ylacombe/globe_processed

该数据集包含音频、名称、原始文本、语言、数据集来源、分割类型、说话者、持续时间、简短描述、掩码描述、文本描述和文本等多个特征。数据集分为测试集和训练集，测试集包含14304个样本，训练集包含567421个样本。数据集的下载大小为92480481986字节，总大小为93039849030字节。

The dataset includes multiple features such as audio, name, original text, language, dataset source, split type, speaker, duration, short captions, masked captions, text description, and text. The dataset is divided into a test set and a training set, with the test set containing 14,304 samples and the training set containing 567,421 samples. The download size of the dataset is 92,480,481,986 bytes, and the total size is 93,039,849,030 bytes.