dleemiller/BigCat-NLI
收藏官方服务：
资源简介：
BigCatNLI数据集是一个综合性的自然语言推理数据集，由6个主要的数据集编译而成，包括NLI-FEVER、ANLI、SNLI、MNLI、LingNLI和WANLI。数据集采用统一的三标签分类体系（蕴含、中立、矛盾），并经过了去重和一致性检查。总共有约243万个示例，分为训练集、验证集和测试集。
The BigCatNLI dataset is a comprehensive compilation of natural language inference datasets, consisting of NLI-FEVER, ANLI, SNLI, MNLI, LingNLI, and WANLI. It uses a unified three-label classification system (entailment, neutral, contradiction) and has been deduplicated and consistency-checked. There are approximately 2.43 million examples, divided into training, validation, and test sets.
提供机构：
dleemiller
