Management method, system and device applied to model conversion, and electronic equipment

Abstract

The embodiment of the invention provides a management method, system and device applied to model conversion, and electronic equipment. According to the embodiment of the invention, the second deep learning model matched with the target AI equipment is embodied in the target deep learning model, and the model description information used for describing the second deep learning model is also embodied, so that key information required when the target deep learning model is used for deep learning reasoning can be visually marked, fragment information such as AI equipment information, a deep learning reasoning framework and a model training framework corresponding to the target deep learning model does not need to be additionally recorded, and unified management of a model conversion process of a multi-AI equipment platform (a platform mixed with AI equipment of different application scenes) is realized; and furthermore, when the model description information used for describing the second deep learning model is reflected in the target deep learning model, the target deep learning model can be upgraded based on the model description information subsequently.

Description

technical field [0001] The present application relates to computer technology, in particular to a management method, system, device and electronic equipment applied to model conversion. Background technique [0002] AI devices can be used for deep learning training and deep learning inference. The so-called deep learning training refers to sending a set of training data sets (data with labels) into the neural network, and adjusting each network layer in the neural network according to the difference between the output information of the neural network and the preset expected output information. The connection weights, and finally train an ideal neural network. The so-called deep learning reasoning refers to the use of trained neural networks for data reasoning, such as: recognizing images, recognizing and recording spoken language, detecting diseases in blood, or recommending clothes that match someone's style, etc. As an embodiment, the AI ​​device may be a device deploye...

AI Technical Summary

Problems solved by technology

[0004] At present, when deploying a deep learning model on an AI device, the following problems often occur: the deep learning model does not match the AI ​​device (for example, the deep learning model does not match the deep learning reasoning framework corresponding to the AI ​​device, and cannot be based on the corresponding depth of the AI ​​device. learning reasoning framework for learning reasoning, etc.)

Once the deep learning model does not match the AI ​​device, the deep learning model cannot rely on the deep learning reasoning framework corresponding to the AI ​​device for learning reasoning, etc.

Images