Deprecated API

Contents

• Deprecated Interfaces
• Deprecated Classes
• Deprecated Enums
• Deprecated Fields
• Deprecated Methods
• Deprecated Constructors
• Deprecated Enum Constants

Deprecated Interfaces

Interface and Description
org.deeplearning4j.eval.IEvaluation Use IEvaluation

Deprecated Classes

Class and Description
org.deeplearning4j.eval.BaseEvaluation Use BaseEvaluation
org.deeplearning4j.optimize.listeners.ComposableIterationListener Not required - DL4J networks can use multiple listeners simultaneously
org.deeplearning4j.eval.ConfusionMatrix Use ConfusionMatrix
org.deeplearning4j.earlystopping.scorecalc.DataSetLossCalculatorCG Use DataSetLossCalculator instead for both MultiLayerNetwork and ComputationGraph
org.deeplearning4j.eval.Evaluation Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.EvaluationBinary
org.deeplearning4j.eval.EvaluationCalibration Use EvaluationCalibration
org.deeplearning4j.eval.EvaluationUtils Use EvaluationUtils
org.deeplearning4j.nn.conf.distribution.GaussianDistribution Use NormalDistribution which is identical to this implementation
org.deeplearning4j.nn.conf.layers.GravesBidirectionalLSTM use Bidirectional instead. With the Bidirectional layer wrapper you can make any recurrent layer bidirectional, in particular GravesLSTM. Note that this layer adds the output of both directions, which translates into "ADD" mode in Bidirectional. Usage: .layer(new Bidirectional(Bidirectional.Mode.ADD, new GravesLSTM.Builder()....build()))
org.deeplearning4j.nn.conf.layers.GravesLSTM Will be eventually removed. Use LSTM instead, which has similar prediction accuracy, but supports CuDNN for faster network training on CUDA (Nvidia) GPUs
org.deeplearning4j.nn.layers.recurrent.GravesLSTM Will be eventually removed. Use LSTM instead, which has similar prediction accuracy, but supports CuDNN for faster network training on CUDA (Nvidia) GPUs
org.deeplearning4j.eval.curves.Histogram Use Histogram
org.deeplearning4j.optimize.api.IterationListener Use TrainingListener instead
org.deeplearning4j.eval.curves.PrecisionRecallCurve Use ReliabilityDiagram
org.deeplearning4j.eval.meta.Prediction Use Prediction
org.deeplearning4j.eval.RegressionEvaluation Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.curves.ReliabilityDiagram Use ReliabilityDiagram
org.deeplearning4j.eval.ROC Use ROC
org.deeplearning4j.eval.ROC.CountsForThreshold Use ROC.CountsForThreshold
org.deeplearning4j.eval.ROCBinary Use ROCBinary
org.deeplearning4j.eval.curves.RocCurve Use RocCurve
org.deeplearning4j.eval.ROCMultiClass Use ROCMultiClass

Deprecated Enums

Enum and Description
org.deeplearning4j.eval.Evaluation.Metric
org.deeplearning4j.eval.EvaluationAveraging Use EvaluationAveraging
org.deeplearning4j.eval.RegressionEvaluation.Metric Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation.Metric
org.deeplearning4j.nn.conf.layers.SpaceToDepthLayer.DataFormat Use CNN2DFormat instead

Deprecated Fields

Field and Description
org.deeplearning4j.eval.EvaluationBinary.DEFAULT_EDGE_VALUE
org.deeplearning4j.eval.EvaluationBinary.DEFAULT_PRECISION
org.deeplearning4j.eval.ROCBinary.DEFAULT_STATS_PRECISION Use ROCBinary
org.deeplearning4j.eval.ROCMultiClass.DEFAULT_STATS_PRECISION Use ROCMultiClass

