mbzuai-ugrip-statement-tuning/pawsx_instruction

该数据集是一个多语言数据集，包含德语（de）、英语（en）、西班牙语（es）、法语（fr）和中文（zh）五种语言配置。每个语言配置下都有指导和输出两种字符串类型的特征，且都只有一个训练集分割，每个训练集包含49401个示例。

The dataset consists of five language configurations (de, en, es, fr, zh), each with two string-type features: instruction and output. The dataset is split into a training set for each language, with each languages training set containing 49401 examples, and the size of the training sets and download sizes vary by language.