gistfile1.txt

import org.deeplearning4j.nn.api.Layer;
import org.deeplearning4j.nn.conf.WorkspaceMode;
import org.deeplearning4j.nn.conf.dropout.IDropout;
import org.deeplearning4j.nn.conf.inputs.InputType;
import org.deeplearning4j.nn.conf.layers.ConvolutionLayer;
import org.deeplearning4j.nn.conf.layers.DenseLayer;
import org.deeplearning4j.nn.conf.layers.OutputLayer;
import org.deeplearning4j.nn.graph.ComputationGraph;
import org.deeplearning4j.nn.transferlearning.FineTuneConfiguration;
import org.deeplearning4j.nn.transferlearning.TransferLearning;
import org.deeplearning4j.nn.weights.WeightInit;
import org.deeplearning4j.zoo.ZooModel;
import org.deeplearning4j.zoo.model.ResNet50;
import org.nd4j.linalg.activations.Activation;
import org.nd4j.linalg.api.ndarray.INDArray;
import org.nd4j.linalg.dataset.DataSet;
import org.nd4j.linalg.factory.Nd4j;
import org.nd4j.linalg.learning.config.Nesterovs;
import org.nd4j.linalg.lossfunctions.LossFunctions;

public class Debug6756 {

    public static void main(String[] args) throws Exception {

        long seed = 12345;
        int height = 112;
        int width = 112;
        int channels = 3;
        int numLabels = 1000;


        ZooModel zooModel = ResNet50.builder().workspaceMode(WorkspaceMode.NONE).build();
        ComputationGraph resnet50 = (ComputationGraph) zooModel.initPretrained();
        System.out.println(resnet50.summary());

        FineTuneConfiguration fineTuneConf = new FineTuneConfiguration.Builder()
                .activation(Activation.LEAKYRELU)
                .weightInit(WeightInit.RELU)
                .updater(new Nesterovs(5e-5))
                .dropOut(0.5)
                .seed(seed)
                .build();

        ComputationGraph graph = new TransferLearning.GraphBuilder(resnet50)
                .fineTuneConfiguration(fineTuneConf)
                .setInputTypes(InputType.convolutional(height, width, channels))
                .removeVertexKeepConnections("conv1")
                .addLayer("conv1", new ConvolutionLayer.Builder(new int[]{3, 3})
                        .nIn(channels).nOut(64).activation( Activation.RELU).build(), "input_1")
                .addLayer("fc2048",new DenseLayer.Builder().activation(Activation.TANH).nIn(1000).nOut(2048).build(),"fc1000")
                .removeVertexAndConnections("output")
                .addLayer("newOutput",new OutputLayer
                        .Builder(LossFunctions.LossFunction.NEGATIVELOGLIKELIHOOD)
                        .activation(Activation.SOFTMAX)
                        .nIn(2048)
                        .nOut(numLabels)
                        .build(),"fc2048")
                .setOutputs("newOutput")
                .build();
        System.out.println(graph.summary());

        /////////////////////////////////////////////////////
        //WORKAROUND FOR 1.0.0-beta3
        for(Layer l : graph.getLayers()){
            org.deeplearning4j.nn.conf.layers.Layer conf = (org.deeplearning4j.nn.conf.layers.Layer)l.getConfig();
            if(!(l.getConfig() instanceof org.deeplearning4j.nn.conf.layers.BaseLayer)){
                conf.setIDropout(null);
            } else {
                IDropout d = conf.getIDropout();
                conf.setIDropout(d == null ? null : d.clone());
            }
        }
        /////////////////////////////////////////////////////

        INDArray in = Nd4j.rand(new int[]{1, channels, height, width});
        INDArray labels = Nd4j.create(1, numLabels);
        labels.putScalar(0, 0, 1.0);

        DataSet ds = new DataSet(in, labels);

        graph.fit(ds);

    }
}

Hi Alex: Has this bug in 1.0.0-beta3 been fixed or should I switch to 1.0.0-beta? Best wishes, Hamaad.

P.S., Yes to come confirm that it works without the above bug fix requirement in 1.0.0-beta. Many thanks, Hamaad.