Update shaded Jackson version to 2.10.1 (#82)
* Update shaded Jackson version to 2.10.1 Signed-off-by: AlexDBlack <blacka101@gmail.com> * Remove no longer needed scala compiler plugin from UI Signed-off-by: AlexDBlack <blacka101@gmail.com> * Fix op name for BitwiseAnd op Signed-off-by: AlexDBlack <blacka101@gmail.com> * TimeDistributedLayer mask array fix + test Signed-off-by: AlexDBlack <blacka101@gmail.com>master
Alex Black
GitHub
parent
4b50b920c7
commit
8843c7377a
|
|
@ -21,12 +21,14 @@ import org.deeplearning4j.TestUtils;
|
|||
import org.deeplearning4j.nn.conf.MultiLayerConfiguration;
|
||||
import org.deeplearning4j.nn.conf.NeuralNetConfiguration;
|
||||
import org.deeplearning4j.nn.conf.inputs.InputType;
|
||||
import org.deeplearning4j.nn.conf.layers.DenseLayer;
|
||||
import org.deeplearning4j.nn.conf.layers.LSTM;
|
||||
import org.deeplearning4j.nn.conf.layers.OutputLayer;
|
||||
import org.deeplearning4j.nn.conf.layers.RnnOutputLayer;
|
||||
import org.deeplearning4j.nn.conf.layers.recurrent.Bidirectional;
|
||||
import org.deeplearning4j.nn.conf.layers.recurrent.LastTimeStep;
|
||||
import org.deeplearning4j.nn.conf.layers.recurrent.SimpleRnn;
|
||||
import org.deeplearning4j.nn.conf.layers.recurrent.TimeDistributed;
|
||||
import org.deeplearning4j.nn.multilayer.MultiLayerNetwork;
|
||||
import org.deeplearning4j.nn.weights.WeightInit;
|
||||
import org.junit.Ignore;
|
||||
|
|
@ -289,4 +291,66 @@ public class RnnGradientChecks extends BaseDL4JTest {
|
|||
}
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
|
||||
|
||||
@Test
|
||||
public void testTimeDistributedDense() {
|
||||
int nIn = 3;
|
||||
int nOut = 5;
|
||||
int tsLength = 4;
|
||||
int layerSize = 8;
|
||||
|
||||
Random r = new Random(12345);
|
||||
for (int mb : new int[]{1, 3}) {
|
||||
for (boolean inputMask : new boolean[]{false, true}) {
|
||||
|
||||
|
||||
INDArray in = Nd4j.rand(new int[]{mb, nIn, tsLength});
|
||||
INDArray labels = TestUtils.randomOneHotTimeSeries(mb, nOut, tsLength);
|
||||
String maskType = (inputMask ? "inputMask" : "none");
|
||||
|
||||
INDArray inMask = null;
|
||||
if (inputMask) {
|
||||
inMask = Nd4j.ones(mb, tsLength);
|
||||
for (int i = 0; i < mb; i++) {
|
||||
int firstMaskedStep = tsLength - 1 - i;
|
||||
if (firstMaskedStep == 0) {
|
||||
firstMaskedStep = tsLength;
|
||||
}
|
||||
for (int j = firstMaskedStep; j < tsLength; j++) {
|
||||
inMask.putScalar(i, j, 0.0);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
String name = "testLastTimeStepLayer() - mb=" + mb + ", tsLength = " + tsLength + ", maskType=" + maskType;
|
||||
if (PRINT_RESULTS) {
|
||||
System.out.println("Starting test: " + name);
|
||||
}
|
||||
|
||||
MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
|
||||
.dataType(DataType.DOUBLE)
|
||||
.activation(Activation.TANH)
|
||||
.updater(new NoOp())
|
||||
.weightInit(WeightInit.XAVIER)
|
||||
.list()
|
||||
.layer(new LSTM.Builder().nOut(layerSize).build())
|
||||
.layer(new TimeDistributed(new DenseLayer.Builder().nOut(layerSize).activation(Activation.SOFTMAX).build(), 2))
