TeoGchx/BEAT_HML3D_english

该数据集包含语音、动作、文本和元数据四个主要特征。语音数据为float32类型的序列，动作数据为float32类型的序列的序列，文本数据为string类型的序列，元数据包含duration、name、num_frames、num_words和speaker等字段。数据集分为train、test、val和additional四个分割，每个分割的字节数和示例数也被详细列出。数据集的下载大小为21023501958字节，总大小为20984449331字节。

The dataset includes four main features: speech, motion, text, and meta_data. The speech feature contains a sequence of float32, the motion feature is a nested sequence of float32, and the text feature contains a sequence of strings. The meta_data feature is a struct containing duration, name, number of frames, number of words, and speaker information. The dataset is divided into four parts: train, test, val, and additional, each with corresponding bytes and number of examples. The download size and actual size of the dataset are 21023501958 bytes and 20984449331 bytes, respectively.