biMNLI

该数据集是平衡的多领域自然语言推理（MNLI）语料库的一个子集，主要关注蕴含和矛盾关系。在数据集中，蕴含、矛盾和中性类别的比例达到了1:1:1的平衡状态。该数据集的规模属于中等，所涉及的任务是自然语言推理。

This dataset is a subset of the balanced multi-domain natural language inference (MNLI) corpus, primarily focusing on entailment and contradiction relations. The dataset maintains a perfectly balanced 1:1:1 ratio across the entailment, contradiction, and neutral categories. It is a medium-sized dataset targeting the natural language inference task.