mlfoundations-dev/oh_v3.1_wo_caseus_custom

该数据集包含两个主要特征：conversations和source_label_exact。conversations特征是一个列表，包含from和value两个字段，均为字符串类型。source_label_exact特征是一个字符串序列。数据集只有一个训练集（train），包含998,830个例子，总大小为1,993,061,483字节。下载大小为1,047,559,675字节。

The dataset contains two main features: conversations and source_label_exact. The conversations feature is a list containing two fields: from and value, both of which are of string type. The source_label_exact feature is a sequence of strings. The dataset has only one training set (train) with 998,830 examples, totaling 1,993,061,483 bytes in size. The download size is 1,047,559,675 bytes.