ChristianYang/librivox-tracks-vad
收藏Hugging Face2026-04-23 更新2026-04-26 收录
下载链接：
https://hf-mirror.com/datasets/ChristianYang/librivox-tracks-vad
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是从pykeio/librivox-tracks经过单阅读器过滤和Silero VAD分割处理得到的。数据文件包括训练数据的parquet文件，每行存储了源元数据、单声道WAV音频数据（audio_bytes）和采样率（sampling_rate）。数据收集直到达到全局总音频预算（默认为(2442/5994)*3600 * multiplier秒）为止。
This dataset is produced from `pykeio/librivox-tracks` with single-reader filtering and Silero VAD segmentation. The data files include training data in parquet format, each row storing source metadata, mono WAV audio data (`audio_bytes`), and `sampling_rate`. Data collection continues until a global total audio budget (default `(2442/5994)*3600 * multiplier` seconds) is reached.
提供机构：
ChristianYang
