vector-institute/hotpotqa

这是一个包含问题和预期答案的数据集，适用于某种问答系统。数据集中的每个问题都有一个唯一的标识符(id)，并且可能包含问题类型(type)和难度级别(level)。此外，每个问题都可能有一系列支持事实和支持性上下文，这些上下文包含标题和一系列句子。数据集分为训练集、验证集和测试集，分别用于训练、验证和测试模型。

This is a dataset containing questions and expected answers, suitable for a question-answering system. Each question in the dataset has a unique identifier (id) and may include the question type (type) and difficulty level (level). Additionally, each question may have a series of supporting facts and supporting contexts, which include titles and a series of sentences. The dataset is divided into training, validation, and test sets, respectively used for training, validation, and testing of models.