Examples with LossFunction org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction used on opensource projects

Example 6 with LossFunction

use of org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction in project deeplearning4j by deeplearning4j.

the class GradientCheckTests method testGradientGravesBidirectionalLSTMFull.

@Test
public void testGradientGravesBidirectionalLSTMFull() {
    Activation[] activFns = { Activation.TANH, Activation.SOFTSIGN };
    LossFunction[] lossFunctions = { LossFunction.MCXENT, LossFunction.MSE };
    //i.e., lossFunctions[i] used with outputActivations[i] here
    Activation[] outputActivations = { Activation.SOFTMAX, Activation.TANH };
    int timeSeriesLength = 4;
    int nIn = 2;
    int layerSize = 2;
    int nOut = 2;
    int miniBatchSize = 3;
    Random r = new Random(12345L);
    INDArray input = Nd4j.zeros(miniBatchSize, nIn, timeSeriesLength);
    for (int i = 0; i < miniBatchSize; i++) {
        for (int j = 0; j < nIn; j++) {
            for (int k = 0; k < timeSeriesLength; k++) {
                input.putScalar(new int[] { i, j, k }, r.nextDouble() - 0.5);
            }
        }
    }
    INDArray labels = Nd4j.zeros(miniBatchSize, nOut, timeSeriesLength);
    for (int i = 0; i < miniBatchSize; i++) {
        for (int j = 0; j < timeSeriesLength; j++) {
            int idx = r.nextInt(nOut);
            labels.putScalar(new int[] { i, idx, j }, 1.0f);
        }
    }
    //use l2vals[i] with l1vals[i]
    double[] l2vals = { 0.4, 0.0, 0.4, 0.4 };
    double[] l1vals = { 0.0, 0.0, 0.5, 0.0 };
    double[] biasL2 = { 0.0, 0.0, 0.0, 0.2 };
    double[] biasL1 = { 0.0, 0.0, 0.6, 0.0 };
    for (Activation afn : activFns) {
        for (int i = 0; i < lossFunctions.length; i++) {
            for (int k = 0; k < l2vals.length; k++) {
                LossFunction lf = lossFunctions[i];
                Activation outputActivation = outputActivations[i];
                double l2 = l2vals[k];
                double l1 = l1vals[k];
                NeuralNetConfiguration.Builder conf = new NeuralNetConfiguration.Builder().regularization(l1 > 0.0 || l2 > 0.0);
                if (l1 > 0.0)
                    conf.l1(l1);
                if (l2 > 0.0)
                    conf.l2(l2);
                if (biasL2[k] > 0)
                    conf.l2Bias(biasL2[k]);
                if (biasL1[k] > 0)
                    conf.l1Bias(biasL1[k]);
                MultiLayerConfiguration mlc = conf.seed(12345L).list().layer(0, new GravesBidirectionalLSTM.Builder().nIn(nIn).nOut(layerSize).weightInit(WeightInit.DISTRIBUTION).dist(new NormalDistribution(0, 1)).activation(afn).updater(Updater.NONE).build()).layer(1, new RnnOutputLayer.Builder(lf).activation(outputActivation).nIn(layerSize).nOut(nOut).weightInit(WeightInit.DISTRIBUTION).dist(new NormalDistribution(0, 1)).updater(Updater.NONE).build()).pretrain(false).backprop(true).build();
                MultiLayerNetwork mln = new MultiLayerNetwork(mlc);
                mln.init();
                if (PRINT_RESULTS) {
                    System.out.println("testGradientGravesBidirectionalLSTMFull() - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", l2=" + l2 + ", l1=" + l1);
                    for (int j = 0; j < mln.getnLayers(); j++) System.out.println("Layer " + j + " # params: " + mln.getLayer(j).numParams());
                }
                boolean gradOK = GradientCheckUtil.checkGradients(mln, DEFAULT_EPS, DEFAULT_MAX_REL_ERROR, DEFAULT_MIN_ABS_ERROR, PRINT_RESULTS, RETURN_ON_FIRST_FAILURE, input, labels);
                String msg = "testGradientGravesLSTMFull() - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", l2=" + l2 + ", l1=" + l1;
                assertTrue(msg, gradOK);
            }
        }
    }
}

Example 7 with LossFunction

use of org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction in project deeplearning4j by deeplearning4j.

the class GradientCheckTests method testGradientMLP2LayerIrisSimple.

