Monitor Deep Learning Training Progress

Question

Zhangyu Huang 2021-2-8

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/738987-monitor-deep-learning-training-progress

回答： Rohit Pappu 2021-2-10

采纳的回答： Rohit Pappu

In MATLAB R2020b, when I use trainingOptions to set up network, MATLAB cannot create a figure of training progress.

By the way, I specify 'Plots' to 'Training-Progress', but it doesn't show the figure like the below.

I never see this kind of figure during training progress.

So what is the problem? Is it a bug?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Rohit Pappu 2021-2-10

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/738987-monitor-deep-learning-training-progress#answer_619947

在 MATLAB Online 中打开

I tried reproducing the above issue using an example provided in the documentation -

digitDatasetPath = fullfile(matlabroot,'toolbox','nnet','nndemos','nndatasets','DigitDataset');
imds = imageDatastore(digitDatasetPath, 'IncludeSubfolders',true,'LabelSource','foldernames');
numTrainFiles = 750;
[imdsTrain,imdsValidation] = splitEachLabel(imds,numTrainFiles,'randomize');
inputSize = [28 28 1];
numClasses = 10;
layers = [
    imageInputLayer(inputSize)
    convolution2dLayer(5,20)
    batchNormalizationLayer
    reluLayer
    fullyConnectedLayer(numClasses)
    softmaxLayer
    classificationLayer];
options = trainingOptions('sgdm', ...
    'MaxEpochs',4, ...
    'ValidationData',imdsValidation, ...
    'ValidationFrequency',30, ...
    'Verbose',false, ...
    'Plots','training-progress');
net = trainNetwork(imdsTrain,layers,options);