weka.dl4j

Interface ApiWrapper<T>

Type Parameters:

T - Class which is to be wrapped.

All Known Implementing Classes:

AbstractDropout, AbstractWeightNoise, Activation, ActivationCube, ActivationELU, ActivationHardSigmoid, ActivationHardTanH, ActivationIdentity, ActivationLayer, ActivationLReLU, ActivationRationalTanh, ActivationReLU, ActivationRReLU, ActivationSigmoid, ActivationSoftmax, ActivationSoftPlus, ActivationSoftSign, ActivationSwish, ActivationTanH, AdaDelta, AdaGrad, Adam, AdaMax, AlgoMode, AlphaDropout, BatchNormalization, BinomialDistribution, BroadcastLambdaLayer, CenterLossOutputLayer, CharacterNGramTokenizerFactory, CommonPreProcessor, ConstantDistribution, ConstantSchedule, ConvolutionLayer, ConvolutionMode, Cropping2D, DefaultStepFunction, DefaultTokenizerFactory, DenseLayer, DepthwiseConvolution2DLayer, Disabled, Disabled, Disabled, Distribution, DropConnect, Dropout, DropoutLayer, EndingPreProcessor, ExponentialSchedule, FeedForwardLayer, GaussianDropout, GaussianNoise, GlobalPoolingLayer, GradientNormalization, GradientStepFunction, GravesLSTM, InverseSchedule, Layer, LocalResponseNormalization, LogNormalDistribution, LossBinaryXENT, LossCosineProximity, LossFMeasure, LossFunction, LossHinge, LossKLD, LossL1, LossL2, LossLayer, LossMAPE, LossMCXENT, LossMixtureDensity, LossMSE, LossMSLE, LossMultiLabel, LossNegativeLogLikelihood, LossPoisson, LossSquaredHinge, LowCasePreProcessor, LSTM, MapSchedule, Nadam, NegativeDefaultStepFunction, NegativeGradientStepFunction, Nesterovs, NGramTokenizerFactory, NoOp, NoParamLayer, NormalDistribution, OrthogonalDistribution, OutputLayer, PolySchedule, PoolingType, PretrainedType, RmsProp, RnnOutputLayer, SameDiffLayer, Schedule, ScheduleType, SeperableConvolution2DLayer, Sgd, SigmoidSchedule, StemmingPreProcessor, StepFunction, StepSchedule, SubsamplingLayer, TokenizerFactory, TokenPreProcess, TruncatedNormalDistribution, TweetNLPTokenizerFactory, UniformDistribution, Updater, WeightNoise, ZeroPaddingLayer

public interface ApiWrapper<T>

A general interface to access a backend object of a certain deeplearning4j class which is wrapped in a class to make it usable in weka.

Method Summary

Modifier and Type	Method and Description
T	getBackend() Access the DL4J backend.
void	initializeBackend() Initialize the DL4J backend.
void	setBackend(T newBackend) Set the DL4J backend.

Method Detail

getBackend

T getBackend()

Access the DL4J backend.

Returns:

DL4J backend for this wrapper

setBackend

void setBackend(T newBackend)

Set the DL4J backend.

Parameters:

newBackend - Backend that should be wrapped by this class

initializeBackend

void initializeBackend()

Initialize the DL4J backend.