acul3/laion-2b-indonesia-subset

该数据集包含多个特征，如URL、文本内容、宽度、高度、相似度、语言、水印概率、不安全内容概率、英文文本和预测值。数据集分为一个训练集，包含21,909,328个样本，总大小为5,678,773,557字节。下载大小为4,290,558,447字节。数据集的配置为默认配置，数据文件路径为data/train-*。

This dataset includes multiple features such as URL, text content, width, height, similarity, language, watermark probability, unsafe content probability, English text, and prediction values. The dataset is divided into a training set containing 21,909,328 samples, with a total size of 5,678,773,557 bytes. The download size is 4,290,558,447 bytes. The dataset configuration is default, and the data file path is data/train-*.