|
||||
.layer(new RnnOutputLayer.Builder().nOut(nOut).activation(Activation.SOFTMAX)
|
||||
.lossFunction(LossFunctions.LossFunction.MCXENT).build())
|
||||
.setInputType(InputType.recurrent(nIn))
|
||||
.build();
|
||||
|
||||
MultiLayerNetwork net = new MultiLayerNetwork(conf);
|
||||
net.init();
|
||||
|
||||
boolean gradOK = GradientCheckUtil.checkGradients(net, DEFAULT_EPS, DEFAULT_MAX_REL_ERROR,
|
||||
DEFAULT_MIN_ABS_ERROR, PRINT_RESULTS, RETURN_ON_FIRST_FAILURE, in, labels, inMask, null, true, 16);
|
||||
assertTrue(name, gradOK);
|
||||
TestUtils.testModelSerialization(net);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -16,6 +16,7 @@
|
|||
|
||||
package org.deeplearning4j.nn.dtypes;
|
||||
|
||||
import org.deeplearning4j.nn.conf.layers.recurrent.TimeDistributed;
|
||||
import org.nd4j.shade.guava.collect.ImmutableSet;
|
||||
import org.nd4j.shade.guava.reflect.ClassPath;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
|
|
@ -811,7 +812,8 @@ public class DTypeTests extends BaseDL4JTest {
|
|||
.layer(new DenseLayer.Builder().nOut(5).build())
|
||||
.layer(new GravesBidirectionalLSTM.Builder().nIn(5).nOut(5).activation(Activation.TANH).build())
|
||||
.layer(new Bidirectional(new LSTM.Builder().nIn(5).nOut(5).activation(Activation.TANH).build()))
|
||||
.layer(new SimpleRnn.Builder().nIn(10).nOut(5).build())
|
||||
.layer(new TimeDistributed(new DenseLayer.Builder().nIn(10).nOut(5).activation(Activation.TANH).build(), 2))
|
||||
.layer(new SimpleRnn.Builder().nIn(5).nOut(5).build())
|
||||
.layer(new MaskZeroLayer.Builder().underlying(new SimpleRnn.Builder().nIn(5).nOut(5).build()).maskValue(0.0).build())
|
||||
.layer(secondLast)
|
||||
.layer(ol)
|
||||
|
|
|
|||
|
|
@ -1,10 +1,12 @@
|
|||
package org.deeplearning4j.nn.layers.recurrent;
|
||||
|
||||
import org.deeplearning4j.nn.api.Layer;
|
||||
import org.deeplearning4j.nn.api.MaskState;
|
||||
import org.deeplearning4j.nn.gradient.Gradient;
|
||||
import org.deeplearning4j.nn.layers.wrapper.BaseWrapperLayer;
|
||||
import org.deeplearning4j.nn.workspace.ArrayType;
|
||||
import org.deeplearning4j.nn.workspace.LayerWorkspaceMgr;
|
||||
import org.deeplearning4j.util.TimeSeriesUtils;
|
||||
import org.nd4j.linalg.api.ndarray.INDArray;
|
||||
import org.nd4j.linalg.primitives.Pair;
|
||||
import org.nd4j.linalg.util.ArrayUtil;
|
||||
|
|
@ -107,4 +109,30 @@ public class TimeDistributedLayer extends BaseWrapperLayer {
|
|||
INDArray permuted = reshaped.permute(permute);
|
||||
return permuted;
|
||||
}
|
||||
|
||||
@Override
|
||||
public void setMaskArray(INDArray maskArray) {
|
||||
if(maskArray == null){
|
||||
underlying.setMaskArray(null);
|
||||
} else {
|
||||
INDArray reshaped = TimeSeriesUtils.reshapeTimeSeriesMaskToVector(maskArray, LayerWorkspaceMgr.noWorkspaces(), ArrayType.ACTIVATIONS);
|
||||
underlying.setMaskArray(reshaped);
|
||||
}
|
||||
}
|
||||
|
||||
@Override
|
||||
public Pair<INDArray, MaskState> feedForwardMaskArray(INDArray maskArray, MaskState currentMaskState, int minibatchSize) {
