Main Content


Parameters to configure deep learning code generation that does not depend on third-party libraries

Since R2021a


The coder.DeepLearningCodeConfig object contains the parameters that the codegen function uses to generate generic C or C++ code for deep neural networks.


Create an DeepLearningCodeConfig configuration object by using the coder.DeepLearningConfig function with target library set to 'none'.


expand all

Compression type, specified as "none" or "bfloat16". To enable learnables compression, use "bfloat16".

Name of target library, specified as a character vector.


collapse all

Create an entry-point function resnet50 that uses the coder.loadDeepLearningNetwork function to load the resnet50 (Deep Learning Toolbox) object.

function out = resnet_predict(in)

persistent mynet;
if isempty(mynet)
    mynet = coder.loadDeepLearningNetwork('resnet50','myresnet');

out = predict(mynet,in);

Create a coder.config configuration object for MEX code generation.

cfg = coder.config('mex');

Set the target language to C++.

cfg.TargetLang = 'C++';

Create a coder.DeepLearningCodeConfig deep learning configuration object. Assign it to the DeepLearningConfig property of the cfg configuration object.

dlcfg = coder.DeepLearningConfig(TargetLibrary = 'none');
cfg.DeepLearningConfig = dlcfg;

Use the -config option of the codegen function to specify the cfg configuration object. The codegen function must determine the size, class, and complexity of MATLAB® function inputs. Use the -args option to specify the size of the input to the entry-point function.

codegen -args {ones(224,224,3,'single')} -config cfg resnet_predict

The codegen command places the generated files in the codegen folder. This folder contains the C++ code for the entry-point function resnet_predict.cpp, the header, and the source files that contain the C++ class definitions for the convolutional neural network (CNN), weight, and bias files.

Version History

Introduced in R2021a

expand all