seongil-dn/seongil-dn_mteb-gooaq_naive

该数据集包含了查询及其相关正例和负例文本信息。每个样本由一个唯一标识符(id)、一个查询(query)、一组正例(positives)和一组负例(negatives)组成。正例和负例都包括文本(text)、分数(score)和标识符(id)，负例还包括topk排名(topk_rank)。数据集仅包含训练集(train split)，共有100万个样本，数据大小为约6.2GB。

The dataset consists of queries and their associated positive and negative text examples. Each sample includes a unique identifier (id), a query (query), a set of positives (positives), and a set of negatives (negatives). Both positives and negatives contain text (text), score (score), and identifier (id), while negatives also include the topk rank (topk_rank). The dataset contains only a training set (train split) with a total of 1 million samples, and the data size is approximately 6.2GB.