amuvarma/smoltalk-audio-speech-raw

该数据集包含多个对话文本和音频数据，主要特征包括对话双方的文本内容（human和assistant）以及助手的音频数据（assistant_speech）。此外，还包含一些序列数据（ass_facodec）。数据集仅包含一个训练集，共有2260个样本，总大小为1558423272.5字节。

This dataset contains multiple dialogue texts and audio data, with main features including text content from both sides of the dialogue (human and assistant) and audio data from the assistant (assistant_speech). Additionally, it includes some sequence data (ass_facodec). The dataset contains only a training set with 2260 samples, totaling 1558423272.5 bytes in size.