Bilingual Audio Dataset

该数据集包含了从互联网收集的高质量中英双语音频数据，其中包含25,000小时中文音频和9,000小时英文音频。该数据集经过处理，利用自动语音识别技术生成了伪标签，并根据文本语义和语音停顿添加了标点符号，从而可以灵活控制停顿。总规模达到34,000小时（包括25,000小时中文和9,000小时英文），适用于文本到语音合成和声音克隆任务。

This dataset contains high-quality Mandarin-English bilingual audio data collected from the Internet, including 25,000 hours of Chinese audio and 9,000 hours of English audio. The dataset has been processed, where automatic speech recognition technology is used to generate pseudo-labels, and punctuation marks are added based on text semantics and speech pauses to enable flexible pause control. The total scale reaches 34,000 hours (including 25,000 hours of Chinese audio and 9,000 hours of English audio), which is suitable for text-to-speech synthesis and voice cloning tasks.