deu05232/promptriever-RQ1-diff_version_fair
收藏资源简介：
该数据集是一个用于信息检索或自然语言处理任务的数据集，包含查询ID、查询文本、正面和负面段落、指令等特征。正面段落包括文档ID、解释、FollowIR评分、联合ID、文本和标题；负面段落包括文档ID、文本和标题。此外，还包含仅指令、仅查询、是否包含指令等字段，以及新的负面段落列表。数据集适用于训练和评估查询-文档匹配模型或指令遵循模型，支持大规模数据处理。
This dataset is designed for information retrieval or natural language processing tasks, featuring query ID, query text, positive and negative passages, instructions, and other attributes. Positive passages include document ID, explanation, FollowIR score, joint ID, text, and title; negative passages include document ID, text, and title. It also contains fields such as only instruction, only query, has instruction, and a list of new negative passages. The dataset is suitable for training and evaluating query-document matching models or instruction-following models, supporting large-scale data processing.
