Tino3141/mls_sub1-seed3603964567

该数据集包含多个音频相关的特征字段，如信号噪声比（snr），以及两个不同语言的音频源（src1和src2）。每个样本可能包含两个说话者的信息（speaker1和speaker2）。训练集和测试集的大小分别为30000和3000个样本。但是具体的数据集描述在README中未提供，因此无法给出详细的数据集中文描述。

The dataset includes multiple audio-related feature fields, such as signal-to-noise ratio (snr) and two audio sources in different languages (src1 and src2). Each sample may contain information about two speakers (speaker1 and speaker2). The training and test sets contain 30,000 and 3,000 samples respectively. However, the specific dataset description is not provided in the README, so no detailed Chinese description of the dataset can be given.