Model online method and device

Abstract

The invention provides a model online method and device, relates to the field of communication, and can improve the model online efficiency. The method comprises the following steps: receiving a datamodeling instruction input by a user, wherein the data modeling instruction is used for requesting to create a machine learning model; creating a machine learning model according to the data modelinginstruction, and generating a serialized model file and a model description file of the machine learning model with the serialized model file being a file in a serialized format of the machine learning model and the model description file being a file representing attributes of the machine learning model; receiving a model deployment instruction input by a user; generating a new service script according to the universal service script template and the model description file; generating a mirror image of the model according to the new service script and the serialized model file; and generatingmodel service POD according to the mirror image of the model, so that the machine learning model is deployed online.

Description

technical field [0001] The present application relates to the communication field, and in particular to a method and device for online model. Background technique [0002] As a branch of artificial intelligence, machine learning is an important business support method in today's big data era. Machine learning can be applied to various application scenarios and can significantly improve the performance indicators of enterprises. For example, machine learning technology can be applied to predicting advertising clicks, product recommendations, image recognition, anomaly detection and other scenarios, and has produced significant beneficial effects. [0003] Usually, the development process of a machine learning model includes data exploration, feature extraction, model creation, and model online. Usually, the operations of each link of model development are completed by different developers. For example, developer A is responsible for model creation, and developer B is respons...

AI Technical Summary

Problems solved by technology

[0004] However, in the process of model launch, since each model is created in a different way, the developer responsible for model deployment and launch must not only be familiar with the professional knowledge in the field of machine learning, but also need to understand the characteristics of the model to be deployed in advance, for example, need to understand The algorithm of the model to be deployed and the data source of the model to be deployed, etc., and the model is deployed online according to the characteristics of the model to be deployed, which consumes the time and energy of the developer, and the efficiency of the model online is low

Images