When training a network with "trainNetwork", how to keep the trainable parameters from the begining to the end?

Question

shuai ma 2021-12-26

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1617665-when-training-a-network-with-trainnetwork-how-to-keep-the-trainable-parameters-from-the-begining

回答： yanqi liu 2021-12-27

I want to see the 'sgdm' or 'adam' trace on the loss surface, the network at each iteration is needed? how can I keep the network at each iteration when training?

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

yanqi liu 2021-12-27

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1617665-when-training-a-network-with-trainnetwork-how-to-keep-the-trainable-parameters-from-the-begining#answer_862400

在 MATLAB Online 中打开

yes，sir，may be set options to train，such as

options = trainingOptions('sgdm', ...
    'MaxEpochs',20,...
    'InitialLearnRate',1e-4, ...
    'Verbose',false, ...
    'Plots','training-progress');
[tnet,tinfo] = trainNetwork(imdsTrain,layers,options);