A computing service system for machine learning and a method for machine learning

Abstract

The invention discloses a computing service system for machine learning and a method for machine learning, and relates to the field of computer information processing. The system comprises an online experiment module which is used for providing basic services for machine learning codes of a user, wherein the basic services comprise data maintenance, data editing and code operation; a training andtesting module used for training and testing the machine learning task submitted by the user, and realized by adopting a Docker container; and a remote debugging module used for providing a remote debugging function for the training test module for a user. According to the computing service system for machine learning and the method for machine learning disclosed by the invention, the environmentdeployment and data selection of the machine learning task can be simply and quickly carried out, the special requirement for the machine learning task is met, and the isolation of a machine learningexperiment environment is ensured.

Description

technical field [0001] The present invention relates to the field of computer information processing, in particular to a system for computing services for machine learning and a method for machine learning. Background technique [0002] With the advent of the era of big data and the continuous development of artificial intelligence technology, more and more engineers are devoted to the research of machine learning algorithms. This depends not only on the individual's theoretical basis, but also on the ability to operate on high-performance hardware (such as GPU) resources. This is because the current machine learning research, especially deep learning, often involves more complex environment configuration, parameter debugging, etc., which brings great challenges to inexperienced developers. After the experiment environment is changed, the developer needs to repeat the process of setting up the environment and compiling the code, resulting in a waste of time and a decrease i...

AI Technical Summary

Problems solved by technology

This method of running services in containers is not suitable for machine learning computing platforms, because machine learning models are not unified, and modular services in containers cannot be used to meet model training / testing requirements

[0004] In the existing technology, machine learning services can be provided, for example, through general demand cloud platforms and cloud platforms based on virtual machines. It does not bring convenience to the research and development of machine learning algorithms

In the cloud platform based on the virtual machine, the resource usage rate of the virtual machine technology is high, and the performance overhead is large.

Images