marcov/super_glue_rte_promptsource

该数据集包含多个字段，用于自然语言处理任务，特别是文本推理任务。字段包括前提、假设、索引、标签、模板名称、模板、渲染输入和渲染输出。数据集分为训练集、验证集和测试集，分别包含24900、2770和30000个样本。

This dataset is designed for natural language inference tasks, featuring premises, hypotheses, indices, labels, template names, templates, rendered inputs, and rendered outputs. The labels are categorized into entailment and not_entailment. The dataset is divided into training, validation, and test sets, containing 24900, 2770, and 30000 samples respectively.