CIDAR

该数据集专为阿拉伯语自然语言处理任务设计，包含了指令与输出结果的对，旨在帮助模型理解和生成不同类型任务下的回应。每条记录都包含了独特的阿拉伯语指令及其对应的输出结果，这样的设计有助于提升模型遵循指令的能力。该数据集的主要任务是指导遵循。

This dataset is specifically developed for Arabic natural language processing (NLP) tasks. It comprises instruction-output pairs, with the objective of assisting models in understanding and generating responses across various task types. Each entry contains a unique Arabic instruction and its corresponding output, a design that helps improve the model's instruction-following capability. The primary task of this dataset is instruction following.