Hyper-parameter tuning method and device and storage medium

Abstract

The invention discloses a hyper-parameter adjustment and optimization method and device and a storage medium, and the method comprises the steps: carrying out the adjustment and optimization of a hyper-parameter adjustment and optimization strategy when the current available calculation resource is obtained because the corresponding relation between the resource quantity of the calculation resource and the hyper-parameter adjustment and optimization strategy is stored in advance; selecting the hyper-parameter tuning strategy corresponding to the currently available computing resources according to the corresponding relationship, and then turning the hyper-parameter according to the hyper-parameter tuning strategy. By applying the technical scheme, different hyper-parameter tuning strategies are selected according to available computing resources in a specific scene, the existing computing resources can be utilized to the maximum extent, the optimal tuning effect is achieved, and the requirement of training precision is met as far as possible.

Description

technical field [0001] The present application relates to the technical field of deep learning, in particular to a hyperparameter tuning method, device and storage medium. Background technique [0002] Currently, deep learning is being applied in various fields such as big data analysis, computer vision, and natural language processing. For complex problems in reality, deep learning models are often complex in structure and require a huge amount of data to train the network, and both training and reasoning require a large amount of computing resources to run quickly. [0003] The hyperparameter tuning of the deep learning model is to replace the expert experience with computing resources, set several hyperparameters to establish the deep learning model, train the deep learning model, and set the set of hyperparameters with the best effect of the deep learning model as the final hyperparameters. In the above process, each set of hyperparameters needs to correspond to a deep...

AI Technical Summary

Problems solved by technology

In the above process, each set of hyperparameters needs to correspond to a deep learning model, and the training results are compared, so the requirements for computing resources are put forward. If the amount of calculation is greatly reduced, the accuracy of the deep learning model will be reduced. Therefore, the training accuracy and Computational resources create conflicts

Images