AdoCleanCode/french-multi-mfa_train_v1_preview

该数据集是一个包含音频和文本特征的数据集，主要用于语音处理相关的研究。数据集包含100个样本，每个样本包括左右声道音频、移除的音频、完整音频、完整标记音频以及对应的音频持续时间。此外，还包括完整转录文本、原始转录文本、移除的单词、标注音素、实际音素和xcodec2标记。音频的采样率为16000Hz。数据集总大小约为96.7MB。

This dataset is a collection of audio and text features, primarily used for speech processing research. It contains 100 samples, each including left and right channel audio, removed audio, full audio, full tokens audio, and corresponding audio durations. Additionally, it includes full transcription text, original transcription text, removed words, annotated phonemes, actual phonemes, and xcodec2 tokens. The audio sampling rate is 16000Hz. The total dataset size is approximately 96.7MB.