Deep learning model reasoning method, system and equipment and computer medium

Abstract

The invention discloses a deep learning model reasoning method, system and device, and a computer medium. The method comprises the following steps: determining channel quantization parameters of each channel in a target deep learning model; determining a branch quantization parameter of the model branch based on the channel quantization parameter according to a relationship between a channel in the target deep learning model and the model branch; quantizing the target deep learning model based on the branch quantization parameters to obtain a quantized deep learning model; and reasoning based on the quantized deep learning model to obtain a reasoning result. In the application, the channel quantization parameter of each channel in the target deep learning module is determined, fine-grained quantization parameter determination is realized, the branch quantization parameter of the model branch is determined based on the channel quantization parameter, reasoning is performed based on the quantized deep learning model, and the reasoning result is obtained. The deep learning model reasoning based on the fine-grained channel quantization parameters is realized, and the reasoning precision of the deep learning model is improved.

Description

technical field [0001] The present application relates to the technical field of deep learning, and more specifically, relates to a deep learning model reasoning method, system, device and computer medium. Background technique [0002] Deep learning (that is, deep neural network) is a branch of machine learning. It is an algorithm that uses artificial neural networks as a framework to realize pattern recognition (that is, reasoning) by learning (that is, training) information features. Deep learning is characterized by the use of unsupervised or semi-supervised feature learning and hierarchical feature extraction to replace manual feature extraction. The engineering implementation of deep learning training and inference process generally depends on the deep learning framework. The deep learning framework mainly focuses on the rich deep learning model structure and operator expression, but has no idea how to efficiently implement inference tasks on different hardware backends...

AI Technical Summary

Problems solved by technology

However, the existing quantization methods are all hierarchical quantization, which makes the inference accuracy of the deep learning model poor

Images