Using @SuperBuilder for LayerConfigurations

Signed-off-by: brian <brian@brutex.de>

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/gradientcheck/RnnGradientChecks.java

@@ -267,11 +267,11 @@ public class RnnGradientChecks extends BaseDL4JTest {
                                 .activation(Activation.TANH)
                                 .updater(new NoOp())
                                 .weightInit(WeightInit.XAVIER)
-                                .list()
+
                                 .layer(simple ? SimpleRnn.builder().nOut(layerSize).hasLayerNorm(hasLayerNorm).build() :
                                         LSTM.builder().nOut(layerSize).build())
-                                .layer(new LastTimeStep(simple ? SimpleRnn.builder().nOut(layerSize).hasLayerNorm(hasLayerNorm).build() :
+                                .layer(LastTimeStep.builder().underlying(simple ? SimpleRnn.builder().nOut(layerSize).hasLayerNorm(hasLayerNorm).build() :
-                                        LSTM.builder().nOut(layerSize).build()))
+                                        LSTM.builder().nOut(layerSize).build()).build())
                                 .layer(OutputLayer.builder().nOut(nOut).activation(Activation.SOFTMAX)
                                         .lossFunction(LossFunctions.LossFunction.MCXENT).build())
                                 .inputType(InputType.recurrent(nIn))
@@ -335,7 +335,7 @@ public class RnnGradientChecks extends BaseDL4JTest {
                         .weightInit(WeightInit.XAVIER)
                         .list()
                         .layer(LSTM.builder().nOut(layerSize).build())
-                        .layer(new TimeDistributed(DenseLayer.builder().nOut(layerSize).activation(Activation.SOFTMAX).build()))
+                        .layer(TimeDistributed.builder().underlying(DenseLayer.builder().nOut(layerSize).activation(Activation.SOFTMAX).build()).build())
                         .layer(RnnOutputLayer.builder().nOut(nOut).activation(Activation.SOFTMAX)
                                 .lossFunction(LossFunctions.LossFunction.MCXENT).build())
                         .inputType(InputType.recurrent(nIn))

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/dtypes/DTypeTests.java

@@ -482,7 +482,7 @@ public class DTypeTests extends BaseDL4JTest {
                             break;
                         case 1:
                             ol = LossLayer.builder().activation(Activation.SOFTMAX).lossFunction(LossFunctions.LossFunction.MCXENT.getILossFunction()).build();
-                            secondLast = new FrozenLayerWithBackprop(DenseLayer.builder().nOut(10).activation(Activation.SIGMOID).build());
+                            secondLast = FrozenLayerWithBackprop.builder().underlying(DenseLayer.builder().nOut(10).activation(Activation.SIGMOID).build()).build();
                             break;
                         case 2:
                             ol =CenterLossOutputLayer.builder().nOut(10).activation(Activation.SOFTMAX).lossFunction(LossFunctions.LossFunction.MCXENT).build();
@@ -889,7 +889,7 @@ public class DTypeTests extends BaseDL4JTest {
                             break;
                         case 2:
                             ol = OutputLayer.builder().nOut(5).build();
-                            secondLast = new LastTimeStep(SimpleRnn.builder().nOut(5).activation(Activation.TANH).build());
+                            secondLast = LastTimeStep.builder().underlying(SimpleRnn.builder().nOut(5).activation(Activation.TANH).build()).build();
                             break;
                         default:
                             throw new RuntimeException();
@@ -905,7 +905,7 @@ public class DTypeTests extends BaseDL4JTest {
                             .layer(DenseLayer.builder().nOut(5).build())
                             .layer(GravesBidirectionalLSTM.builder().nIn(5).nOut(5).activation(Activation.TANH).build())
                             .layer(Bidirectional.builder(LSTM.builder().nIn(5).nOut(5).activation(Activation.TANH).build()).build())
-                            .layer(new TimeDistributed(DenseLayer.builder().nIn(10).nOut(5).activation(Activation.TANH).build()))
+                            .layer(TimeDistributed.builder().underlying(DenseLayer.builder().nIn(10).nOut(5).activation(Activation.TANH).build()).build())
                             .layer(SimpleRnn.builder().nIn(5).nOut(5).build())
                             .layer(MaskZeroLayer.builder().underlying(SimpleRnn.builder().nIn(5).nOut(5).build()).maskingValue(0.0).build())
                             .layer(secondLast)
@@ -1062,7 +1062,7 @@ public class DTypeTests extends BaseDL4JTest {
                         INDArray input;
                         if (test == 0) {
                             if (frozen) {
-                                conf.layer("0", new FrozenLayer(EmbeddingLayer.builder().nIn(5).nOut(5).build()), "in");
+                                conf.layer("0", FrozenLayer.builder(EmbeddingLayer.builder().nIn(5).nOut(5).build()).build(), "in");
                             } else {
                                 conf.layer("0", EmbeddingLayer.builder().nIn(5).nOut(5).build(), "in");
                             }
@@ -1071,7 +1071,7 @@ public class DTypeTests extends BaseDL4JTest {
                             conf.setInputTypes(InputType.feedForward(1));
                         } else if (test == 1) {
                             if (frozen) {
-                                conf.layer("0", new FrozenLayer(EmbeddingSequenceLayer.builder().nIn(5).nOut(5).build()), "in");
+                                conf.layer("0", FrozenLayer.builder(EmbeddingSequenceLayer.builder().nIn(5).nOut(5).build()).build(), "in");
                             } else {
                                 conf.layer("0", EmbeddingSequenceLayer.builder().nIn(5).nOut(5).build(), "in");
                             }

