NEURAL NETWORK MODEL QUANTIFICATION METHOD AND APPARATUS, ELECTRONIC DEVICE AND COMPUTER-READABLE STORAGE MEDIUM

摘要

The present disclosure relates to the technical field of machine learning, and provided are a neural network model quantification method and apparatus, an electronic device, and a computer-readable storage medium. The method comprises: acquiring a neural network model, wherein the neural network model comprises a convolutional layer, a normalization layer and a quantized activation layer, and output features of the quantized activation layer are integer-type features; converting the parameters of the convolutional layer into integer-type parameters; merging the normalization layer and the quantized activation layer to obtain a merged layer; and converting the parameters of the merged layer into integer-type parameters to obtain a quantized neural network model. The present disclosure reduces the computational complexity of the neural network model, so that the quantized neural network model may run faster on target devices that only support fixed-point or low-bit width operations.