Check patentability & draft patents in minutes with Patsnap Eureka AI!

A tensorflow-based training model storage method, driver, and computing server

A computing server and training model technology, applied in computing, instrumentation, program control design, etc., can solve problems such as failure to save training models

Active Publication Date: 2022-04-05
HUAWEI TECH CO LTD
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Therefore, this application provides a Tensorflow-based training model storage method, driver, and computing server, the purpose of which is to enable the computing server and parameter server that save the training model to run on the same computing device, so as to solve the problem of failure to save the training 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
  • A tensorflow-based training model storage method, driver, and computing server
  • A tensorflow-based training model storage method, driver, and computing server
  • A tensorflow-based training model storage method, driver, and computing server

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] The training model preservation method described in this application can be applied in figure 2 In the Tensorflow On Spark architecture shown. and figure 1 compared to, figure 2 The shown Tensorflow On Spark architecture adds a storage system, and improves SparkDriver, computing server, and parameter server, so that the computing server and parameter server that save the model run on the same computing device.

[0020] figure 2 The computing device shown is a device running an operating system, and may include a physical machine, a virtual machine, or a Docker container.

[0021] image 3 The Tensorflow-based training model preservation method disclosed in the embodiment of the present application includes the following steps:

[0022] S301: After the Spark Driver schedules the training tasks to multiple computing servers, it obtains an Internet Protocol (Internet Protocol, IP) address of the parameter server, and stores the IP address of the parameter server in...

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

This application provides a method for saving a training model based on Tensorflow: the driver obtains and stores the first identifier and the second identifier, the first identifier is the identifier of the computing device running the parameter server, and the second identifier is the identifier running on the same machine as the parameter server A compute server identifier for a compute device. The computing server obtains the first identification and the second identification from the driver, and stores the training model when it is confirmed that the identification of the computing device running on the computing server is the same as the first identification, and the identification of the computing server is the same as the second identification, thereby improving training The success rate of model saving.

Description

technical field [0001] The present application relates to the field of electronic information, in particular to a Tensorflow-based training model storage method, driver, and computing server. Background technique [0002] Tensorflow Tensorflow is a machine learning model produced by Google, which provides distributed machine learning and deep learning capabilities. [0003] figure 1 It is a structural schematic diagram of Tensorflow On Spark, a common architecture of Tensorflow. Tensorflow On Spark includes the following logical units: Spark driver Driver, computing server and parameter server. Among them, Spark Driver dispatches training tasks to multiple computing servers and distributes training data to each computing server. The calculation server executes the training process according to the training tasks and training data, and obtains the feedback value of the model parameters. The parameter server corrects the model parameters (for example, the model parameters in...

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): G06F9/48G06F9/50
CPCG06F9/485G06F9/5027
Inventor 袁建勇余远铭王超
Owner HUAWEI TECH CO LTD
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com