marcov/discovery_discovery_promptsource

该数据集包含两个句子（sentence1和sentence2）以及它们之间的关系标签（label），标签共有173个类别，每个类别代表一个特定的连接词或短语。数据集还包含索引（idx）、模板名称（template_name）、模板（template）、渲染输入（rendered_input）和渲染输出（rendered_output）等特征。数据集分为训练集、测试集和验证集，分别包含7830000、435000和435000个样本。数据集的下载大小为4045772130字节，总大小为21559924610字节。

The dataset contains two sentences (sentence1 and sentence2) and a relationship label (label) between them, with 173 categories, each representing a specific connector or phrase. The dataset also includes features such as index (idx), template name (template_name), template (template), rendered input (rendered_input), and rendered output (rendered_output). The dataset is divided into training, test, and validation sets, containing 7830000, 435000, and 435000 samples respectively. The download size of the dataset is 4045772130 bytes, and the total size is 21559924610 bytes.