Method for cutting neural network model and electronic equipment

A technology of neural network model and electronic equipment, applied in the field of tailoring neural network model, can solve problems such as inability to support neural network model deployment, limited edge device resources, etc.

Active Publication Date: 2019-11-19
LENOVO (BEIJING) CO LTD
View PDF7 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In view of this, this application provides a method for tailoring the neural network model to solve the problem that the limited resources of the edge device in the prior art cannot support the deployment of the neural network model

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
  • Method for cutting neural network model and electronic equipment
  • Method for cutting neural network model and electronic equipment
  • Method for cutting neural network model and electronic equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the application with reference to the drawings in the embodiments of the application. Apparently, the described embodiments are only some of the embodiments of the application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.

[0054] Such as figure 1 As shown, it is a flow chart of Embodiment 1 of a method for clipping a neural network model provided by the present application. The method is applied to an electronic device, and the method includes the following steps:

[0055] Step S101: Select a neuron layer to be clipped, and the neuron layer to be clipped includes at least two neurons;

[0056] Wherein, the neural network model includes multiple neuron layers, and each neuron layer may include multipl...

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 invention provides a method for cutting a neural network model. The method includes: based on the lowest importance of neurons in a to-be-clipped neuron layer, determining that the first neuron isa to-be-clipped neuron, and determining the second neuron as a to-be-clipped neuron based on the minimum difference degree with the first neuron, and so on, determining the neuron with the minimum difference degree with the previous to-be-clipped neuron as a new to-be-clipped neuron until the number of clipped neurons meets the preset number of to-be-clipped neurons. The cut neurons have the lowest importance degree; the accuracy of the neural network model is not greatly influenced by the neural network model with the lowest importance degree and the neuron with the lowest difference degreefrom the neuron with the lowest importance degree, the compression cutting of the neural network model is completed on the premise of ensuring the data processing accuracy of the neural network model,and the neural network model is conveniently deployed on edge equipment.

Description

technical field [0001] The present application relates to the field of electronic equipment, and more specifically, relates to a method for clipping a neural network model and electronic equipment. Background technique [0002] "Edge intelligence", as the last mile of artificial intelligence to implement in real life, needs to run complex deep neural network models on terminal devices. [0003] With the increasing demand for machine intelligence in the field of artificial intelligence, the design of the structure of the neural network becomes more and more complex, and the amount of calculation and storage space required also increases greatly. However, portable devices at the edge (such as FaceID ( face authentication), UAVs, decentralized autonomous driving systems, etc.) have limited computing and storage resources, which makes it difficult for deep neural networks to be efficiently deployed on edge devices. Therefore, the neural network model needs to be compressed and ...

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 Applications(China)
IPC IPC(8): G06N3/063
CPCG06N3/063
Inventor 舒红乔王奇刚李远辉杨安荣邓建林
Owner LENOVO (BEIJING) 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