mteb/CMedQAv2-reranking

该数据集包含四个配置：文本数据集（corpus）、查询与文本匹配数据集（default）、查询数据集（queries）和高排名文本数据集（top_ranked）。文本数据集包含文档的ID和文本内容；查询与文本匹配数据集包含查询ID、文本ID和匹配得分；查询数据集包含查询的ID和文本内容；高排名文本数据集包含查询ID和一系列相关文本的ID。每个配置都有测试集分割，包含一定数量的样本和对应的文件大小。

The dataset consists of four configurations: text dataset (corpus), query-text matching dataset (default), query dataset (queries), and top-ranked text dataset (top_ranked). The text dataset includes document ID and text content; the query-text matching dataset includes query ID, text ID, and matching score; the query dataset includes query ID and text content; the top-ranked text dataset includes query ID and a sequence of IDs of related texts. Each configuration has a test split with a certain number of samples and corresponding file sizes.