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/graph/TestComputationGraphNetwork.java

@@ -1925,7 +1925,7 @@ public class TestComputationGraphNetwork extends BaseDL4JTest {
                 .setOutputs("output")
                 .addLayer("0", ConvolutionLayer.builder().nOut(5).convolutionMode(ConvolutionMode.Same).build(),"input" )
                 .addVertex("dummyAdd", new ElementWiseVertex(ElementWiseVertex.Op.Add), "0")
-                .addLayer("output", new CnnLossLayer(), "dummyAdd")
+                .addLayer("output", CnnLossLayer.builder(), "dummyAdd")
                 .build());
         graph.init();
         graph.outputSingle(Nd4j.randn(1, 2, 10, 10));

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/layers/FrozenLayerTest.java

@@ -289,11 +289,11 @@ public class FrozenLayerTest extends BaseDL4JTest {
                         .build();
 
         NeuralNetConfiguration conf2 = NeuralNetConfiguration.builder().seed(12345).list().layer(0,
-                        new org.deeplearning4j.nn.conf.layers.misc.FrozenLayer(DenseLayer.builder().nIn(10).nOut(10)
+                        org.deeplearning4j.nn.conf.layers.misc.FrozenLayer.builder(DenseLayer.builder().nIn(10).nOut(10)
-                                        .activation(Activation.TANH).weightInit(WeightInit.XAVIER).build()))
+                                        .activation(Activation.TANH).weightInit(WeightInit.XAVIER).build()).build())
-                        .layer(1, new org.deeplearning4j.nn.conf.layers.misc.FrozenLayer(
+                        .layer(1, org.deeplearning4j.nn.conf.layers.misc.FrozenLayer.builder(
                                         DenseLayer.builder().nIn(10).nOut(10).activation(Activation.TANH)
-                                                        .weightInit(WeightInit.XAVIER).build()))
+                                                        .weightInit(WeightInit.XAVIER).build()).build())
                         .layer(2, org.deeplearning4j.nn.conf.layers.OutputLayer.builder().lossFunction(
                                         LossFunctions.LossFunction.MCXENT).activation(Activation.SOFTMAX).nIn(10)
                                                         .nOut(10).build())

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/layers/FrozenLayerWithBackpropTest.java

@@ -60,11 +60,11 @@ public class FrozenLayerWithBackpropTest extends BaseDL4JTest {
                 .build();
 
         NeuralNetConfiguration conf2 = NeuralNetConfiguration.builder().seed(12345).list().layer(0,
-                new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(DenseLayer.builder().nIn(10).nOut(10)
+                org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(DenseLayer.builder().nIn(10).nOut(10)
-                        .activation(Activation.TANH).weightInit(WeightInit.XAVIER).build()))
+                        .activation(Activation.TANH).weightInit(WeightInit.XAVIER).build()).build())
-                .layer(1, new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .layer(1, org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                         DenseLayer.builder().nIn(10).nOut(10).activation(Activation.TANH)
-                                .weightInit(WeightInit.XAVIER).build()))
+                                .weightInit(WeightInit.XAVIER).build()).build())
                 .layer(2, OutputLayer.builder(
                         LossFunctions.LossFunction.MCXENT).activation(Activation.SOFTMAX).nIn(10)
                         .nOut(10).build())
@@ -113,10 +113,10 @@ public class FrozenLayerWithBackpropTest extends BaseDL4JTest {
 
         ComputationGraphConfiguration conf2 = NeuralNetConfiguration.builder().seed(12345).graphBuilder()
                 .addInputs("in")
-                .addLayer("0", new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer("0", org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                         DenseLayer.builder().nIn(10).nOut(10).activation(Activation.TANH)
                                 .weightInit(WeightInit.XAVIER).build()), "in")
-                .addLayer("1", new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer("1", org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                         DenseLayer.builder().nIn(10).nOut(10).activation(Activation.TANH)
                                 .weightInit(WeightInit.XAVIER).build()), "0")
                 .addLayer("2", OutputLayer.builder(
@@ -160,11 +160,11 @@ public class FrozenLayerWithBackpropTest extends BaseDL4JTest {
                 .updater(new Sgd(2))
                 .list()
                 .layer(DenseLayer.builder().nIn(4).nOut(3).build())
-                .layer(new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .layer(org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                                 DenseLayer.builder().nIn(3).nOut(4).build()))
-                .layer(new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .layer(org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                                 DenseLayer.builder().nIn(4).nOut(2).build()))
-                .layer(new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .layer(org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                                 OutputLayer.builder().lossFunction(LossFunctions.LossFunction.MSE).activation(Activation.TANH).nIn(2).nOut(1).build()))
                 .build();
 
