amuvarma/mls-train-200

该数据集包含音频值、转录ID、标签和输入ID等特征，主要用于处理与音频相关的自然语言处理任务。数据集包含200,000个训练样本，总大小为480,328,800,000字节，下载大小为295,360,838,917字节。

This dataset includes features such as audio values, transcript IDs, labels, and input IDs, primarily used for natural language processing tasks related to audio. The dataset contains 200,000 training samples, with a total size of 480,328,800,000 bytes and a download size of 295,360,838,917 bytes.