A method and a system for constructing and optimizing a depth learning environment

Abstract

The invention discloses a method for constructing and optimizing a depth learning environment, which comprises the following steps: 1) acquiring current software information and current hardware information of an artificial intelligence computing platform; 2) installing or updating corresponding software according to the acquired current software information and current hardware information to complete the construction of the software and hardware environment; 3) installing a depth learning framework under the built software and hardware environment; 4) adjust hardware parameters to optimize hardware performance for that depth learn framework; And 5) optimizing the performance of the depth learning framework according to the software configuration of the software and hardware environment and the configuration of the depth learning framework. The invention can realize the automatic deployment of the depth learning software environment and the automatic optimization of the software and hardware parameters, greatly saves the human resource cost and the time cost, has strong practicability, wide use range and is easy to popularize.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, and more specifically, to a method and system for building and optimizing a deep learning environment. Background technique [0002] With the continuous development of deep learning algorithms and big data, AI (Artificial Intelligence) has gradually penetrated into all walks of life, demonstrating its powerful advantages. AI relies on big data, large models, and deep learning algorithms. In addition, the implementation of AI algorithms and the training and reasoning of models also depend on the hardware of the computing platform, such as CPU, memory, GPU and other hardware. AI model training is often a time-consuming and labor-intensive process. Choosing a high-performance computing platform for model training can improve efficiency, get twice the result with half the effort, and shorten the algorithm development cycle and development time. In addition, the rapid construc...

AI Technical Summary

Problems solved by technology

[0006] In view of this, the purpose of the embodiments of the present invention is to propose a method and system for building and optimizing a deep learning environment, which can solve the problem of human resources caused by manually building a deep learning environment and optimizing hardware and software in the prior art. The problem of huge waste of cost and time cost

Images