AdoCleanCode/mfa_en_voxopouli_correct

该数据集是一个语音数据集，包含音频文件及其对应的转录文本、单词和音素的时间标注信息。具体特征包括：id（唯一标识符）、speaker_id（说话者标识符）、音频（采样率为16kHz）、转录文本（音频的文本内容）、单词列表（每个单词及其在音频中的起止时间）和音素列表（每个音素及其在音频中的起止时间）。数据集仅包含训练集，共有167,033个样本，总大小约为48.68GB。

This dataset is a speech dataset containing audio files along with their corresponding transcriptions, word-level and phoneme-level time alignments. The features include: id (unique identifier), speaker_id (speaker identifier), audio (with a sampling rate of 16kHz), transcript (text content of the audio), words list (each word with its start and end times in the audio), and phonemes list (each phoneme with its start and end times in the audio). The dataset only includes a training split with 167,033 samples and a total size of approximately 48.68GB.