Multi-machine cluster-based neural network training method and system

A neural network training and clustering technology, applied in the field of neural network training, can solve the problems of low single-machine training efficiency and low model iteration efficiency, and achieve the effect of speeding up data communication and improving the speed.

Active Publication Date: 2022-06-10
BEIJING KUANGSHI TECH CO LTD +1
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Especially for the general object detection model, since this model usually requires the input of very large pictures, the efficiency of stand-alone training is particularly low. A 100,000-sized data set takes 7 days or even half a month to train
This long time overhead leads to the increasingly low efficiency of model iteration

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Multi-machine cluster-based neural network training method and system
  • Multi-machine cluster-based neural network training method and system
  • Multi-machine cluster-based neural network training method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] In order to make the objects, technical solutions and advantages of the present invention more apparent, exemplary embodiments according to the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some of the embodiments of the present invention, but not all of the embodiments of the present invention, and it should be understood that the present invention is not limited by the example embodiments described herein. Based on the embodiments of the present invention described in the present invention, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present invention.

[0027] As mentioned earlier, the current training of neural networks on a single machine is particularly inefficient. Therefore, the present invention proposes a multi-machine cluster-based neural network training scheme, which can improv...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The present invention provides a neural network training system and method based on multi-machine clusters and computing equipment and servers used in the system and method. For training a plurality of training machines, the network cluster connection is: each of the training machines is connected to any other training machines in the plurality of training machines. According to the embodiment of the present invention, the neural network training system and method based on multi-machine cluster and the computing equipment and server used in the system and method perform neural network training based on multiple training machines, and connect multiple The training machine can speed up the data communication speed between different training machines and improve the speed of neural network training.

Description

technical field [0001] The present invention relates to the technical field of neural network training, and more particularly to a multi-machine cluster-based neural network training system and method, as well as a computing device and a server used by the system and method. Background technique [0002] With the widening and deepening of the neural network model itself, and the exponential explosion of data in various industries, the computational power required to train a usable neural network has also grown exponentially. Especially the general object detection model, because this model usually requires the input of large pictures, which leads to the low efficiency of single-machine training, and a 100,000-sized data set takes 7 days or even half a month to train. This long time overhead leads to an increasingly inefficient model iteration. SUMMARY OF THE INVENTION [0003] The present invention has been made in view of the above-mentioned problems. According to an as...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06N3/08
CPCG06N3/08
Inventor 彭超贾开俞刚
Owner BEIJING KUANGSHI TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap