mteb/MIRACLRetrieval_th_top_250_only_w_correct_v1

该数据集包含两个部分：default和queries。default部分具有query-id，corpus-id和score字段，主要用于记录查询、文档及其相关得分。queries部分具有_id和text字段，主要用于存储查询文本。数据集的训练部分分别包含default和queries的示例，可用于训练相关模型，如信息检索系统。The dataset consists of two parts: default and queries. The default part has fields for query-id, corpus-id, and score, which are used to record queries, documents, and their associated scores. The queries part has fields for _id and text, which are used to store the query texts. The training part of the dataset includes examples from both the default and queries parts, which can be used to train relevant models, such as information retrieval systems.