seongil-dn/mteb_msmarco_naive

该数据集包含查询及其相关正例和负例文本信息。每个示例包括一个唯一的标识符id，查询内容query，以及与之相关的正例列表positives和负例列表negatives。正例和负例列表中的每个元素都包含一个文本标识符id，相关性得分score和文本内容text。负例列表还包含一个topk_rank字段，表示该负例在所有负例中的排名。训练集包含502,939个示例，数据集总大小为3,995,504,072字节。

The dataset consists of queries and their associated positive and negative text examples. Each example includes a unique identifier id, the query content query, and a list of related positive examples positives and a list of negative examples negatives. Each element in the positive and negative lists contains a text identifier id, a relevance score score, and the text content text. The negative list also includes a topk_rank field, indicating the rank of the negative example among all negative examples. The training set contains 502,939 examples, and the total size of the dataset is 3,995,504,072 bytes.