@Test
public void testGradientMLP2LayerIrisSimple() {
    //Parameterized test, testing combinations of:
    // (a) activation function
    // (b) Whether to test at random initialization, or after some learning (i.e., 'characteristic mode of operation')
    // (c) Loss function (with specified output activations)
    //activation functions such as relu and hardtanh: may randomly fail due to discontinuities
    String[] activFns = { "sigmoid", "tanh", "softplus" };
    //If true: run some backprop steps first
    boolean[] characteristic = { false, true };
    LossFunction[] lossFunctions = { LossFunction.MCXENT, LossFunction.MSE };
    //i.e., lossFunctions[i] used with outputActivations[i] here
    String[] outputActivations = { "softmax", "tanh" };
    DataNormalization scaler = new NormalizerMinMaxScaler();
    DataSetIterator iter = new IrisDataSetIterator(150, 150);
    scaler.fit(iter);
    iter.setPreProcessor(scaler);
    DataSet ds = iter.next();
    INDArray input = ds.getFeatureMatrix();
    INDArray labels = ds.getLabels();
    for (String afn : activFns) {
        for (boolean doLearningFirst : characteristic) {
            for (int i = 0; i < lossFunctions.length; i++) {
                LossFunction lf = lossFunctions[i];
                String outputActivation = outputActivations[i];
                MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder().regularization(false).optimizationAlgo(OptimizationAlgorithm.CONJUGATE_GRADIENT).learningRate(1.0).seed(12345L).list().layer(0, new DenseLayer.Builder().nIn(4).nOut(3).weightInit(WeightInit.DISTRIBUTION).dist(new NormalDistribution(0, 1)).activation(afn).updater(Updater.SGD).build()).layer(1, new OutputLayer.Builder(lf).activation(outputActivation).nIn(3).nOut(3).weightInit(WeightInit.DISTRIBUTION).dist(new NormalDistribution(0, 1)).updater(Updater.SGD).build()).pretrain(false).backprop(true).build();
                MultiLayerNetwork mln = new MultiLayerNetwork(conf);
                mln.init();
                if (doLearningFirst) {
                    //Run a number of iterations of learning
                    mln.setInput(ds.getFeatures());
                    mln.setLabels(ds.getLabels());
                    mln.computeGradientAndScore();
                    double scoreBefore = mln.score();
                    for (int j = 0; j < 10; j++) mln.fit(ds);
                    mln.computeGradientAndScore();
                    double scoreAfter = mln.score();
                    //Can't test in 'characteristic mode of operation' if not learning
                    String msg = "testGradMLP2LayerIrisSimple() - score did not (sufficiently) decrease during learning - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", doLearningFirst=" + doLearningFirst + " (before=" + scoreBefore + ", scoreAfter=" + scoreAfter + ")";
                    assertTrue(msg, scoreAfter < 0.8 * scoreBefore);
                }
                if (PRINT_RESULTS) {
                    System.out.println("testGradientMLP2LayerIrisSimpleRandom() - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", doLearningFirst=" + doLearningFirst);
                    for (int j = 0; j < mln.getnLayers(); j++) System.out.println("Layer " + j + " # params: " + mln.getLayer(j).numParams());
                }
                boolean gradOK = GradientCheckUtil.checkGradients(mln, DEFAULT_EPS, DEFAULT_MAX_REL_ERROR, DEFAULT_MIN_ABS_ERROR, PRINT_RESULTS, RETURN_ON_FIRST_FAILURE, input, labels);
                String msg = "testGradMLP2LayerIrisSimple() - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", doLearningFirst=" + doLearningFirst;
                assertTrue(msg, gradOK);
            }
        }
    }
}

Example 8 with LossFunction

use of org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction in project deeplearning4j by deeplearning4j.

the class GradientCheckTests method testGradientMLP2LayerIrisL1L2Simple.

