diff --git a/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/gradientcheck/LossFunctionGradientCheck.java b/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/gradientcheck/LossFunctionGradientCheck.java
index 45e98246e449..99b568939166 100644
--- a/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/gradientcheck/LossFunctionGradientCheck.java
+++ b/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/gradientcheck/LossFunctionGradientCheck.java
@@ -51,6 +51,8 @@
 
 import static org.junit.Assert.assertEquals;
 import static org.junit.Assert.assertTrue;
+import static org.nd4j.linalg.indexing.NDArrayIndex.all;
+import static org.nd4j.linalg.indexing.NDArrayIndex.point;
 
 /**
  * Created by Alex on 12/09/2016.
@@ -411,7 +413,9 @@ public static INDArray[] getFeaturesAndLabels(ILossFunction l, long[] featuresSh
             case "LossKLD":
                 //KL divergence: should be a probability distribution for labels??
                 ret[1] = Nd4j.rand(labelsShape);
-                Nd4j.getExecutioner().exec(new OldSoftMax(ret[1]), 1);
+                for(int i=0; i<labelsShape[2]; i++ ) {
+                    Nd4j.getExecutioner().exec(new OldSoftMax(ret[1].get(all(), all(), point(i))), 1);
+                }
                 break;
             case "LossMCXENT":
             case "LossNegativeLogLikelihood":
diff --git a/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/nn/layers/objdetect/TestYolo2OutputLayer.java b/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/nn/layers/objdetect/TestYolo2OutputLayer.java
index 87493db568ab..47dbdecc7b60 100644
--- a/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/nn/layers/objdetect/TestYolo2OutputLayer.java
+++ b/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/nn/layers/objdetect/TestYolo2OutputLayer.java
@@ -495,7 +495,7 @@ public void testYoloOverfitting() throws Exception {
 
         MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
                 .convolutionMode(ConvolutionMode.Same)
-                .updater(new Adam(1e-3))
+                .updater(new Adam(2e-3))
                 .gradientNormalization(GradientNormalization.ClipElementWiseAbsoluteValue)
                 .gradientNormalizationThreshold(3)
                 .activation(Activation.LEAKYRELU)
diff --git a/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/optimize/solver/BackTrackLineSearchTest.java b/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/optimize/solver/BackTrackLineSearchTest.java
index 0be8c7e16dd9..cce562c32d3d 100644
--- a/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/optimize/solver/BackTrackLineSearchTest.java
+++ b/deeplearning4j/deeplearning4j-core/src/test/java/org/deeplearning4j/optimize/solver/BackTrackLineSearchTest.java
@@ -39,6 +39,7 @@
 import org.nd4j.linalg.dataset.DataSet;
 import org.nd4j.linalg.dataset.api.iterator.DataSetIterator;
 import org.nd4j.linalg.factory.Nd4j;
+import org.nd4j.linalg.learning.config.Adam;
 import org.nd4j.linalg.learning.config.Nesterovs;
 import org.nd4j.linalg.lossfunctions.LossFunctions;
 
@@ -240,7 +241,7 @@ public void testBackTrackLineLBFGS() {
 
     private static MultiLayerConfiguration getIrisMultiLayerConfig(Activation activationFunction, OptimizationAlgorithm optimizer) {
         MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder().optimizationAlgo(optimizer)
-                        .miniBatch(false).updater(new Nesterovs(0.9)).seed(12345L).list()
+                        .updater(new Adam(0.1)).seed(12345L).list()
                         .layer(0, new DenseLayer.Builder().nIn(4).nOut(100).weightInit(WeightInit.XAVIER)
                                         .activation(activationFunction).build())
                         .layer(1, new org.deeplearning4j.nn.conf.layers.OutputLayer.Builder(
diff --git a/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/recurrent/RnnLossLayer.java b/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/recurrent/RnnLossLayer.java
index 95c0394fabdd..ffe674a5f282 100644
--- a/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/recurrent/RnnLossLayer.java
+++ b/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/recurrent/RnnLossLayer.java
@@ -166,7 +166,9 @@ public INDArray activate(boolean training, LayerWorkspaceMgr workspaceMgr) {
             throw new UnsupportedOperationException(
                             "Input must be rank 3. Got input with rank " + input.rank() + " " + layerId());
 
-        return layerConf().getActivationFn().getActivation(workspaceMgr.dup(ArrayType.ACTIVATIONS, input), training);
+        INDArray as2d = TimeSeriesUtils.reshape3dTo2d(input);
+        INDArray out2d = layerConf().getActivationFn().getActivation(workspaceMgr.dup(ArrayType.ACTIVATIONS, as2d, as2d.ordering()), training);
+        return workspaceMgr.leverageTo(ArrayType.ACTIVATIONS, TimeSeriesUtils.reshape2dTo3d(out2d, (int)input.size(0), workspaceMgr, ArrayType.ACTIVATIONS));
     }
 
     @Override
diff --git a/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/variational/VariationalAutoencoder.java b/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/variational/VariationalAutoencoder.java
index b6f913b9a68f..298d5a3da2b1 100644
--- a/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/variational/VariationalAutoencoder.java
+++ b/deeplearning4j/deeplearning4j-nn/src/main/java/org/deeplearning4j/nn/layers/variational/VariationalAutoencoder.java
@@ -245,7 +245,7 @@ public void computeGradientAndScore(LayerWorkspaceMgr workspaceMgr) {
                 INDArray temp = meanZ.mul(meanZ).addi(pzxSigmaSquared).negi();
                 temp.addi(logStdev2Z).addi(1.0);
                 double scorePt1 = -0.5 / minibatch * temp.sumNumber().doubleValue();
-                this.score = scorePt1 + (calcL1(false) + calcL2(false)) / minibatch;
+                this.score = scorePt1 + (calcL1(false) + calcL2(false));
             }
 
             INDArray pxzDistributionPreOut = current.mmul(pxzw).addiRowVector(pxzb);