@@ -213,15 +213,15 @@ public class FrozenLayerWithBackpropTest extends BaseDL4JTest {
                 .addInputs("input")
                 .addLayer(initialLayer, DenseLayer.builder().nIn(4).nOut(4).build(),"input")
                 .addLayer(frozenBranchUnfrozenLayer0, DenseLayer.builder().nIn(4).nOut(3).build(),initialLayer)
-                .addLayer(frozenBranchFrozenLayer1, new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer(frozenBranchFrozenLayer1, org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                         DenseLayer.builder().nIn(3).nOut(4).build()),frozenBranchUnfrozenLayer0)
-                .addLayer(frozenBranchFrozenLayer2, new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer(frozenBranchFrozenLayer2, org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                         DenseLayer.builder().nIn(4).nOut(2).build()),frozenBranchFrozenLayer1)
                 .addLayer(unfrozenLayer0, DenseLayer.builder().nIn(4).nOut(4).build(),initialLayer)
                 .addLayer(unfrozenLayer1, DenseLayer.builder().nIn(4).nOut(2).build(),unfrozenLayer0)
                 .addLayer(unfrozenBranch2, DenseLayer.builder().nIn(2).nOut(1).build(),unfrozenLayer1)
                 .addVertex("merge", new MergeVertex(), frozenBranchFrozenLayer2, unfrozenBranch2)
-                .addLayer(frozenBranchOutput,new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer(frozenBranchOutput,org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                         OutputLayer.builder().lossFunction(LossFunctions.LossFunction.MSE).activation(Activation.TANH).nIn(3).nOut(1).build()),"merge")
                 .setOutputs(frozenBranchOutput)
                 .build();
@@ -269,9 +269,9 @@ public class FrozenLayerWithBackpropTest extends BaseDL4JTest {
                 .updater(new Sgd(2))
                 .list()
                 .layer(0,DenseLayer.builder().nIn(4).nOut(3).build())
-                .layer(1,new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(DenseLayer.builder().nIn(3).nOut(4).build()))
+                .layer(1,org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(DenseLayer.builder().nIn(3).nOut(4).build()))
-                .layer(2,new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(DenseLayer.builder().nIn(4).nOut(2).build()))
+                .layer(2,org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(DenseLayer.builder().nIn(4).nOut(2).build()))
-                .layer(3,new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(OutputLayer.builder().lossFunction(LossFunctions.LossFunction.MSE).activation(Activation.TANH).nIn(2).nOut(1).build()))
+                .layer(3,org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(OutputLayer.builder().lossFunction(LossFunctions.LossFunction.MSE).activation(Activation.TANH).nIn(2).nOut(1).build()))
                 .build();
         MultiLayerNetwork frozenNetwork = new MultiLayerNetwork(confFrozen);
         frozenNetwork.init();
@@ -327,16 +327,16 @@ public class FrozenLayerWithBackpropTest extends BaseDL4JTest {
                 .addInputs("input")
                 .addLayer(initialLayer,DenseLayer.builder().nIn(4).nOut(4).build(),"input")
                 .addLayer(frozenBranchUnfrozenLayer0,DenseLayer.builder().nIn(4).nOut(3).build(), initialLayer)
-                .addLayer(frozenBranchFrozenLayer1,new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer(frozenBranchFrozenLayer1,org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                                 DenseLayer.builder().nIn(3).nOut(4).build()),frozenBranchUnfrozenLayer0)
                 .addLayer(frozenBranchFrozenLayer2,
-                        new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                        org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                                 DenseLayer.builder().nIn(4).nOut(2).build()),frozenBranchFrozenLayer1)
                 .addLayer(unfrozenLayer0,DenseLayer.builder().nIn(4).nOut(4).build(),initialLayer)
                 .addLayer(unfrozenLayer1,DenseLayer.builder().nIn(4).nOut(2).build(),unfrozenLayer0)
                 .addLayer(unfrozenBranch2,DenseLayer.builder().nIn(2).nOut(1).build(),unfrozenLayer1)
                 .addVertex("merge",new MergeVertex(), frozenBranchFrozenLayer2, unfrozenBranch2)
-                .addLayer(frozenBranchOutput, new org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop(
+                .addLayer(frozenBranchOutput, org.deeplearning4j.nn.conf.layers.misc.FrozenLayerWithBackprop.builder(
                                 OutputLayer.builder().lossFunction(LossFunctions.LossFunction.MSE).activation(Activation.TANH).nIn(3).nOut(1).build()),"merge")
                 .setOutputs(frozenBranchOutput)
                 .build();

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/layers/recurrent/RnnDataFormatTests.java

@@ -243,7 +243,7 @@ public class RnnDataFormatTests extends BaseDL4JTest {
             layer = MaskZeroLayer.builder().maskingValue(0.).underlying(layer).build();
         }
         if(lastTimeStep){
-            layer = new LastTimeStep(layer);
+            layer = LastTimeStep.builder(layer);
         }
         NeuralNetConfiguration.NeuralNetConfigurationBuilder builder = (NeuralNetConfiguration.NeuralNetConfigurationBuilder) NeuralNetConfiguration.builder()
                 .seed(12345)

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/layers/recurrent/TestLastTimeStepLayer.java

@@ -63,7 +63,7 @@ public class TestLastTimeStepLayer extends BaseDL4JTest {
     public void testLastTimeStepVertex() {
 
         ComputationGraphConfiguration conf = NeuralNetConfiguration.builder().graphBuilder().addInputs("in")
-                .addLayer("lastTS", new LastTimeStep(SimpleRnn.builder()
+                .addLayer("lastTS", LastTimeStep.builder(SimpleRnn.builder()
                         .nIn(5).nOut(6).dataFormat(rnnDataFormat).build()), "in")
                 .setOutputs("lastTS")
                 .build();
@@ -134,7 +134,7 @@ public class TestLastTimeStepLayer extends BaseDL4JTest {
                 .graphBuilder()
                 .addInputs("in")
                 .setInputTypes(InputType.recurrent(1, rnnDataFormat))
-                .addLayer("RNN", new LastTimeStep(LSTM.builder()
+                .addLayer("RNN", LastTimeStep.builder(LSTM.builder()
                         .nOut(10).dataFormat(rnnDataFormat)
                         .build()), "in")
                 .addLayer("dense", DenseLayer.builder()

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/layers/recurrent/TestTimeDistributed.java

