AdoCleanCode/giga_mfa_v0_wprev
收藏Hugging Face2025-12-14 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/AdoCleanCode/giga_mfa_v0_wprev
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含音频及其相关转录信息，特征包括序列、完整转录、原始转录、移除的单词、标注的音素、实际的音素、xcodec2标记、左音频、移除的音频、右音频和完整音频。音频的采样率为16000Hz。数据集分为9个批次，每个批次包含20000个样本，总样本量为180000。
This dataset contains audio and related transcription information, with features including sequence, full transcription, original transcription, removed words, annotated phonemes, actual phonemes, xcodec2 tokens, left audio, removed audio, right audio, and full audio. The audio sampling rate is 16000Hz. The dataset is divided into 9 batches, each containing 20000 samples, with a total sample size of 180000.
提供机构：
AdoCleanCode
