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

Searching method of machine learning model and related device, and equipment

A machine learning model and search method technology, applied in the terminal field, can solve the problem of low model quantification efficiency, and achieve the effect of improving efficiency and ensuring accuracy

Active Publication Date: 2020-05-19
HUAWEI TECH CO LTD
View PDF6 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Embodiments of the present invention provide a machine learning model search method and related devices and equipment to solve the problem of low efficiency in the process of model quantification and search

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
  • Searching method of machine learning model and related device, and equipment
  • Searching method of machine learning model and related device, and equipment
  • Searching method of machine learning model and related device, and equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0077] Firstly, the technical terms and concepts involved in this application are introduced.

[0078] (1) neural network

[0079] A neural network can be composed of neural units, which can be referred to as x s and intercept 1 are the input operation unit, the output of the operation unit can be:

[0080]

[0081] Wherein, s=1, 2, ... n, n is a natural number greater than 1, W s for x s The weight of , b is the bias of the neuron unit. f is the activation function of the neural unit, which is used to introduce nonlinear characteristics into the neural network to convert the input signal in the neural unit into an output signal. The output signal of this activation function can be used as the input of the next convolutional layer. The activation function may be a sigmoid function. A neural network is a network formed by connecting many of the above-mentioned single neural units, that is, the output of one neural unit can be the input of another neural unit. The input ...

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 embodiment of the invention discloses a searching method of a machine learning model and a related device, and equipment. Specifically, the invention relates to the technical field of artificial intelligence, the method comprises the following steps of: searching and quantifying a model; generating a plurality of pure bit models according to the to-be-quantized model, further obtaining an evaluation parameter of each layer structure in the plurality of pure bit models; further, selecting one candidate model from the candidate set for training and testing; obtaining a target model, the quantitative weight of each layer of structure in a target model can be determined based on the network structure of the target model and the evaluation parameter of each layer of structure in the targetmodel; therefore, the layer structure with the maximum quantization weight in the target model is quantified, the quantified model is added into the candidate set, frequent information interaction with the terminal can be reduced, and the efficiency of model search and model quantization is improved.

Description

technical field [0001] The present invention relates to the field of terminal technology, in particular to a search method for a machine learning model and related devices and equipment. Background technique [0002] With the rapid development of deep learning, deep neural networks are widely used in image, voice, text classification, recognition and other fields. The network structure of the deep neural network is usually complex, the reasoning time is long, and the operation requires a large amount of memory. Due to the limitations of processor computing power and memory storage resources of mobile terminals, deep neural networks cannot be applied to mobile terminals. Therefore, how to deploy the deep learning model with an increasing amount of calculations on mobile terminals is an urgent problem to be solved. [0003] Hybrid bit quantization for deep neural networks is an efficient solution to this problem. Mixed bit quantization means that the model parameters origin...

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
IPC IPC(8): G06N20/00G06N5/04
CPCG06N20/00G06N5/04Y02D10/00G06N3/09G06N3/0495G06N3/0464G06N3/04G06N3/063G06N3/08
Inventor 俞清华刘默翰隋志成周力白立勋
Owner HUAWEI TECH CO LTD
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com