ai-bond/ru-alpaca-logic

该数据集是一个文本生成任务的数据集，使用MIT许可证，语言为俄语。数据集包含三个主要特征：instruction（指令）、input（输入）和output（输出），均为字符串类型。数据集配置包括默认配置，数据文件分为训练集和测试集。该数据集是对MERA-evaluation和Vikhrmodels数据集的重新格式化，以适应alpaca-friendly格式。

This is a dataset formatted in the alpaca-friendly style, derived from MERA-evaluation and Vikhrmodels. The datasets task category is text generation, using Russian, and is licensed under MIT. The dataset includes three features: instruction, input, and output, all of which are string types. The dataset configuration includes a default setup with paths for training and testing data files.