A method and device for machine learning based on cloud computing

Abstract

A method of machine learning based on cloud computing, comprising: pulling a corresponding image according to a selected deep learning framework, creating a container according to the image and configuration information including the number of specified processor resources, and importing an input model file The storage path of the training data and the storage path of the training data are bound to the container; during the training process, the container is used to execute the model file obtained according to the storage path of the model file, and read according to the storage path of the training data The training data uses the read training data to train and execute the model established by the model file, and publishes the trained model. A device for machine learning based on cloud computing. This solution can eliminate the complexity in the process of using machine learning and lower the threshold for learning and using it.

Description

technical field [0001] The present invention relates to cloud computing technology, in particular to a method and device for machine learning based on cloud computing. Background technique [0002] Machine Learning (ML) is a multi-domain interdisciplinary subject that studies how computers simulate or realize human learning behaviors to acquire new knowledge or skills, and reorganize existing knowledge structures to continuously improve their own knowledge. performance. It is the core of artificial intelligence and the fundamental way to make computers intelligent. It is applied in all fields of artificial intelligence. It mainly uses induction and synthesis instead of deduction. [0003] Currently, most developers generally find machine learning much harder than it is, because the process of building and training a model and then deploying it into production is too complicated and too slow. The process of using machine learning mainly has the following parts: collecting a...

AI Technical Summary

Problems solved by technology

[0003] Currently, most developers generally feel that machine learning is much harder than it is because the process of building and training a model and then deploying it to production is too complex and slow

The process of using machine learning mainly has the following parts: collecting and preparing training data, selecting the algorithm and framework to be used, teaching the model how to make predictions through training, adjusting the model, integrating the model with the application and deploying several major steps, each step Both require extremely tedious manual work, combined with extensive expertise, access to large amounts of compute and storage, and time spent experimenting and optimizing every aspect of the process, factors that prevent most developers from using machine learning alone

Images