marcov/paws_labeled_swap_promptsource

该数据集包含多个特征，如id、sentence1、sentence2、label、template_name、template、rendered_input和rendered_output。label是一个二分类标签，类别为0和1。数据集仅包含一个训练集，共有337,297个样本，文件大小为239,165,844.38字节，下载大小为105,746,694字节。

The dataset includes multiple features such as id, sentence1, sentence2, label, template_name, template, rendered_input, and rendered_output. Among them, id is of integer type, sentence1 and sentence2 are of string type, label is a classification label with values 0 or 1. template_name, template, rendered_input, and rendered_output are also of string type. The dataset is divided into a training set, containing 337,297 samples.