Model structure adjusting method and device
A technology of model structure and adjustment method, which is applied in neural learning methods, biological neural network models, neural architectures, etc., can solve the problems of being unable to test the model structure on the platform, prolonging the model training time, and simplifying the model training process , the effect of improving the accuracy
Active Publication Date: 2019-11-12
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF13 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, the model structure is deployed on the deep learning platform during training, and the actual time-consuming of the model structure on the terminal cannot be tested on the platform.
If the model structure is deployed on the terminal to obtain the actual time-consuming for each training, a large number of operations will be generated and the model training time will be extended
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0061] Exemplary embodiments of the present application are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present application to facilitate understanding, and they should be regarded as exemplary only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the application. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
[0062] figure 1 The flow chart of the model structure adjustment method provided by the embodiment of this application, such as figure 1 As shown, the model structure adjustment methods of our own strength include:
[0063] Step S11: Test the structure of the model to be tested deployed on the model training platform by using the test data, and obtain the test resu...
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
Login to View More Abstract
The invention discloses a model structure adjusting method and device, and relates to the field of computers. According to the specific implementation scheme, the method comprises the steps of testinga to-be-tested model structure deployed on a model training platform by utilizing test data to obtain a test result; obtaining return information of the to-be-tested model structure according to thetest result and the test time consumption; wherein the test time consumption is the time consumption for testing a to-be-tested model structure deployed on the test terminal by using the test data, and is obtained according to a preset mapping relationship; wherein the preset mapping relationship comprises a corresponding relationship among the to-be-tested model structure, the test data and the test time consumption; and adjusting the to-be-tested model structure according to the return information to obtain an optimized model structure.
Description
technical field [0001] The present application relates to the field of computer technology, in particular to the field of model structure testing. Background technique [0002] MANS is the abbreviation of Neural Architecture Search for Mobile, that is, "neural architecture search on the mobile phone side", which is a part of AutoML, which automatically searches the network structure. AutoML is an AI capable of only sub-AI (Artificial Intelligence, artificial intelligence), consisting of a controller (Controller) and a sub-network (Child). Automated network search enables machines to design corresponding network structures based on different tasks, and then to design corresponding model structures based on different tasks. At present, AutoML has been applied to many important aspects such as image classification, image detection, and image segmentation. When using AutoML to design the model structure, it is necessary to calculate the reward based on the actual time spent on...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/214Y02T10/40
Inventor 姚锟孙逸鹏黎健成韩钧宇刘经拓丁二锐
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD