Language model tool calling method and device, computer equipment and storage medium

The invention relates to a language model tool calling method and device, computer equipment and a storage medium. The method comprises the following steps: constructing training data for fine tuning of a language model; the training data comprises prompt statements, a local function library, a question set and an answer set; the question set serves as input of user roles in the training data, function names and parameters marked in the answer set, prompt statements and subsets of the local function library serve as input of assistant roles in the training data, and functions marked in the answer set are called and returned to serve as input of tool roles in the training data; performing fine tuning training on the language model by using the training data; wherein during fine tuning training, the language model does not calculate loss values of a user role and a tool role in back propagation; and reasoning according to a problem input by a user by adopting the language model subjected to fine tuning training