@@ -79,9 +79,8 @@ public class TestTimeDistributed extends BaseDL4JTest {
                     .inferenceWorkspaceMode(wsm)
                     .seed(12345)
                     .updater(new Adam(0.1))
-                    .list()
                     .layer(LSTM.builder().nIn(3).nOut(3).dataFormat(rnnDataFormat).build())
-                    .layer(new TimeDistributed(DenseLayer.builder().nIn(3).nOut(3).activation(Activation.TANH).build(), rnnDataFormat))
+                    .layer(TimeDistributed.builder().underlying(DenseLayer.builder().nIn(3).nOut(3).activation(Activation.TANH).build()).rnnDataFormat(rnnDataFormat))
                     .layer(RnnOutputLayer.builder().nIn(3).nOut(3).activation(Activation.SOFTMAX).dataFormat(rnnDataFormat)
                             .lossFunction(LossFunctions.LossFunction.MCXENT).build())
                     .inputType(InputType.recurrent(3, rnnDataFormat))

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/multilayer/TestMasking.java

@@ -314,7 +314,7 @@ public class TestMasking extends BaseDL4JTest {
                 )
                 .addInputs("m1", "m2")
                 .addVertex("stack", new StackVertex(), "m1", "m2")
-                .addLayer("lastUnStacked", new LastTimeStep(LSTM.builder().nIn(3).nOut(1).activation(Activation.TANH).build()), "stack")
+                .addLayer("lastUnStacked", LastTimeStep.builder(LSTM.builder().nIn(3).nOut(1).activation(Activation.TANH).build()), "stack")
                 .addVertex("unstacked1", new UnstackVertex(0, 2), "lastUnStacked")
                 .addVertex("unstacked2", new UnstackVertex(1, 2), "lastUnStacked")
                 .addVertex("restacked", new StackVertex(), "unstacked1", "unstacked2")

cavis-dnn/cavis-dnn-core/src/test/java/org/deeplearning4j/nn/transferlearning/TransferLearningCompGraphTest.java

@@ -336,12 +336,12 @@ public class TransferLearningCompGraphTest extends BaseDL4JTest {
                         new ComputationGraph(overallConf.graphBuilder().addInputs("layer0In")
                                 .setInputTypes(InputType.convolutionalFlat(28,28, 3))
                                 .addLayer("layer0",
-                                        new FrozenLayer(ConvolutionLayer.builder(5, 5).nIn(3)
+                                        FrozenLayer.builder(ConvolutionLayer.builder(5, 5).nIn(3)
                                                 .stride(1, 1).nOut(20)
                                                 .activation(Activation.IDENTITY).build()),
                                         "layer0In")
                                 .addLayer("layer1",
-                                        new FrozenLayer(SubsamplingLayer.builder(
+                                        FrozenLayer.builder(SubsamplingLayer.builder(
                                                 SubsamplingLayer.PoolingType.MAX)
                                                 .kernelSize(2, 2).stride(2, 2)
                                                 .build()),
@@ -430,11 +430,11 @@ public class TransferLearningCompGraphTest extends BaseDL4JTest {
                         .weightInit(WeightInit.XAVIER)
                         .graphBuilder().addInputs("in")
                         .addLayer("blstm1",
-                                        new FrozenLayer(GravesBidirectionalLSTM.builder().nIn(10).nOut(10)
+                                        FrozenLayer.builder(GravesBidirectionalLSTM.builder().nIn(10).nOut(10)
                                                         .activation(Activation.TANH).build()),
                                         "in")
-                        .addLayer("pool", new FrozenLayer(GlobalPoolingLayer.builder().build()), "blstm1")
+                        .addLayer("pool", FrozenLayer.builder(GlobalPoolingLayer.builder().build()), "blstm1")
-                        .addLayer("dense", new FrozenLayer(DenseLayer.builder().nIn(10).nOut(10).build()), "pool")
+                        .addLayer("dense", FrozenLayer.builder(DenseLayer.builder().nIn(10).nOut(10).build()), "pool")
                         .addLayer("out", OutputLayer.builder().nIn(10).nOut(5).activation(Activation.SOFTMAX)
                                         .updater(new Adam(0.1))
                                         .lossFunction(LossFunctions.LossFunction.MCXENT).build(), "dense")

cavis-dnn/cavis-dnn-modelimport/src/main/java/org/deeplearning4j/nn/modelimport/keras/layers/recurrent/KerasLSTM.java

@@ -203,7 +203,7 @@ public class KerasLSTM extends KerasLayer {
 
         this.layer = builder.build();
         if (!returnSequences) {
-            this.layer = new LastTimeStep(this.layer);
+            this.layer = LastTimeStep.builder(this.layer);
         }
         if (maskingConfig.getFirst()) {
             this.layer = new MaskZeroLayer(this.layer, maskingConfig.getSecond());

cavis-dnn/cavis-dnn-modelimport/src/main/java/org/deeplearning4j/nn/modelimport/keras/layers/recurrent/KerasSimpleRnn.java

@@ -174,7 +174,7 @@ public class KerasSimpleRnn extends KerasLayer {
 
         this.layer = builder.build();
         if (!returnSequences) {
-            this.layer = new LastTimeStep(this.layer);
+            this.layer = LastTimeStep.builder(this.layer);
         }
         if (maskingConfig.getFirst()) {
             this.layer = new MaskZeroLayer(this.layer, maskingConfig.getSecond());

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/ComputationGraphConfiguration.java

@@ -819,6 +819,9 @@ public class ComputationGraphConfiguration implements Serializable, Cloneable {
         public GraphBuilder addLayer(String layerName, LayerConfiguration layer, String... layerInputs) {
             return addLayer(layerName, layer, null, layerInputs);
         }
+        public GraphBuilder addLayer(String layerName, LayerConfiguration.LayerConfigurationBuilder<?,?> layer, String... layerInputs) {
+            return addLayer(layerName, layer.build(), null, layerInputs);
+        }
 
