A simulation evaluation method and device for running an algorithm model on a mobile terminal

Abstract

Embodiments of the present application provide a simulation evaluation method and device for running an algorithm model on a mobile terminal. In the embodiment of the present application, in the model evaluation oriented to the mobile terminal, for some device evaluation environments, a cloud evaluation environment adapted to the cloud side resources is simulated, which can replace some real device devices for model evaluation, and solve the problem of The problem of model evaluation for all real machine equipment; for another part of the equipment evaluation environment, the corresponding real machine equipment is used for real machine evaluation, and the machine learning model is evaluated based on the real machine environment and the cloud evaluation environment. The problem of model evaluation for real machine equipment provides conditions for the implementation of end-to-end intelligence. It can also take into account the evaluation advantages of both the model evaluation based on the real machine environment and the model evaluation based on the cloud evaluation environment. The evaluation efficiency of the machine learning model, the correctness and authenticity of the evaluation results, and the automatic evaluation for the mobile terminal.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, and in particular, to a simulation evaluation method and device for running an algorithm model on a mobile terminal. Background technique [0002] Machine learning is the study of computer algorithms, which hopes to automatically optimize the performance of computer algorithms (or models) through data or previous experience. Usually, machine learning models are obtained by machine learning engineers completing data collection and algorithm design for the problem, and then using traditional cloud-side machine learning frameworks such as TensorFlow and PyTorch for model training in the cloud. In the model inference process, the mobile device initiates a model request to the cloud, and the cloud uses the machine learning model to perform related computing tasks and returns the task results to the mobile device for further processing by the mobile device. This way of using ...

AI Technical Summary

Problems solved by technology

However, model evaluation for mobile devices is a pain point for end-side intelligence

Images