chiyuanhsiao/audio_no-replay-15_trivia_qa-audio

该数据集包含了问题及其相关信息的记录，每个记录包括问题文本、问题ID、问题来源、相关实体页面信息（如文档来源、文件名、标题和维基百科上下文）、搜索结果信息（如描述、文件名、排名、标题、URL和搜索上下文）、答案信息（包括别名、标准化别名、匹配的维基实体名称、标准化的匹配维基实体名称、标准化值、类型和值）、问题单位、响应交错的文本、响应文本、响应令牌和响应语音。数据集包含一个验证集，其大小为1000个示例，总共占用的字节数为1,088,453,056字节。

The dataset consists of records containing questions and their related information, each including the question text, question ID, question source, related entity page information (such as document source, filename, title, and Wikipedia context), search result information (such as description, filename, rank, title, URL, and search context), answer information (including aliases, normalized aliases, matched Wikipedia entity name, normalized matched Wikipedia entity name, normalized value, type, and value), question unit, interleaved response text, response text, response tokens, and response speech. The dataset includes a validation set with 1000 examples, totaling 1,088,453,056 bytes in size.