MYJOKERML/chinese-dialogue-speech-dataset

这是一个包含46,080个多轮对话的大规模中文多轮对话语音合成数据集，覆盖文学问答、自然对话和诗词文化等多个领域。数据集包含约275,000个WAV格式音频文件，音频总时长约1,000-1,200小时，采样率为16kHz。This is a large-scale Chinese multi-turn dialogue speech synthesis dataset containing 46,080 multi-turn dialogues, covering fields such as literary Q&A, natural dialogue, and poetry culture. The dataset includes approximately 275,000 WAV format audio files, with a total audio duration of about 1,000-1,200 hours, and a sampling rate of 16kHz.