An optimization method and apparatus for a deep learning network model server

Abstract

The present application discloses an optimization method and apparatus for a deep learning network model server. The optimization method of the deep learning network model server includes: determiningan image tag of a pre-stored image training set on the server; inputting the image training set with an image tag to a server, and training a neural network initialized based on a pre-training modelto obtain a target neural network; identifying the class attribute of the image to be measured by the server through the target neural network, wherein the residual unit of the neural network adopts attention mechanism when learning and extracting features on the server. The present application solves the technical problem of slow server operation caused by slow convergence speed during network training in deep learning.

Description

technical field [0001] The present application relates to the technical field of deep learning servers, in particular, to an optimization method and device for a deep learning network model server. Background technique [0002] Deep learning is a new field in machine learning research. Its motivation is to establish and simulate the neural network of human brain for analysis and learning. It imitates the mechanism of human brain to explain data, such as images, sounds and texts. Deep machine learning methods can also be divided into supervised learning and unsupervised learning. The learning models established under different learning frameworks are very different. For example, convolutional neural networks (CNNs for short) are a kind of deep supervised learning. Deep Belief Nets (DBNs) is a machine learning model under unsupervised learning. [0003] The current deep learning network model server mainly has the following defects: slow convergence speed, high computational ...

AI Technical Summary

Problems solved by technology

[0003] The current deep learning network model server mainly has the following defects: slow convergence speed, high computational complexity resulting in slow calculation speed

[0004] For the problem of slow running speed caused by slow convergence speed of network training in the deep learning network model server in related technologies, no effective solution has been proposed so far

Images