illuin-cde/squad-chunked-par-500

该数据集包含两部分：documents和queries。documents部分包含文档的chunk信息，包括chunk的标识符(chunk_id)、文本内容(chunk)和偏移量(offset)。queries部分包含查询及其对应的答案，包括查询的标识符(chunk_id)、查询文本(query)和答案(answer)。数据集分为训练集和验证集，以便进行模型训练和验证。

The dataset consists of two parts: documents and queries. The documents part includes chunk information of the documents, including chunk identifier (chunk_id), text content (chunk), and offset. The queries part includes queries and their corresponding answers, including query identifier (chunk_id), query text (query), and answer (answer). The dataset is split into training and validation sets for model training and validation purposes.