Deprecated Methods

Method and Description
org.deeplearning4j.nn.conf.layers.ActivationLayer.Builder.activation(String) Use ActivationLayer.Builder.activation(Activation) or @activation(IActivation)
org.deeplearning4j.gradientcheck.GradientCheckUtil.checkGradients(MultiLayerNetwork, double, double, double, boolean, boolean, INDArray, INDArray)
org.deeplearning4j.gradientcheck.GradientCheckUtil.checkGradients(MultiLayerNetwork, double, double, double, boolean, boolean, INDArray, INDArray, INDArray, INDArray, boolean, int, Set<String>, Integer)
org.deeplearning4j.nn.conf.inputs.InputType.convolutional3D(long, long, long, long) Use InputType.convolutional3D(Convolution3D.DataFormat, long, long, long, long)
org.deeplearning4j.nn.conf.layers.ConvolutionLayer.BaseConvBuilder.cudnnAllowFallback(boolean) Use ConvolutionLayer.BaseConvBuilder.helperAllowFallback(boolean)
org.deeplearning4j.nn.conf.layers.LocalResponseNormalization.Builder.cudnnAllowFallback(boolean) Use LocalResponseNormalization.Builder.helperAllowFallback(boolean)
org.deeplearning4j.nn.conf.layers.SubsamplingLayer.BaseSubsamplingBuilder.cudnnAllowFallback(boolean) Use SubsamplingLayer.BaseSubsamplingBuilder.helperAllowFallback(boolean)
org.deeplearning4j.nn.conf.layers.Subsampling3DLayer.BaseSubsamplingBuilder.cudnnAllowFallback(boolean) Use Subsampling3DLayer.BaseSubsamplingBuilder.helperAllowFallback(boolean)
org.deeplearning4j.nn.conf.layers.BatchNormalization.Builder.cudnnAllowFallback(boolean) Use BatchNormalization.Builder.helperAllowFallback(boolean)
org.deeplearning4j.nn.conf.layers.SpaceToDepthLayer.Builder.dataFormat(SpaceToDepthLayer.DataFormat) Use SpaceToDepthLayer.Builder.dataFormat(CNN2DFormat)
org.deeplearning4j.optimize.solvers.accumulation.EncodingHandler.decodeUpdates(INDArray)
org.deeplearning4j.nn.conf.NeuralNetConfiguration.Builder.dist(Distribution) Use NeuralNetConfiguration.Builder.weightInit(Distribution)
org.deeplearning4j.nn.conf.layers.BaseLayer.Builder.dist(Distribution)
org.deeplearning4j.nn.transferlearning.FineTuneConfiguration.Builder.dist(Distribution)
org.deeplearning4j.nn.graph.ComputationGraph.evaluateROC(DataSetIterator) To be removed - use ComputationGraph.evaluateROC(DataSetIterator, int) to enforce selection of appropriate ROC/threshold configuration
org.deeplearning4j.nn.multilayer.MultiLayerNetwork.evaluateROC(DataSetIterator) To be removed - use MultiLayerNetwork.evaluateROC(DataSetIterator, int) to enforce selection of appropriate ROC/threshold configuration
org.deeplearning4j.nn.graph.ComputationGraph.evaluateROC(MultiDataSetIterator) To be removed - use ComputationGraph.evaluateROC(DataSetIterator, int) to enforce selection of appropriate ROC/threshold configuration
org.deeplearning4j.nn.graph.ComputationGraph.evaluateROCMultiClass(DataSetIterator) To be removed - use ComputationGraph.evaluateROCMultiClass(DataSetIterator, int) to enforce selection of appropriate ROC/threshold configuration
org.deeplearning4j.nn.multilayer.MultiLayerNetwork.evaluateROCMultiClass(DataSetIterator) To be removed - use MultiLayerNetwork.evaluateROCMultiClass(DataSetIterator, int) to enforce selection of appropriate ROC/threshold configuration
org.deeplearning4j.nn.api.Model.fit()
org.deeplearning4j.eval.EvaluationBinary.fromJson(String)
org.deeplearning4j.eval.Evaluation.fromJson(String) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.curves.RocCurve.fromJson(String) Use RocCurve
org.deeplearning4j.eval.curves.Histogram.fromJson(String) Use Histogram
org.deeplearning4j.eval.EvaluationBinary.fromYaml(String)
org.deeplearning4j.eval.Evaluation.fromYaml(String) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.curves.RocCurve.fromYaml(String) Use RocCurve
org.deeplearning4j.eval.curves.Histogram.fromYaml(String) Use Histogram
org.deeplearning4j.nn.conf.inputs.InputType.InputTypeConvolutional.getDepth()
org.deeplearning4j.util.ConvolutionUtils.getOutputSize(INDArray, int[], int[], int[], ConvolutionMode)
org.deeplearning4j.util.ConvolutionUtils.getOutputSize(INDArray, int[], int[], int[], ConvolutionMode, int[]) Use ConvolutionUtils.getOutputSize(INDArray, int[], int[], int[], ConvolutionMode, int[], CNN2DFormat)
org.deeplearning4j.nn.conf.layers.InputTypeUtil.getOutputTypeCnnLayers(InputType, int[], int[], int[], int[], ConvolutionMode, long, long, String, Class<?>) Use InputTypeUtil.getOutputTypeCnnLayers(InputType, int[], int[], int[], int[], ConvolutionMode, long, long, String, CNN2DFormat, Class)
org.deeplearning4j.nn.multilayer.MultiLayerNetwork.getTrainingListeners() Use MultiLayerNetwork.getListeners()
org.deeplearning4j.nn.conf.MultiLayerConfiguration.Builder.inferenceWorkspaceMode(WorkspaceMode) Use NeuralNetConfiguration.Builder.inferenceWorkspaceMode(WorkspaceMode)
org.deeplearning4j.nn.weights.WeightInitUtil.initWeights(double, double, int[], WeightInit, Distribution, char, INDArray)
org.deeplearning4j.nn.weights.WeightInitUtil.initWeights(double, double, int[], WeightInit, Distribution, INDArray)
org.deeplearning4j.nn.api.Classifier.numLabels() Will be removed in a future release
org.deeplearning4j.nn.multilayer.MultiLayerNetwork.numLabels() Will be removed in a future release
org.deeplearning4j.nn.graph.ComputationGraph.params(boolean) To be removed. Use ComputationGraph.params()
org.deeplearning4j.nn.multilayer.MultiLayerNetwork.params(boolean) To be removed. Use MultiLayerNetwork.params() instead
org.deeplearning4j.eval.Evaluation.precision(EvaluationAveraging)
org.deeplearning4j.eval.Evaluation.recall(EvaluationAveraging)
org.deeplearning4j.util.ConvolutionUtils.reshapeMaskIfRequired(INDArray, INDArray, LayerWorkspaceMgr, ArrayType) Use ConvolutionUtils.reshapeMaskIfRequired(INDArray, INDArray, CNN2DFormat, LayerWorkspaceMgr, ArrayType)
org.deeplearning4j.eval.Evaluation.scoreForMetric(Evaluation.Metric) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.RegressionEvaluation.scoreForMetric(RegressionEvaluation.Metric) Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.nn.conf.inputs.InputType.InputTypeConvolutional.setDepth(long)
org.deeplearning4j.nn.conf.NeuralNetConfiguration.Builder.stepFunction(StepFunction)
org.deeplearning4j.nn.conf.MultiLayerConfiguration.Builder.trainingWorkspaceMode(WorkspaceMode) Use NeuralNetConfiguration.Builder.trainingWorkspaceMode(WorkspaceMode)
org.deeplearning4j.nn.conf.NeuralNetConfiguration.Builder.updater(Updater) Use NeuralNetConfiguration.Builder.updater(IUpdater)
org.deeplearning4j.nn.conf.layers.BaseLayer.Builder.updater(Updater)
org.deeplearning4j.nn.transferlearning.FineTuneConfiguration.Builder.updater(Updater) Use FineTuneConfiguration.Builder.updater(IUpdater)

