Lots-of-LoRAs/task085_unnatural_addsub_arithmetic

该数据集名为task085_unnatural_addsub_arithmetic，属于文本生成任务类别，包含5200个训练样本、650个验证样本和650个测试样本。数据集的特征包括输入、输出和ID，均为字符串类型。数据集的主页和相关论文提供了更多详细信息。This dataset is crowdsourced, licensed under Apache 2.0, and primarily used for text generation tasks. It includes three fields: input, output, and ID, all of which are string types. The dataset is divided into training, validation, and test sets, containing 5200, 650, and 650 examples respectively. The dataset is named task085_unnatural_addsub_arithmetic, part of the Natural Instructions project.