@Test
public void testGradientMLP2LayerIrisL1L2Simple() {
    //As above (testGradientMLP2LayerIrisSimple()) but with L2, L1, and both L2/L1 applied
    //Need to run gradient through updater, so that L2 can be applied
    String[] activFns = { "sigmoid", "tanh" };
    //If true: run some backprop steps first
    boolean[] characteristic = { false, true };
    LossFunction[] lossFunctions = { LossFunction.MCXENT, LossFunction.MSE };
    //i.e., lossFunctions[i] used with outputActivations[i] here
    String[] outputActivations = { "softmax", "tanh" };
    DataNormalization scaler = new NormalizerMinMaxScaler();
    DataSetIterator iter = new IrisDataSetIterator(150, 150);
    scaler.fit(iter);
    iter.setPreProcessor(scaler);
    DataSet ds = iter.next();
    INDArray input = ds.getFeatureMatrix();
    INDArray labels = ds.getLabels();
    //use l2vals[i] with l1vals[i]
    double[] l2vals = { 0.4, 0.0, 0.4, 0.4 };
    double[] l1vals = { 0.0, 0.0, 0.5, 0.0 };
    double[] biasL2 = { 0.0, 0.0, 0.0, 0.2 };
    double[] biasL1 = { 0.0, 0.0, 0.6, 0.0 };
    for (String afn : activFns) {
        for (boolean doLearningFirst : characteristic) {
            for (int i = 0; i < lossFunctions.length; i++) {
                for (int k = 0; k < l2vals.length; k++) {
                    LossFunction lf = lossFunctions[i];
                    String outputActivation = outputActivations[i];
                    double l2 = l2vals[k];
                    double l1 = l1vals[k];
                    MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder().regularization(true).l2(l2).l1(l1).l2Bias(biasL2[k]).l1Bias(biasL1[k]).optimizationAlgo(OptimizationAlgorithm.CONJUGATE_GRADIENT).seed(12345L).list().layer(0, new DenseLayer.Builder().nIn(4).nOut(3).weightInit(WeightInit.DISTRIBUTION).dist(new NormalDistribution(0, 1)).updater(Updater.NONE).activation(afn).build()).layer(1, new OutputLayer.Builder(lf).nIn(3).nOut(3).weightInit(WeightInit.DISTRIBUTION).dist(new NormalDistribution(0, 1)).updater(Updater.NONE).activation(outputActivation).build()).pretrain(false).backprop(true).build();
                    MultiLayerNetwork mln = new MultiLayerNetwork(conf);
                    mln.init();
                    if (doLearningFirst) {
                        //Run a number of iterations of learning
                        mln.setInput(ds.getFeatures());
                        mln.setLabels(ds.getLabels());
                        mln.computeGradientAndScore();
                        double scoreBefore = mln.score();
                        for (int j = 0; j < 10; j++) mln.fit(ds);
                        mln.computeGradientAndScore();
                        double scoreAfter = mln.score();
                        //Can't test in 'characteristic mode of operation' if not learning
                        String msg = "testGradMLP2LayerIrisSimple() - score did not (sufficiently) decrease during learning - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", doLearningFirst=" + doLearningFirst + ", l2=" + l2 + ", l1=" + l1 + " (before=" + scoreBefore + ", scoreAfter=" + scoreAfter + ")";
                        assertTrue(msg, scoreAfter < 0.8 * scoreBefore);
                    }
                    if (PRINT_RESULTS) {
                        System.out.println("testGradientMLP2LayerIrisSimpleRandom() - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", doLearningFirst=" + doLearningFirst + ", l2=" + l2 + ", l1=" + l1);
                        for (int j = 0; j < mln.getnLayers(); j++) System.out.println("Layer " + j + " # params: " + mln.getLayer(j).numParams());
                    }
                    boolean gradOK = GradientCheckUtil.checkGradients(mln, DEFAULT_EPS, DEFAULT_MAX_REL_ERROR, DEFAULT_MIN_ABS_ERROR, PRINT_RESULTS, RETURN_ON_FIRST_FAILURE, input, labels);
                    String msg = "testGradMLP2LayerIrisSimple() - activationFn=" + afn + ", lossFn=" + lf + ", outputActivation=" + outputActivation + ", doLearningFirst=" + doLearningFirst + ", l2=" + l2 + ", l1=" + l1;
                    assertTrue(msg, gradOK);
                }
            }
        }
    }
}