sarpba/CV22-hun-cleaned

这是一个包含音频文件和同名.txt转录文件的数据集。字段包括：id（基于文件名的标识符）、audio（可以通过datasets加载的音频）、text（完整的UTF-8转录）、text_vibevoice（带有Speaker 1: 前缀的转录，例如VibeVoice格式）、relpath（相对于数据根的相对路径）、duration（大致长度）、sample_rate（采样率）、channels（通道数）、bitrate_kbps（估计比特率）、age（年龄）、gender（性别）和accents（口音）。

This dataset consists of audio files and corresponding .txt transcripts. Fields include: id (filename-based identifier), audio (audio loadable by datasets), text (complete UTF-8 transcript), text_vibevoice (transcript with Speaker 1: prefix, for example VibeVoice format), relpath (relative path to the data root), duration (approximate length), sample_rate (sampling rate), channels (number of channels), bitrate_kbps (estimated bitrate), age, gender, and accents.