hon9kon9ize/zoengjyutgaai_saamgwokjinji
收藏官方服务：
资源简介：
该数据集名为張悦楷三國演義，是一个粤语音频数据集。数据集通过对原始wav文件进行重新分割和对齐，确保音频与字幕文件（srt）的同步，并过滤了过短的样本。数据集包含音频文件、说话者信息、语言信息和转录文本，主要用于训练bert-vits2模型，支持44.1k的采样率。
This dataset contains Cantonese audio files with corresponding text transcriptions. The features of the dataset include file name, speaker, language, and transcription text. The dataset is divided into a training set with 307 samples. The original files of the dataset were not split correctly, so the author provided unsplit wav files and srt files, and re-split and aligned them. The audio files in the dataset have been resampled to support specific training models.
提供机构：
hon9kon9ize