         /**
          * Add a layer, with no {@link InputPreProcessor}, with the specified name

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/NeuralNetBaseBuilderConfiguration.java

@@ -661,7 +661,17 @@ public abstract class NeuralNetBaseBuilderConfiguration implements INeuralNetwor
     public B layer(Integer index, @NonNull LayerConfiguration layer) {
       innerConfigurations$value.add(index, layer);
       innerConfigurations$set = true;
-      return (B) this;
+      return self();
+    }
+    /**
+     * Set layer at index
+     *
+     * @param index where to insert
+     * @param layer the layer
+     * @return builder
+     */
+    public B layer(Integer index, @NonNull LayerConfiguration.LayerConfigurationBuilder<?,?> layer) {
+      return this.layer(index, layer.build());
     }
 
     /**
@@ -675,6 +685,9 @@ public abstract class NeuralNetBaseBuilderConfiguration implements INeuralNetwor
       innerConfigurations$set = true;
       return (B) this;
     }
+    public B layer(@NonNull LayerConfiguration.LayerConfigurationBuilder<?, ?> layer) {
+      return this.layer(layer.build());
+    }
 
     // TODO this is a dirty workaround
     public boolean isOverrideNinUponBuild() {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/BaseLayerConfiguration.java

@@ -212,6 +212,16 @@ public abstract class BaseLayerConfiguration extends LayerConfiguration
           C extends BaseLayerConfiguration, B extends BaseLayerConfigurationBuilder<C, B>>
       extends LayerConfigurationBuilder<C, B> {
 
+public B updater(Updater upd) {
+  this.updater = upd.getIUpdaterWithDefaultConfig();
+  return self();
+}
+
+
+    public B updater(IUpdater upd) {
+      this.updater = upd;
+      return self();
+    }
 
     /**
      * Set weight initialization scheme to random sampling via the specified distribution.

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/CapsuleLayer.java

@@ -38,7 +38,7 @@ import org.nd4j.linalg.api.memory.MemoryWorkspace;
 import org.nd4j.linalg.api.ndarray.INDArray;
 import org.nd4j.linalg.factory.Nd4j;
 
-@Data
+
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")
 public class CapsuleLayer extends SameDiffLayer {
@@ -50,33 +50,33 @@ public class CapsuleLayer extends SameDiffLayer {
      * @param hasBias
      * @return
      */
-    @Builder.Default private boolean hasBias = false;
+    @Builder.Default @Getter @Setter private boolean hasBias = false;
     /**
      * Usually inferred automatically.
      * @param inputCapsules
      * @return
      */
-    @Builder.Default private long inputCapsules = 0;
+    @Builder.Default @Getter @Setter private long inputCapsules = 0;
     /**
      * Usually inferred automatically.
      * @param inputCapsuleDimensions
      * @return
      */
-    @Builder.Default private long inputCapsuleDimensions = 0;
+    @Builder.Default @Getter @Setter private long inputCapsuleDimensions = 0;
     /**
      * Set the number of capsules to use.
      * @param capsules
      * @return
      */
-    private int capsules;
+    @Getter @Setter private int capsules;
-    private int capsuleDimensions;
+    @Getter @Setter private int capsuleDimensions;
     /**
      * Set the number of dynamic routing iterations to use.
      * The default is 3 (recommendedded in Dynamic Routing Between Capsules)
      * @param routings
      * @return
      */
-    @Builder.Default private int routings = 3;
+    @Builder.Default @Getter @Setter private int routings = 3;
 
     @Override
     public void setNIn(InputType inputType, boolean override) {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/CapsuleStrengthLayer.java

@@ -32,7 +32,6 @@ import org.nd4j.autodiff.samediff.SDVariable;
 import org.nd4j.autodiff.samediff.SameDiff;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder
 public class CapsuleStrengthLayer extends SameDiffLambdaLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/CnnLossLayer.java

@@ -39,7 +39,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 import org.nd4j.linalg.lossfunctions.ILossFunction;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/ConvolutionLayer.java

@@ -46,7 +46,7 @@ import org.nd4j.linalg.api.ndarray.INDArray;
  * to be used in the net or in other words the channels The builder specifies the filter/kernel
  * size, the stride and padding The pooling layer takes the kernel size
  */
-@Data
+
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")
@@ -72,7 +72,7 @@ public class ConvolutionLayer extends FeedForwardLayer {
    *
    * @param format Format for activations (in and out)
    */
-  @Builder.Default
+  @Builder.Default @Getter @Setter
   private CNN2DFormat convFormat =
       CNN2DFormat.NCHW; // default value for legacy serialization reasons
 
@@ -86,24 +86,29 @@ public class ConvolutionLayer extends FeedForwardLayer {
    * http://deeplearning.net/software/theano/tutorial/conv_arithmetic.html#dilated-convolutions</a>
    * <br>
    */
+  @Getter @Setter
   private @Builder.Default int[] dilation = new int[] {1, 1};
   /** Default is 2. Down-sample by a factor of 2 */
+  @Getter @Setter
   private @Builder.Default int[] stride = new int[] {1, 1};
-
+@Getter @Setter
   private @Builder.Default int[] padding = new int[] {0, 0};
   /**
    * When using CuDNN and an error is encountered, should fallback to the non-CuDNN implementatation
    * be allowed? If set to false, an exception in CuDNN will be propagated back to the user. If
    * false, the built-in (non-CuDNN) implementation for ConvolutionLayer will be used
    */
+  @Getter
   @Builder.Default private boolean cudnnAllowFallback = true;
 
