Lots-of-LoRAs/task1631_openpi_answer_generation

该数据集名为Natural Instructions，主要用于文本生成任务。数据集包含输入、输出和ID三个特征，分别存储为字符串类型。数据集分为训练集、验证集和测试集，分别包含2392、299和299个样本。数据集的创建者和语言创建者均为众包，语言为英语，许可证为Apache-2.0。更多详细信息可以参考相关论文和数据集主页。

The dataset is named Natural Instructions and is primarily used for text generation tasks. It includes three features: input, output, and ID, all stored as string types. The dataset is divided into training, validation, and test sets, containing 2392, 299, and 299 samples respectively. The creators and language creators of the dataset are crowdsourced, the language is English, and the license is Apache-2.0. For more details, refer to the related papers and the dataset homepage.