lion-ai/EskulapASRData

该数据集包含多个配置（anoni、bigos、gemini、human、youtube），每个配置都有音频和文本特征。音频特征采样率为16000Hz，文本特征为字符串类型。数据集分为训练集和测试集，每个配置的样本数量和大小各不相同。具体来说，anoni配置有13757个训练样本和1000个测试样本；bigos配置有20000个训练样本和1000个测试样本；gemini配置有1987个训练样本和220个测试样本；human配置有5661个训练样本和1000个测试样本；youtube配置有3710个训练样本和261个测试样本。

The dataset includes multiple configurations (anoni, bigos, gemini, human, youtube), each with audio and text features. The audio features have a sampling rate of 16000Hz, and the text features are of string type. The dataset is divided into training and test sets, with varying numbers of samples and sizes for each configuration. Specifically, the anoni configuration has 13757 training samples and 1000 test samples; the bigos configuration has 20000 training samples and 1000 test samples; the gemini configuration has 1987 training samples and 220 test samples; the human configuration has 5661 training samples and 1000 test samples; the youtube configuration has 3710 training samples and 261 test samples.