   /** Defaults to "PREFER_FASTEST", but "NO_WORKSPACE" uses less memory. */
+  @Getter
   @Builder.Default private AlgoMode cudnnAlgoMode = AlgoMode.PREFER_FASTEST;
 
   private FwdAlgo cudnnFwdAlgo;
   private BwdFilterAlgo cudnnBwdFilterAlgo;
   private BwdDataAlgo cudnnBwdDataAlgo;
+  @Getter @Setter
   @Builder.Default private int convolutionDim = 2; // 2D convolution by default
   /** Causal convolution - allowed for 1D only */
   @Builder.Default private boolean allowCausal = false;

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/EmbeddingLayer.java

@@ -45,7 +45,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/EmbeddingSequenceLayer.java

@@ -44,7 +44,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/GravesBidirectionalLSTM.java

@@ -39,7 +39,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 import java.util.*;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @Deprecated

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/LSTM.java

@@ -41,7 +41,6 @@ import java.util.Collections;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/LayerConfiguration.java

@@ -48,7 +48,7 @@ import org.nd4j.linalg.learning.config.IUpdater;
 import org.nd4j.linalg.learning.regularization.Regularization;
 
 /** A neural network layer. */
-@JsonTypeInfo(use = JsonTypeInfo.Id.CLASS, include = JsonTypeInfo.As.PROPERTY, property = "@class")
+//@JsonTypeInfo(use = JsonTypeInfo.Id.CLASS, include = JsonTypeInfo.As.PROPERTY, property = "@class")
 @EqualsAndHashCode
 // @JsonIdentityInfo(generator= ObjectIdGenerators.IntSequenceGenerator.class, property="@id")
 @Slf4j

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/LocalResponseNormalization.java

@@ -40,7 +40,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 import org.nd4j.linalg.learning.regularization.Regularization;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/LocallyConnected2D.java

@@ -47,7 +47,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 import org.nd4j.linalg.factory.Nd4j;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @JsonIgnoreProperties({"paramShapes"})
 @SuperBuilder(buildMethodName = "initBuild")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/LossLayer.java

@@ -40,7 +40,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 import org.nd4j.linalg.lossfunctions.ILossFunction;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/OutputLayer.java

@@ -37,7 +37,6 @@ import org.nd4j.linalg.lossfunctions.ILossFunction;
 import org.nd4j.linalg.lossfunctions.LossFunctions;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/PReLULayer.java

@@ -38,7 +38,6 @@ import org.nd4j.linalg.api.buffer.DataType;
 import org.nd4j.linalg.api.ndarray.INDArray;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Pooling1D.java

@@ -32,7 +32,6 @@ import lombok.ToString;
  * @author Max Pumperla
  */
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 public class Pooling1D extends Subsampling1DLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Pooling2D.java

@@ -32,7 +32,6 @@ import lombok.ToString;
  * @author Max Pumperla
  */
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 public class Pooling2D extends SubsamplingLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/PrimaryCapsules.java

@@ -40,7 +40,6 @@ import org.nd4j.linalg.api.ops.impl.layers.convolution.config.Conv2DConfig;
 import org.nd4j.linalg.factory.Nd4j;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")
 public class PrimaryCapsules extends SameDiffLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/RecurrentAttentionLayer.java

@@ -42,7 +42,6 @@ import org.nd4j.linalg.factory.Nd4j;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild")
 public class RecurrentAttentionLayer extends SameDiffLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/RnnLossLayer.java

@@ -40,7 +40,6 @@ import org.nd4j.linalg.lossfunctions.ILossFunction;
 import org.nd4j.linalg.lossfunctions.LossFunctions;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/RnnOutputLayer.java

@@ -40,7 +40,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 import org.nd4j.linalg.lossfunctions.LossFunctions;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/SelfAttentionLayer.java

@@ -38,7 +38,6 @@ import org.nd4j.linalg.factory.Nd4j;
 
 @Data
 @EqualsAndHashCode(callSuper = true)
-@NoArgsConstructor()
 @SuperBuilder(buildMethodName = "initBuild")
 public class SelfAttentionLayer extends SameDiffLayer {
   private static final String WEIGHT_KEY_QUERY_PROJECTION = "Wq";

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/SpaceToBatchLayer.java

@@ -39,7 +39,6 @@ import org.nd4j.linalg.api.buffer.DataType;
 import org.nd4j.linalg.api.ndarray.INDArray;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/SpaceToDepthLayer.java

@@ -38,7 +38,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Subsampling1DLayer.java

@@ -48,7 +48,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
  * wide.
  */
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Subsampling3DLayer.java

@@ -43,7 +43,6 @@ import org.nd4j.linalg.exception.ND4JArraySizeException;
 import org.nd4j.linalg.learning.regularization.Regularization;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder", buildMethodName = "initBuild")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/SubsamplingLayer.java

@@ -43,7 +43,6 @@ import org.nd4j.linalg.api.buffer.DataType;
 import org.nd4j.linalg.api.ndarray.INDArray;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(buildMethodName = "initBuild", builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Upsampling1D.java

@@ -41,7 +41,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Upsampling2D.java

@@ -39,7 +39,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/Upsampling3D.java

@@ -37,7 +37,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/ZeroPadding1DLayer.java

@@ -38,7 +38,6 @@ import org.nd4j.linalg.api.buffer.DataType;
 import org.nd4j.linalg.api.ndarray.INDArray;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")
 public class ZeroPadding1DLayer extends NoParamLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/ZeroPadding3DLayer.java

