lbox-open-v0.2.1

#lbox-open-v0.2.1

Dataset

• case_corpus: 150k Korean legal cases.
• defamation_corpus: 1,536 defamation (명예훼손) precedents. See also #4.
• act_on_special_cases_concerning_the_settlement_of_traffic_accidents_corpus: 768 교통사고처리특례법위반(치상) precedents
• casename_classification: 10,000 (facts, case name) pairs.
• casename_classification_plus: 31,283 (facts, case name) pairs.
• statute_classification: 2,760 (facts, statutes) pairs.
• statute_classification_plus: 17,730 (facts, statutes) pairs.
• summarization: 20,000 (precedent, summary) pairs.
• summarization_plus: 51,114 (precedent, summary) pairs.

Model

• Release legal-mt5-small.
• Release legal-mt5-small fine-tuned on the summarization dataset.

lbox-open-v0.1

lbox-open-v0.1

• case_corpus: 150k Korean legal cases.
• casename_classification: 10,000 (facts, case name) pairs.
• statute_classification: 2,760 (facts, statutes) pairs
• summarization: 20,000 (precedent, summary) pairs.