Java Code Examples for org.deeplearning4j.nn.weights.WeightInit#DISTRIBUTION

public static PretrainedNNFilter getPretrainedNNFilterByName(final String name, final int layer,
                                                             final long[] shape) {
    // Thanks to a pointless API requirement, the zoo models require int[] shapes
    // while dl4j uses long[] at any other place
    final int[] intShape = Arrays.stream(shape).mapToInt(i -> (int) i).toArray();

    switch (name) {
        case "AlexNet":
            return new PretrainedNNFilter(new AlexNet(42, intShape, 10, new Nesterovs(1e-2, 0.9), CacheMode.NONE,
                    WorkspaceMode.ENABLED, AlgoMode.PREFER_FASTEST), layer, shape, name);
        case "LeNet":
            return new PretrainedNNFilter(new LeNet(42, intShape, 10, new Nesterovs(1e-2, 0.9), CacheMode.NONE,
                    WorkspaceMode.ENABLED, AlgoMode.PREFER_FASTEST), layer, shape, name);
        case "VGG19":
            return new PretrainedNNFilter(new VGG19(42, intShape, 10, new Nesterovs(1e-2, 0.9), CacheMode.NONE,
                    WorkspaceMode.ENABLED, AlgoMode.PREFER_FASTEST), layer, shape, name);
        case "ResNet50":
            return new PretrainedNNFilter(new ResNet50(42, intShape, 10, WeightInit.DISTRIBUTION,
                    new Nesterovs(1e-2, 0.9), CacheMode.NONE, WorkspaceMode.ENABLED, AlgoMode.PREFER_FASTEST), layer,
                    shape, name);
        default:
            return new PretrainedNNFilter(new VGG16(42, intShape, 10, new Nesterovs(1e-2, 0.9), CacheMode.NONE,
                    WorkspaceMode.ENABLED, AlgoMode.PREFER_FASTEST), layer, shape, "VGG16");
    }
}
 

public GraphBuilder nOutReplace(String layerName, int nOut, WeightInit scheme, Distribution dist) {
    if(scheme == WeightInit.DISTRIBUTION) {
        throw new UnsupportedOperationException("Not supported!, Use " +
                "nOutReplace(layerNum, nOut, new WeightInitDistribution(dist), new WeightInitDistribution(distNext)) instead!");
    }
    return nOutReplace(layerName, nOut, scheme.getWeightInitFunction(), new WeightInitDistribution(dist));
}
 

public GraphBuilder nOutReplace(String layerName, int nOut, Distribution dist, WeightInit scheme) {
    if(scheme == WeightInit.DISTRIBUTION) {
        throw new UnsupportedOperationException("Not supported!, Use " +
                "nOutReplace(layerNum, nOut, new WeightInitDistribution(dist), new WeightInitDistribution(distNext)) instead!");
    }
    return nOutReplace(layerName, nOut, new WeightInitDistribution(dist), scheme.getWeightInitFunction());
}
 

public GraphBuilder nOutReplace(String layerName, int nOut, WeightInit scheme, WeightInit schemeNext) {
    if(scheme == WeightInit.DISTRIBUTION || schemeNext == WeightInit.DISTRIBUTION) {
        throw new UnsupportedOperationException("Not supported!, Use " +
                "nOutReplace(layerNum, nOut, new WeightInitDistribution(dist), new WeightInitDistribution(distNext)) instead!");
    }
    return nOutReplace(layerName, nOut, scheme.getWeightInitFunction(), schemeNext.getWeightInitFunction());
}
 

/**
 * Set the weight initialization for the recurrent weights. Not that if this is not set explicitly, the same
 * weight initialization as the layer input weights is also used for the recurrent weights.
 *
 * @param weightInit Weight initialization for the recurrent weights only.
 */
public T weightInitRecurrent(WeightInit weightInit) {
    if (weightInit == WeightInit.DISTRIBUTION) {
        throw new UnsupportedOperationException(
                        "Not supported!, Use weightInit(Distribution distribution) instead!");
    }

    this.setWeightInitFnRecurrent(weightInit.getWeightInitFunction());
    return (T) this;
}
 

/**
 * Modify the architecture of a layer by changing nOut
 * Note this will also affect the layer that follows the layer specified, unless it is the output layer
 * Can specify different weight init schemes for the specified layer and the layer that follows it.
 *
 * @param layerNum   The index of the layer to change nOut of
 * @param nOut       Value of nOut to change to
 * @param scheme     Weight Init scheme to use for params in the layerNum
 * @param schemeNext Weight Init scheme to use for params in the layerNum+1
 * @return Builder
 */
public Builder nOutReplace(int layerNum, int nOut, WeightInit scheme, WeightInit schemeNext) {
    if(scheme == WeightInit.DISTRIBUTION || schemeNext == WeightInit.DISTRIBUTION) {
        throw new UnsupportedOperationException("Not supported!, Use " +
                "nOutReplace(layerNum, nOut, new WeightInitDistribution(dist), new WeightInitDistribution(distNext)) instead!");
    }
    return nOutReplace(layerNum, nOut, scheme.getWeightInitFunction(), schemeNext.getWeightInitFunction());
}
 

/**
 * Modify the architecture of a layer by changing nOut
 * Note this will also affect the layer that follows the layer specified, unless it is the output layer
 * Can specify different weight init schemes for the specified layer and the layer that follows it.
 *
 * @param layerNum The index of the layer to change nOut of
 * @param nOut     Value of nOut to change to
 * @param scheme   Weight init scheme to use for params in layerNum
 * @param distNext Distribution to use for parmas in layerNum+1
 * @return Builder
 * @see org.deeplearning4j.nn.weights.WeightInitDistribution
 */
public Builder nOutReplace(int layerNum, int nOut, WeightInit scheme, Distribution distNext) {
    if(scheme == WeightInit.DISTRIBUTION) {
        throw new UnsupportedOperationException("Not supported!, Use " +
                "nOutReplace(int layerNum, int nOut, Distribution dist, Distribution distNext) instead!");
    }
    return nOutReplace(layerNum, nOut, scheme.getWeightInitFunction(), new WeightInitDistribution(distNext));
}