@@ -39,7 +39,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")
 public class ZeroPadding3DLayer extends NoParamLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/ZeroPaddingLayer.java

@@ -40,7 +40,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder", buildMethodName = "initBuild")
 public class ZeroPaddingLayer extends NoParamLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/convolutional/Cropping1D.java

@@ -40,7 +40,6 @@ import org.nd4j.linalg.api.ndarray.INDArray;
 
 /** Amount of cropping to apply to both the top and the bottom of the input activations */
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")
 public class Cropping1D extends NoParamLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/convolutional/Cropping2D.java

@@ -41,7 +41,6 @@ import org.nd4j.linalg.api.buffer.DataType;
 import org.nd4j.linalg.api.ndarray.INDArray;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")
 public class Cropping2D extends NoParamLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/convolutional/Cropping3D.java

@@ -40,7 +40,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder(builderMethodName = "innerBuilder")
 public class Cropping3D extends NoParamLayer {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/misc/ElementWiseMultiplicationLayer.java

@@ -42,7 +42,6 @@ import java.util.Map;
 @Data
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
-@NoArgsConstructor
 @SuperBuilder
 public class ElementWiseMultiplicationLayer extends org.deeplearning4j.nn.conf.layers.FeedForwardLayer {
 

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/misc/FrozenLayerWithBackprop.java

@@ -57,6 +57,9 @@ public class FrozenLayerWithBackprop extends BaseWrapperLayerConfiguration {
     public static FrozenLayerWithBackpropBuilder<?, ?> builder(LayerConfiguration innerConfiguration) {
         return innerBuilder().underlying(innerConfiguration);
     }
+    public static FrozenLayerWithBackpropBuilder<?, ?> builder(LayerConfigurationBuilder<?,?> innerConfiguration) {
+        return innerBuilder().underlying(innerConfiguration.build());
+    }
     public NeuralNetConfiguration getInnerConf(NeuralNetConfiguration conf) {
         NeuralNetConfiguration nnc = conf.clone();
         nnc.getLayerConfigurations().add(0, underlying);

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/misc/RepeatVector.java

@@ -39,7 +39,6 @@ import java.util.Collection;
 import java.util.Map;
 
 @Data
-@NoArgsConstructor
 @ToString(callSuper = true)
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/recurrent/Bidirectional.java

@@ -48,7 +48,6 @@ import java.util.Map;
 
 import static org.nd4j.linalg.indexing.NDArrayIndex.interval;
 
-@NoArgsConstructor
 @Data
 @EqualsAndHashCode(callSuper = true, exclude = {"initializer"})
 @JsonIgnoreProperties({"initializer"})

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/samediff/AbstractSameDiffLayer.java

@@ -61,6 +61,7 @@ public abstract class AbstractSameDiffLayer extends LayerConfiguration {
    * @param regularization Regularization to apply for the network parameters/weights (excluding
    *     biases)
    */
+  @Getter
   protected List<Regularization> regularization;
   /**
    * The regularization for the biases only - for example {@link WeightDecay} -- SETTER -- Set the
@@ -68,6 +69,7 @@ public abstract class AbstractSameDiffLayer extends LayerConfiguration {
    *
    * @param regularizationBias Regularization to apply for the network biases only
    */
+  @Getter
   protected List<Regularization> regularizationBias;
   /**
    * Gradient updater. For example, {@link org.nd4j.linalg.learning.config.Adam} or {@link
@@ -83,10 +85,11 @@ public abstract class AbstractSameDiffLayer extends LayerConfiguration {
    * @param biasUpdater Updater to use for bias parameters
    */
   protected @Getter @Setter IUpdater biasUpdater;
-
+@Getter  @Setter
   protected GradientNormalization gradientNormalization;
+@Getter @Setter
   protected double gradientNormalizationThreshold = Double.NaN;
-
+@Getter @Setter
   private SDLayerParams layerParams;
 
   @Override

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/samediff/SameDiffLambdaLayer.java

@@ -28,7 +28,6 @@ import org.nd4j.autodiff.samediff.SameDiff;
 import org.nd4j.linalg.api.ndarray.INDArray;
 
 import java.util.Map;
-@NoArgsConstructor
 @SuperBuilder
 public abstract class SameDiffLambdaLayer extends SameDiffLayer {
 

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/layers/variational/VariationalAutoencoder.java

@@ -45,7 +45,6 @@ import org.nd4j.linalg.lossfunctions.ILossFunction;
 import org.nd4j.linalg.lossfunctions.LossFunctions;
 
 @Data
-@NoArgsConstructor
 @EqualsAndHashCode(callSuper = true)
 @SuperBuilder
 public class VariationalAutoencoder extends BasePretrainNetwork {

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/conf/ocnn/OCNNOutputLayer.java

