ryusangwon/deprecated_nq_bm25_document_top_20_atom

该数据集包含id、question、answers和top20四个主要特征。其中，id和question为字符串类型，answers为字符串序列，top20是一个包含atom、bm25_score、docID和rank四个字段的列表。数据集的分割为nq，包含3610个示例，文件大小为26979845字节，下载大小为11185838字节。

The dataset includes four main features: id, question, answers, and top20. Among them, id and question are of string type, answers is a sequence of strings, and top20 is a list containing four fields: atom, bm25_score, docID, and rank. The dataset is split into nq, containing 3610 examples, with a file size of 26979845 bytes and a download size of 11185838 bytes.