Model calling method, apparatus and device, and computer readable storage medium

Abstract

The invention provides a model calling method and device, equipment and a computer readable storage medium. The method comprises the steps: determining a set of models corresponding to a plurality ofapplication scenes; and when in an application scene, determining a model corresponding to the application scene in the set of models, calling the model, and keeping the called model in a starting state.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a model calling method, device, equipment and computer-readable storage medium. Background technique [0002] With the development of science and technology, more and more application scenarios and models are used. In the existing technology, a certain application scenario or a certain function is bound to a fixed model or algorithm. When a certain application scenario needs to call a certain model , a specific model can be invoked by triggering the model or algorithm. When the usage scenario changes, it is necessary to call or close the model in a targeted manner, and the model is not reused. [0003] Recently, with the continuous development of machine learning, the deep learning neural network model has been widely used. The deep learning neural network model has the advantages of short calculation time and stable processing results, but the model occupies more sp...

AI Technical Summary

Problems solved by technology

[0003] Recently, with the continuous development of machine learning, the deep learning neural network model has been widely used. The deep learning neural network model has the advantages of short calculation time and stable processing results, but the model occupies more space and resources. It takes a long time to start, so it is not suitable for frequent calls

Images