Deprecated Constructors

Constructor and Description
org.deeplearning4j.earlystopping.termination.BestScoreEpochTerminationCondition(double, boolean) "lessBetter" argument no longer used
org.deeplearning4j.nn.conf.layers.SpaceToDepthLayer.Builder(int, SpaceToDepthLayer.DataFormat)
org.deeplearning4j.eval.ConfusionMatrix() Use ConfusionMatrix
org.deeplearning4j.eval.ConfusionMatrix(ConfusionMatrix<T>) Use ConfusionMatrix
org.deeplearning4j.eval.ConfusionMatrix(List<T>) Use ConfusionMatrix
org.deeplearning4j.eval.Evaluation() Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(double) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(double, Integer) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(INDArray) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(int) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(int, Integer) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(List<String>) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(List<String>, INDArray) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(List<String>, int) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.Evaluation(Map<Integer, String>) Use ND4J Evaluation class, which has the same interface: Evaluation.Metric
org.deeplearning4j.eval.EvaluationBinary(INDArray)
org.deeplearning4j.eval.EvaluationBinary(int, Integer)
org.deeplearning4j.eval.EvaluationCalibration() Use EvaluationCalibration
org.deeplearning4j.eval.EvaluationCalibration(int, int) Use EvaluationCalibration
org.deeplearning4j.eval.EvaluationCalibration(int, int, boolean) Use EvaluationCalibration
org.deeplearning4j.nn.conf.layers.variational.ExponentialReconstructionDistribution(String) Use ExponentialReconstructionDistribution.ExponentialReconstructionDistribution(Activation)
org.deeplearning4j.eval.curves.Histogram(String, double, double, int[]) Use Histogram
org.deeplearning4j.eval.curves.PrecisionRecallCurve(double[], double[], double[], int[], int[], int[], int) Use ReliabilityDiagram
org.deeplearning4j.eval.RegressionEvaluation() Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.RegressionEvaluation(List<String>) Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.RegressionEvaluation(List<String>, long) Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.RegressionEvaluation(long) Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.RegressionEvaluation(long, long) Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.RegressionEvaluation(String...) Use ND4J RegressionEvaluation class, which has the same interface: RegressionEvaluation
org.deeplearning4j.eval.curves.ReliabilityDiagram(String, double[], double[]) Use ReliabilityDiagram
org.deeplearning4j.eval.ROC() Use ROC
org.deeplearning4j.eval.ROC(int) Use ROC
org.deeplearning4j.eval.ROC(int, boolean) Use ROC
org.deeplearning4j.eval.ROC(int, boolean, int) Use ROC
org.deeplearning4j.eval.ROCBinary() Use ROCBinary
org.deeplearning4j.eval.ROCBinary(int) Use ROCBinary
org.deeplearning4j.eval.ROCBinary(int, boolean) Use ROCBinary
org.deeplearning4j.eval.curves.RocCurve(double[], double[], double[]) Use RocCurve
org.deeplearning4j.eval.ROCMultiClass() Use ROCMultiClass
org.deeplearning4j.eval.ROCMultiClass(int) Use ROCMultiClass
org.deeplearning4j.eval.ROCMultiClass(int, boolean) Use ROCMultiClass

Deprecated Enum Constants

Enum Constant and Description
org.deeplearning4j.nn.conf.Updater.CUSTOM
org.deeplearning4j.nn.conf.WorkspaceMode.SEPARATE Use WorkspaceMode.ENABLED instead
org.deeplearning4j.nn.conf.WorkspaceMode.SINGLE Use WorkspaceMode.ENABLED instead