Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A deep learning model compression method, device, storage medium and terminal equipment

A technology of deep learning and compression methods, applied in neural learning methods, biological neural network models, neural architectures, etc. Effect

Active Publication Date: 2020-12-11
广东宜通联云智能信息有限公司
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The present invention provides a deep learning model compression method, device, storage medium and terminal equipment, aiming to solve the problem that the current lack of methods for compressing deep learning models makes it difficult for deep learning models to run on resource-limited devices

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
  • A deep learning model compression method, device, storage medium and terminal equipment
  • A deep learning model compression method, device, storage medium and terminal equipment
  • A deep learning model compression method, device, storage medium and terminal equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] Such as figure 1 Shown, a kind of deep learning model compression method described in the present invention comprises:

[0032]Model initialization step: initialize the deep learning model, and select an activation function according to the initial value of the connection weight of each node in the model determined by the deep learning model;

[0033] Forward propagation step: According to the connection weight and activation function, the processing value and activation value of the input value in each layer of the model are obtained, and the activation value of the final output layer is the calculated output value of this model;

[0034] Node deletion step: calculate the mean value of the connection weight of each node, delete the node with a mean value of zero, and obtain a new deep learning model;

[0035] Error calculation step: use the loss function to calculate the output error of the new deep learning model. If the error is less than the preset threshold, the m...

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 discloses a deep learning model compression method and device, a storage medium and terminal equipment. the output of each layer of the deep learning model is calculated through a forward propagation algorithm, the output value of the deep learning model is finally obtained, and redundant parameters in the model are sparse by calculating the connection weight mean value of each nodeand deleting the nodes with the mean value of zero, so that the redundant parameters are removed. And the model precision is ensured through the loss function, and on the premise of ensuring that themodel precision is not obviously reduced, the model parameter quantity is greatly reduced, and the calculation amount in the training process is greatly reduced, so that the model can run on resource-limited edge computing equipment. According to the deep learning model compression method, the problem that a deep learning model is difficult to operate on resource-limited equipment due to the lackof a method for compressing the deep learning model at present is solved.

Description

technical field [0001] The invention relates to the field of edge computing in a cloud computing environment, in particular to a deep learning model compression method, device, storage medium and terminal equipment. Background technique [0002] Deep learning solves many challenging problems, and its results have been widely used in computer vision, speech recognition, natural language processing and other fields. Technologies such as image recognition, video processing, and speech recognition based on deep learning have great application prospects and demands on end devices of edge computing systems. However, training and executing deep learning models usually requires a lot of data storage space and super computing power. Existing edge computing devices have insufficient resources such as computing power, storage capacity, network bandwidth, and electricity to run deep learning models. The parameters of the deep learning model are huge and require a lot of computing powe...

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): G06N3/04G06N3/08
Inventor 王永斌张忠平季文翀刘廉如丁雷陈益强彭晓晖李啸海
Owner 广东宜通联云智能信息有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products