Method, system and electronic equipment for task scheduling based on predictive model

Abstract

The invention provides a task scheduling method based on a prediction model. The method comprises the steps that S1, a master controller Master receives a prediction request submitted by a user; S2, respectively calculating the number of prediction requests received by the task queue by the loading counter and the canceling counter; Determining that the prediction request is stored in a task queueaccording to the numerical value of the loading counter; S3, the Worker executes the loading request in the task queue to load the prediction model, executes the prediction task in the task queue, and obtains a prediction result corresponding to the prediction task through calculation of the prediction model; And S4, the Worker executes the cancel requests in the task queue, controls the numerical value of the cancel counter according to the number of the executed cancel requests, and controls unloading of the prediction model according to the numerical value of the cancel counter. The invention further provides a corresponding task scheduling system based on the prediction model and electronic equipment, and the system and the method have the advantages of high task execution efficiency,low mutual interference and the like.

Description

【Technical field】 [0001] The invention relates to the field of artificial intelligence, in particular to a task scheduling method, system and electronic equipment based on a prediction model. 【Background technique】 [0002] In the application of artificial intelligence, people often need to load models at the same time to provide prediction services. For example, a retail e-commerce company needs to make related recommendations and sort products in different formats every day. The priority of these prediction tasks is often the same. In the existing prediction system, once users submit prediction requests at the same time, the model will be loaded repeatedly. In this way, the working efficiency of the whole forecasting system is low. In addition, after the prediction task is completed, the model is still loaded and not released in time, which reduces resource utilization. 【Content of invention】 [0003] In order to overcome the shortcomings of the current predictive task...

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the shortcomings of current predictive task scheduling, the present invention proposes a task scheduling method, system and electronic equipment based on predictive models

Images