mlfoundations-dev/oh_v1.3_metamath_x4

该数据集包含对话数据，每个对话包含来源和内容两个字段，均为字符串类型。数据集还包括一个字符串类型的shard_id字段。数据集有一个训练集分割，包含985,954个示例，总大小为1,605,683,645字节。下载大小为871,391,236字节。

The dataset contains conversation data, with each conversation including two fields: source and content, both of string type. The dataset also includes a shard_id field of string type. There is one training set split containing 985,954 examples, with a total size of 1,605,683,645 bytes. The download size is 871,391,236 bytes.