Training method for parameter configuration model, parameter configuration method, and parameter configuration device

Abstract

The present disclosure relates to a training method for a parameter configuration model, a parameter configuration method, and a parameter configuration device. The training method includes: obtaining a training sample set of a data processing device, wherein the training sample set includes operation parameters and benefit parameters in one-to-one correspondence with the operation parameters, the operation parameter includes at least one of a global operation parameter or a local operation parameter associated with position distribution, the benefit parameter is configured to reflect a ratio of power consumption to computing power of the data processing device; training the parameter configuration model based on the training sample set, wherein the operation parameter is configured as an input of the parameter configuration model, and the benefit parameter is configured as an output of the parameter configuration model; and in a case that a training accuracy of the parameter configuration model is greater than or equal to a first preset accuracy, ending the training and obtaining a trained parameter configuration model.