Resource scheduling method and resource scheduling device in multi-model exploration

Abstract

The invention provides a resource scheduling method and a resource scheduling device in multi-model exploration. The resource scheduling method in the multi-model exploration comprises: conducting a round of hyper-parameter exploration training on a plurality of machine learning models based on the same target data set, each machine learning model at least exploring M sets of hyper-parameters in the round of exploration, and M being a positive integer larger than 1; based on the model evaluation indexes corresponding to the multiple sets of hyper-parameters explored by the multiple machine learning models in the round, calculating the performance score of each machine learning model in the round, and calculating the future potential score of each machine learning model; determining a resource allocation scheme for allocating available resources to each machine learning model by integrating the current round of presentation score and the future potential score of each machine learning model; and performing corresponding resource scheduling in the next round of hyper-parameter exploration training according to the resource allocation scheme.

Description

technical field [0001] The present invention generally relates to the field of artificial intelligence, and more specifically, relates to a resource scheduling method and a resource scheduling device in multi-model exploration. Background technique [0002] With the emergence of massive data, artificial intelligence technology has developed rapidly, and machine learning is an inevitable product of the development of artificial intelligence to a certain stage. It is committed to mining valuable potential information from large amounts of data through computing means. [0003] At present, AutoML (Auto-Machine learning, automatic machine learning) is a very popular direction in machine learning, which is dedicated to automatically determining the optimal parameters and network structure according to the problem. There are endless ways to implement AutoML, and they can be roughly divided into two categories: one is a single model (ie, a machine learning model) to explore the bes...

AI Technical Summary

Problems solved by technology

[0004] However, the current exploration method is usually that each model performs hyperparameter tuning independently, lacking resource coordination and scheduling between models

And no matter which model or performance will continue to run, this will undoubtedly waste a lot of computing and time resources

Lack of effective management mechanism in resource and exploration efficiency

Images