Mollel/Swahili-NLi-Pair

该数据集包含三个主要特征：语言、锚点和正面样本，所有特征的数据类型均为字符串。数据集被分为训练集、开发集和测试集，其中训练集包含314,315个样本，开发集包含6,808个样本，测试集包含6,831个样本。数据集的下载大小为26,852,280字节，总大小为49,285,491.5字节。数据文件的路径分别为data/train-*、data/dev-*和data/test-*。

The dataset contains three main features: language, anchor, and positive, all of which are of string data type. The dataset is divided into training, development, and test sets, with the training set containing 314,315 samples, the development set containing 6,808 samples, and the test set containing 6,831 samples. The download size of the dataset is 26,852,280 bytes, and the total size is 49,285,491.5 bytes. The data file paths are data/train-*, data/dev-*, and data/test-*.