mlfoundations-dev/oh_v1.2_opengpt_x2

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，数据类型均为字符串。数据集还包含一个分片ID字段。数据集分为一个训练集，包含895,111个样本，总大小为1,646,097,412字节。下载大小为890,617,135字节。数据文件路径在默认配置下指定为train-*。

The dataset contains conversation data, with each conversation including fields for the source and content, both of which are of string type. The dataset also includes a shard ID field. It is divided into a training set containing 895,111 samples, with a total size of 1,646,097,412 bytes. The download size is 890,617,135 bytes. The data file path is specified as train-* under the default configuration.