lnm1p/Search-Gen-V-raw

Search-Gen-V-raw数据集包含了用于训练Search-Gen-V模型的原始、未处理的文件。数据集分为训练集和验证集，基于DL21、DL22和DL23数据集。数据集的结构包括字段如qid（查询ID）、query（查询内容）、block_text（文本块）、nuggets_list（信息片段列表）和reasoning（推理过程）。The Search-Gen-V-raw dataset contains the original, unprocessed files used for training the Search-Gen-V model. It consists of a training set and a validation set, based on the DL21, DL22, and DL23 datasets. The dataset structure includes fields such as qid (query ID), query (query content), block_text (text block), nuggets_list (list of information snippets), and reasoning (reasoning process).