fix tests

KraFusion · Sep 8, 2017 · b515399 · b515399
1 parent a3507a0
commit b515399
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 7 deletions.
diff --git a/...rning4j-core/src/test/java/org/deeplearning4j/nn/layers/convolution/Upsampling2DTest.java b/...rning4j-core/src/test/java/org/deeplearning4j/nn/layers/convolution/Upsampling2DTest.java
@@ -48,8 +48,6 @@ public void testUpsampling() throws Exception {
         double[] outArray = new double[] {1., 1., 2., 2., 1., 1., 2., 2., 3., 3., 4., 4., 3., 3., 4., 4.};
         INDArray containedExpectedOut = Nd4j.create(outArray, new int[] {1, 1, 4, 4});
         INDArray containedInput = getContainedData();
-        System.out.println(containedInput);
-        System.out.println(containedExpectedOut);
         INDArray input = getData();
         Layer layer = getUpsamplingLayer();
 
@@ -70,7 +68,7 @@ public void testUpsampling2DBackprop() throws Exception {
                         Nd4j.create(new double[] {1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.},
                                 new int[] {1, 1, 4, 4});
 
-        INDArray expectedContainedEpsilonResult = Nd4j.create(new double[] {1., 1., 1., 1.},
+        INDArray expectedContainedEpsilonResult = Nd4j.create(new double[] {4., 4., 4., 4.},
                         new int[] {1, 1, 2, 2});
 
         INDArray input = getContainedData();
@@ -79,6 +77,7 @@ public void testUpsampling2DBackprop() throws Exception {
         layer.activate(input);
 
         Pair<Gradient, INDArray> containedOutput = layer.backpropGradient(expectedContainedEpsilonInput);
+
         assertEquals(expectedContainedEpsilonResult, containedOutput.getSecond());
         assertEquals(null, containedOutput.getFirst().getGradientFor("W"));
         assertEquals(expectedContainedEpsilonResult.shape().length, containedOutput.getSecond().shape().length);

diff --git a/...4j-nn/src/main/java/org/deeplearning4j/nn/layers/convolution/upsampling/Upsampling2D.java b/...4j-nn/src/main/java/org/deeplearning4j/nn/layers/convolution/upsampling/Upsampling2D.java
@@ -86,19 +86,20 @@ public Pair<Gradient, INDArray> backpropGradient(INDArray epsilon) {
         int size = layerConf().getSize();
 
         INDArray outEpsilon = Nd4j.createUninitialized(miniBatch * inDepth * inH * inW);
+        INDArray reshapedEpsilon = outEpsilon.reshape('c', miniBatch, inDepth, inH, inW);
+
         INDArray forwardOutput  = preOutput(true, true);
 
         Gradient gradient = new DefaultGradient();
 
         CustomOp op = DynamicCustomOp.builder("upsampling_bp")
                 .setIntegerArguments(size)
                 .setInputs(forwardOutput, epsilon)
-                .setOutputs(outEpsilon)
+                .setOutputs(reshapedEpsilon)
                 .callInplace(false)
                 .build();
         Nd4j.getExecutioner().exec(op);
 
-        INDArray reshapedEpsilon = outEpsilon.reshape('c', miniBatch, inDepth, inH, inW);
 
         return new Pair<>(gradient, reshapedEpsilon);
     }
@@ -135,17 +136,18 @@ public INDArray preOutput(boolean training, boolean forBackprop) {
         int outW = inW * size;
 
         INDArray output = Nd4j.createUninitialized(miniBatch * inDepth * outH * outW);
+        INDArray reshapedOutput = output.reshape('c', miniBatch, inDepth, outH, outW);
 
         CustomOp op = DynamicCustomOp.builder("upsampling")
                 .setIntegerArguments(size)
                 .setInputs(input)
-                .setOutputs(output)
+                .setOutputs(reshapedOutput)
                 .callInplace(false)
                 .build();
 
         Nd4j.getExecutioner().exec(op);
 
-        return output.reshape('c', miniBatch, inDepth, outH, outW);
+        return reshapedOutput;
     }
 
     @Override