Title:
DEEP LEARNING MODEL TRAINING METHOD AND APPARATUS, DEVICE AND STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2023/226284
Kind Code:
A1
Abstract:
The present invention relates to the technical field of model training, and in particular to a deep learning model training method and apparatus, a device and a storage medium. The method comprises: firstly, selecting a storage mode to store a training dataset into a local node according to the capacity of the training dataset required by the deep learning model to be trained; then completing an operation of storing the training dataset in the local node; and finally using the training dataset in in the local node to train the deep learning model. According to the invention, the training dataset is stored in the local node according to the capacity of the training dataset, thereby saving the time required for data storage, reducing the total time required by training, and improving the training efficiency.
More Like This:
Inventors:
FAN GAOJUN (CN)
ZENG WEI (CN)
WANG HUI (CN)
ZENG WEI (CN)
WANG HUI (CN)
Application Number:
PCT/CN2022/126231
Publication Date:
November 30, 2023
Filing Date:
October 19, 2022
Export Citation:
Assignee:
PENG CHENG LAB (CN)
International Classes:
G06N3/04; G06N3/08
Foreign References:
| CN114676795A | 2022-06-28 | |||
| CN113344074A | 2021-09-03 | |||
| CN112306623A | 2021-02-02 | |||
| CN111860835A | 2020-10-30 | |||
| CN113792885A | 2021-12-14 | |||
| CN113222118A | 2021-08-06 | |||
| CN112000473A | 2020-11-27 |
Attorney, Agent or Firm:
JOHNSON INTELLECTUAL PROPERTY AGENCY (SHENZHEN) (CN)
Download PDF:
Previous Patent: TAB COATING DEVICE
Next Patent: HIGH-PRECISION ACTIVE PRESSURE CONTROL METHOD FOR ALKALINE WATER ELECTROLYZER
Next Patent: HIGH-PRECISION ACTIVE PRESSURE CONTROL METHOD FOR ALKALINE WATER ELECTROLYZER