marcov/glue_wnli_promptsource

该数据集包含多个特征，如sentence1、sentence2、label、idx、template_name、template、rendered_input和rendered_output。其中，label特征表示两个句子之间的关系，0表示不蕴含，1表示蕴含。数据集分为训练集、验证集和测试集，分别包含3175、355和730个样本。数据集的下载大小为468652字节，总大小为2550437字节。

The dataset includes multiple features such as sentence1, sentence2, label, idx, template_name, template, rendered_input, and rendered_output. The label feature is a classification label with two categories: 0 for not_entailment and 1 for entailment. The dataset is divided into training, validation, and test sets, containing 3175, 355, and 730 samples respectively. The download size of the dataset is 468652 bytes, and the total size is 2550437 bytes.