Unlock instant, AI-driven research and patent intelligence for your innovation.

Neural network model conversion method and device, equipment and storage medium

A neural network model and conversion method technology, applied in the field of neural network model conversion methods, devices, equipment and storage media, can solve problems such as low development efficiency and inconsistent conversion methods, and achieve cost reduction and model conversion efficiency. Effect

Pending Publication Date: 2021-11-30
BEIJING SENSETIME TECH DEV CO LTD
View PDF8 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In related technologies, due to the inconsistency of conversion methods corresponding to different chip platforms, researchers need to learn conversion tools for multiple AI chips, resulting in low development efficiency

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
  • Neural network model conversion method and device, equipment and storage medium
  • Neural network model conversion method and device, equipment and storage medium
  • Neural network model conversion method and device, equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0081] The technical solutions of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Apparently, the described implementations are part of the implementations of the present disclosure, but not all of them. Based on the implementation manners in the present disclosure, all other implementation manners obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present disclosure. In addition, the technical features involved in different embodiments of the present disclosure described below may be combined with each other as long as they do not constitute a conflict with each other.

[0082] The neural network model refers to the artificial neural network model built to simulate the structure and mechanism of biological brain neurons, and the chip platform refers to the hardware platform on which the neural network model is deployed. With the development...

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 relates to the technical field of artificial intelligence, in particular to a neural network model conversion method and device, equipment and a storage medium. The neural network model conversion method comprises the following steps: acquiring input data which comprises general conversion parameters and platform identification information of a target conversion model; determining a target conversion tool in a pre-established tool set according to the platform identification information of the input data, wherein the tool set comprises a plurality of conversion tools, and each conversion tool is used for performing conversion processing on the neural network model to obtain a corresponding conversion model; and based on the target conversion tool and the general conversion parameters of the input data, performing conversion processing on the neural network model to obtain a target conversion model. According to the conversion method provided by the embodiment of the invention, the model conversion cost is reduced, and the model conversion efficiency is improved.

Description

technical field [0001] The present disclosure relates to the technical field of artificial intelligence, and in particular to a neural network model conversion method, device, equipment and storage medium. Background technique [0002] At present, the artificial intelligence (AI) chips of major chip manufacturers support different neural network model formats, and the corresponding conversion tools and usage methods are also quite different. When transplanting the same neural network to a different chip platform, researchers need to perform model conversion and model testing on the neural network model according to the platform conversion tools and corresponding methods, so as to complete the use of the neural network on the AI ​​chip. [0003] In related technologies, due to the inconsistency of conversion methods corresponding to different chip platforms, researchers need to learn a variety of AI chip conversion tools, resulting in low development efficiency. Contents of...

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): G06N3/063G06F8/76G06F11/36
CPCG06N3/063G06F11/3696G06F8/76
Inventor 吕守禄李通陈薇兆孙延刘文韬钱晨
Owner BEIJING SENSETIME TECH DEV CO LTD