ylacombe/emilia-subset-annotated

该数据集包含多个特征字段，如__key__、__url__、dnsmos、duration、id、language、speaker、text、wav等，涵盖了音频文件的元数据、语音质量评估指标、语音特征等。数据集主要包含一个名为en的分割，包含3,387,817个样本，总大小为3,636,409,489字节。数据集的主要用途可能与语音质量评估、语音特征分析等相关。

The dataset contains multiple feature fields such as __key__, __url__, dnsmos, duration, id, language, speaker, text, wav, etc., covering metadata of audio files, speech quality assessment metrics, and speech features. The dataset primarily includes a split named en with 3,387,817 samples and a total size of 3,636,409,489 bytes. The main purpose of the dataset may be related to speech quality assessment and speech feature analysis.