mlfoundations-dev/oh_v1.3_unnatural_instructions_x.5

该数据集包含对话内容，每个对话由from和value两个字段组成，均为字符串类型。数据集还包含一个分片ID（shard_id）字段。数据集分为训练集，包含852,524个样本，总大小为1,439,174,102字节。默认配置下的数据文件路径为data/train-*。

This dataset contains conversation content, with each conversation consisting of two fields: from and value, both of which are string types. The dataset also includes a shard ID (shard_id) field. The dataset is divided into a training set, containing 852,524 samples, with a total size of 1,439,174,102 bytes. The data file path under the default configuration is data/train-*.