Text processing model training method and apparatus, text processing method and apparatus, and question and answer processing method and apparatus

Abstract

The embodiments of the present application relate to the technical field of artificial intelligence, and disclose a text processing model training method and apparatus, a text processing method and apparatus, and a question and answer processing method and apparatus. The text processing model training method comprises: acquiring a plurality of training samples, the training samples at least comprising a text sample; and using the plurality of training samples to distill a large language model as a teacher model, so as to obtain a text processing model as a student model, wherein the parameter scale of the text processing model is smaller than that of the large language model, and during the distillation process, program-of-thought (POT) information and chain-of-thought (COT) information predicted by the large language model for the text sample are used to guide the training of the text processing model. The present application uses the POT information and the COT information predicted by the large language model to jointly guide the training of the text processing model, improving the accuracy of model training.

Images