SKIML-ICL/nq_retrieved

该数据集包含问题及其对应的答案和上下文信息。每个问题有一个唯一的ID（qid），问题文本（question），以及一个答案序列（answers）。上下文信息（ctxs）包括段落ID（pid）、排名（rank）、得分（score）、文本内容（text）和标题（title）。数据集分为训练集和验证集，训练集包含79168个样本，验证集包含8757个样本。数据集的下载大小为333571582字节，数据集总大小为588900147字节。

This dataset contains questions along with their corresponding answers and contextual information. Each question has a unique ID (qid), question text (question), and a sequence of answers (answers). The contextual information (ctxs) includes paragraph ID (pid), rank (rank), score (score), text content (text), and title (title). The dataset is divided into training and validation sets, with the training set containing 79,168 samples and the validation set containing 8,757 samples. The download size of the dataset is 333,571,582 bytes, and the total dataset size is 588,900,147 bytes.