ocisd4/openslr_asr

该数据集包含多个语言配置（如阿拉伯语、孟加拉语、法语、德语、西班牙语、土耳其语等），每个配置包含音频数据和文本消息数据。音频数据为音频文件，文本消息数据包含内容和角色信息。数据集的主要用途可能是语音识别或语音合成任务，因为每个配置都包含音频和对应的文本内容。每个配置的训练集大小和样本数量各不相同。

The dataset contains multiple configurations, each corresponding to a different language (e.g., Arabic, Bengali, French, German, Spanish, Turkish, etc.). Each configuration includes audio data and text message data. The audio data is in audio file format, and the text message data includes content and role information. The primary use of the dataset is likely for speech recognition or speech synthesis tasks, as each configuration contains audio and corresponding text content. The training set size and number of samples vary for each configuration.