jan-hq/instruction-convert-audio-whispervq-llama3.2

该数据集包含三个主要字段：transcript、text和conversations。transcript和text字段均为字符串类型，可能包含文本数据。conversations字段是一个列表，每个元素包含content和role两个子字段，分别表示对话内容和角色。数据集被分割为训练集，包含3,707,675个示例，总大小为34,596,615,896字节。

This dataset includes three main fields: transcript, text, and conversations. Both transcript and text fields are of string type, likely containing textual data. The conversations field is a list, with each element containing two subfields: content and role, representing the dialogue content and role, respectively. The dataset is split into a training set, containing 3,707,675 examples, with a total size of 34,596,615,896 bytes.