" Dataset Summary The LeXFiles is a new diverse English multinational legal corpus that we created including 11 distinct sub-corpora that cover legislation and case law from 6 primarily English-speaking legal syste → Dataset Card for "LexFiles" Dataset Summary The LeXFiles is a new diverse English multinational legal corpus that we created including 11 distinct sub-corpora that cover legislation and case law from 6 primarily English-speaking legal syste
Changes detected by Silytics between refreshes of the source.
More from lexlms
- lexlms/legal_lamaDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.