Beijuka/clean_testcommon_voice

该数据集包含多个特征，如客户端ID、路径、音频、转录文本等。音频数据的采样率为16000。此外，数据集还包含用户的投票信息（赞成票和反对票）、用户信息（年龄、性别、口音、地区）以及音频片段信息（片段和变体）。测试集包含62个样本，总大小为2510757字节。

This dataset includes multiple features such as client_id, path, audio, transcription, etc. The audio feature contains audio data with a sampling rate of 16000. Additionally, the dataset includes voting information (up_votes and down_votes), user information (age, gender, accent, locale), and audio segment information (segment and variant). The test split contains 62 examples with a total size of 2510757 bytes.