elliotjy/train_2M_CN

包含约200万条由BELLE项目生成的中文指令数据，用于文本到文本生成任务。数据样例展示了如何将多个句子组合成一个有意义的段落。数据集包含三个字段：instruction（指令）、input（输入，本数据集均为空）和output（输出）。使用限制规定仅允许用于研究目的，不得用于商业或其他有害用途。

Contains approximately 2 million Chinese instruction data generated by the BELLE project, designed for text-to-text generation tasks. The example demonstrates how to combine multiple sentences into a meaningful paragraph. The dataset includes three fields: instruction, input (empty in this dataset), and output. Usage restrictions specify that it is only allowed for research purposes and not for commercial or other harmful uses.