|
||||
if(maskArray == null){
|
||||
return underlying.feedForwardMaskArray(null, currentMaskState, minibatchSize);
|
||||
} else {
|
||||
INDArray reshaped = TimeSeriesUtils.reshapeTimeSeriesMaskToVector(maskArray, LayerWorkspaceMgr.noWorkspaces(), ArrayType.ACTIVATIONS);
|
||||
Pair<INDArray, MaskState> p = underlying.feedForwardMaskArray(reshaped, currentMaskState, minibatchSize);
|
||||
if(p == null || p.getFirst() == null){
|
||||
return p;
|
||||
}
|
||||
INDArray reshaped2 = TimeSeriesUtils.reshapeVectorToTimeSeriesMask(p.getFirst(), (int)maskArray.size(0));
|
||||
p.setFirst(reshaped2);
|
||||
return p;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -34,36 +34,6 @@
|
|||
<groupId>org.apache.maven.plugins</groupId>
|
||||
<artifactId>maven-compiler-plugin</artifactId>
|
||||
</plugin>
|
||||
<plugin>
|
||||
<groupId>net.alchim31.maven</groupId>
|
||||
<artifactId>scala-maven-plugin</artifactId>
|
||||
<version>${maven-scala-plugin.version}</version>
|
||||
<configuration>
|
||||
<args>
|
||||
<arg>-deprecation</arg>
|
||||
<arg>-explaintypes</arg>
|
||||
<arg>-nobootcp</arg>
|
||||
</args>
|
||||
</configuration>
|
||||
<executions>
|
||||
<execution>
|
||||
<id>scala-compile-first</id>
|
||||
<phase>process-resources</phase>
|
||||
<goals>
|
||||
<goal>add-source</goal>
|
||||
<goal>compile</goal>
|
||||
</goals>
|
||||
</execution>
|
||||
<execution>
|
||||
<id>scala-test-compile</id>
|
||||
<phase>process-test-resources</phase>
|
||||
<goals>
|
||||
<goal>add-source</goal>
|
||||
<goal>testCompile</goal>
|
||||
</goals>
|
||||
</execution>
|
||||
</executions>
|
||||
</plugin>
|
||||
</plugins>
|
||||
</build>
|
||||
<modules>
|
||||
|
|
|
|||
|
|
@ -1,5 +1,6 @@
|
|||
/* ******************************************************************************
|
||||
* Copyright (c) 2015-2018 Skymind, Inc.
|
||||
* Copyright (c) 2019 Konduit K.K.
|
||||
*
|
||||
* This program and the accompanying materials are made available under the
|
||||
* terms of the Apache License, Version 2.0 which is available at
|
||||
|
|
@ -50,7 +51,7 @@ public class BitwiseAnd extends BaseDynamicTransformOp {
|
|||
|
||||
@Override
|
||||
public String opName() {
|
||||
return "BitwiseAnd";
|
||||
return "bitwise_and";
|
||||
}
|
||||
|
||||
|
||||
|
|
|
|||
6
pom.xml
6
pom.xml
|
|
@ -328,9 +328,9 @@
|
|||
<slf4j.version>1.7.21</slf4j.version>
|
||||
<junit.version>4.12</junit.version>
|
||||
<logback.version>1.2.3</logback.version>
|
||||
<jackson.version>2.9.9</jackson.version>
|
||||
<jackson.databind.version>2.9.9.3</jackson.databind.version>
|
||||
<shaded.snakeyaml.version>1.23</shaded.snakeyaml.version>
|
||||
<jackson.version>2.10.1</jackson.version>
|
||||
<jackson.databind.version>2.10.1</jackson.databind.version>
|
||||
<shaded.snakeyaml.version>1.24</shaded.snakeyaml.version>
|
||||
<geo.jackson.version>2.8.7</geo.jackson.version>
|
||||
<lombok.version>1.18.2</lombok.version>
|
||||
<cleartk.version>2.0.0</cleartk.version>
|
||||
|
|
|
|||
Loading…
Reference in New Issue