@@ -48,31 +48,31 @@ public class OCNNOutputLayer extends BaseOutputLayer {
    * The hidden layer size for the one class neural network. Note this would be nOut on a dense
    * layer. NOut in this neural net is always set to 1 though.
    */
-  @Builder.Default private int hiddenLayerSize; // embedded hidden layer size aka "K"
+  @Builder.Default @Getter private int hiddenLayerSize; // embedded hidden layer size aka "K"
   /** For nu definition see the paper */
-  @Builder.Default private double nu = 0.04;
+  @Builder.Default @Getter private double nu = 0.04;
   /**
    * The number of examples to use for computing the quantile for the r value update. This value
    * should generally be the same as the number of examples in the dataset
    */
-  @Builder.Default private int windowSize = 10000;
+  @Builder.Default @Getter private int windowSize = 10000;
   /**
    * The initial r value to use for ocnn for definition, see the paper, note this is only active
    * when {@link #configureR} is specified as true
    */
-  @Builder.Default private double initialRValue = 0.1;
+  @Builder.Default @Getter private double initialRValue = 0.1;
   /**
    * Whether to use the specified {@link #initialRValue} or use the weight initialization with the
    * neural network for the r value
    */
-  @Builder.Default private boolean configureR = true;
+  @Builder.Default @Getter private boolean configureR = true;
   /**
    * Psuedo code from keras: start_time = time.time() for epoch in range(100): # Train with each
    * example sess.run(updates, feed_dict={X: train_X,r:rvalue}) rvalue = nnScore(train_X, w_1, w_2,
    * g) with sess.as_default(): rvalue = rvalue.eval() rvalue = np.percentile(rvalue,q=100*nu)
    * print("Epoch = %d, r = %f" % (epoch + 1,rvalue))
    */
-  @Builder.Default private int lastEpochSinceRUpdated = 0;
+  @Builder.Default @Getter @Setter private int lastEpochSinceRUpdated = 0;
 
   @Override
   public Layer instantiate(

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/layers/ocnn/OCNNOutputLayer.java

@@ -21,8 +21,10 @@
 package org.deeplearning4j.nn.layers.ocnn;
 
 
+import lombok.Builder;
 import lombok.Getter;
 import lombok.Setter;
+import lombok.experimental.SuperBuilder;
 import org.deeplearning4j.nn.api.Layer;
 import org.deeplearning4j.nn.conf.layers.LayerConfiguration;
 import org.deeplearning4j.nn.gradient.DefaultGradient;
@@ -46,10 +48,12 @@ import static org.deeplearning4j.nn.layers.ocnn.OCNNParamInitializer.R_KEY;
 import static org.deeplearning4j.nn.layers.ocnn.OCNNParamInitializer.V_KEY;
 import static org.deeplearning4j.nn.layers.ocnn.OCNNParamInitializer.W_KEY;
 
+
 public class OCNNOutputLayer extends BaseOutputLayer<org.deeplearning4j.nn.conf.ocnn.OCNNOutputLayer> {
     @Setter
     @Getter
     private  IActivation activation = new ActivationReLU();
+
     private  static final IActivation relu = new ActivationReLU();
 
 

cavis-dnn/cavis-dnn-nn/src/main/java/org/deeplearning4j/nn/layers/util/IdentityLayer.java

@@ -21,16 +21,21 @@
 package org.deeplearning4j.nn.layers.util;
 
 import lombok.NoArgsConstructor;
+import lombok.experimental.SuperBuilder;
 import org.deeplearning4j.nn.conf.inputs.InputType;
 import org.deeplearning4j.nn.conf.layers.samediff.SameDiffLambdaLayer;
 import org.nd4j.autodiff.samediff.SDVariable;
 import org.nd4j.autodiff.samediff.SameDiff;
 
-@NoArgsConstructor
+@SuperBuilder(builderMethodName = "innerBuilder")
 public class IdentityLayer extends SameDiffLambdaLayer {
 
-    public IdentityLayer(String name) {
+    public static IdentityLayerBuilder<?,?> builder() {
-        this.name = name;
+        return innerBuilder();
+    }
+    public static IdentityLayerBuilder<?,?> builder(String name) {
+        return innerBuilder()
+                .name(name);
     }
 
     @Override

cavis-native/cavis-native-lib/build.gradle

@@ -235,7 +235,10 @@ chipList.each { thisChip ->
 
         /* Get VCVARS in case we want to build CUDA
         * MinGW64 g++ on MSYS is used otherwise */
-        if (thisChip.equals('cuda') && osdetector.os.startsWith("win") && !VISUAL_STUDIO_INSTALL_DIR.isEmpty()) {
+        if (thisChip.equals('cuda') && osdetector.os.startsWith("win")
+        && project.hasProperty("skip-native")
+                && !project.getProperty("skip-native").equals("true")
+                && !VISUAL_STUDIO_INSTALL_DIR.isEmpty()) {
             def proc = ["cmd.exe", "/c", "${VISUAL_STUDIO_VCVARS_CMD} > nul && set"].execute()
             it.environmentVariables = it.environmentVariables ?: [:]
             def lines = proc.text.split("\\r?\\n")
@@ -329,7 +332,8 @@ chipList.each { thisChip ->
 
         thisTask.properties = getBuildPlatform( thisChip, thisTask )
 
-        if(thisChip.equals('cuda') && osdetector.os.startsWith("win") && !VISUAL_STUDIO_INSTALL_DIR.isEmpty()) {
+        if(thisChip.equals('cuda') && osdetector.os.startsWith("win") && project.hasProperty("skip-native")
+                && !project.getProperty("skip-native").equals("true") && !VISUAL_STUDIO_INSTALL_DIR.isEmpty()) {
             def proc = ["cmd.exe", "/c", "${VISUAL_STUDIO_VCVARS_CMD} > nul && where.exe cl.exe"].execute()
             def outp = proc.text
             def cl = outp.replace("\\", "\\\\").trim()

vsconfig.gradle

@@ -28,7 +28,9 @@
  ****************************************************************************/
 
 if (!hasProperty("VISUAL_STUDIO_INSTALL_DIR") && osdetector.os.equals("windows")) {
+    if (project.hasProperty("skip-native") && !project.getProperty("skip-native").equals("true")) {
         configureVisualStudio()
+    }
 }
 
 def configureVisualStudio() {