badrex/ethiopian-speech

该数据集包含了五种语言的语音数据，分别为afaan_oromo、amharic、sidama、tigrinya和wolaytta。每种语言配置都包括音频ID、说话者ID、音频文件、音频时长、转录文本、性别、年龄组和语言区域等信息。数据集分为验证集、测试集和训练集，提供了每个集合的示例数量和大小。

The dataset includes speech data in five languages: afaan_oromo, amharic, sidama, tigrinya, and wolaytta. Each language configuration contains information such as audio ID, speaker ID, audio file, audio duration, transcription, gender, age group, and locale. The dataset is split into validation, test, and training sets, with the number of examples and size of each set provided.