error using nnet.internal.cnn.layer.util.fullyConnectedGPUImageStrategy
13 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
disp('Track: network training section begins!')
trainOpt = trainingOptions('adam', ...
'InitialLearnRate',options.learningRate, ...
'LearnRateSchedule','piecewise', ...
'LearnRateDropFactor',options.dropFactor, ...
'LearnRateDropPeriod',options.learnRateSch, ...
'L2Regularization',options.weightDecay, ...
'MaxEpochs',options.maxNumEpochs, ...
'MiniBatchSize',options.batchSize, ...
'Shuffle','every-epoch', ...
'ValidationData',dataPrep.val, ...
'ValidationFrequency',options.valFreq, ...
'ExecutionEnvironment','gpu', ...
'ValidationPatience', 10, ... % Disables automatic training break-off
'Plots','none');
gpuDevice(1)
[trainedNet, trainingInfo] = trainNetwork(dataPrep.train{1,:}, dataPrep.train{2,:}, net, trainOpt);
save([options.netSaveDir,'net_ant=',num2str(options.ch(i)),'_pilot=',num2str(options.pilotSize)], 'trainedNet')
% show NMSE on validation dataset
nanLoc = isnan(trainingInfo.ValidationLoss);
valNMSE = trainingInfo.ValidationLoss(~nanLoc);
options.valNMSE = valNMSE;
R_NMSE(j, i) = options.valNMSE(end);
0 commentaires
Réponses (0)
Voir également
Catégories
En savoir plus sur Deep Learning Toolbox dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!