AdoCleanCode/clean_dirty_dac_complex_v007
收藏Hugging Face2025-09-12 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/AdoCleanCode/clean_dirty_dac_complex_v007
下载链接
链接失效反馈官方服务：
资源简介：
这是一个包含语音数据的 dataset，每个样本都包括utterance ID、时长、噪声类型、清理后的token列表、带噪声的token列表、清理后的字符串token、带噪声的字符串token、原始文件名和原始采样率等信息。数据集分为五个批次，每个批次包含25000个示例。
This is a dataset containing speech data, with each sample including utterance ID, duration, noise type, list of cleaned tokens, list of noisy tokens, cleaned string tokens, noisy string tokens, original file name, and original sampling rate, etc. The dataset is divided into five batches, each containing 25,000 examples.
提供机构：
AdoCleanCode
