TaoLi

As ChatGPT attracts the attention of the whole society and various large language models compete to appear, natural language processing tasks in general fields have achieved great success and attracted widespread attention in the international Chinese education field.
International Chinese education professionals have started discussions on large models: whether large models can provide appropriate language expressions based on the learner's level, or provide detailed answers based on the learner's questions, thereby assisting or even serving as a learning tool to a certain extent. Partner, language teacher? However, the current large models in general fields still have limited effect in vertical fields.
In order to solve the above problems, we have comprehensively launched "Taoli" 1.0, a large model suitable for the international Chinese education field, a model that has been additionally trained on data in the international Chinese education field.
We have built an international Chinese education resource library based on more than 500 international Chinese education textbooks and teaching aids, Chinese proficiency test questions, and Chinese learner dictionaries currently circulating in the international Chinese education field. We set up various forms of instructions to make full use of knowledge, constructed a high-quality international Chinese education question and answer data set with a total of 88,000 items, and used the collected data to fine-tune the instructions of the model, so that the model can learn to apply international Chinese education knowledge. ability to apply to specific scenarios.