Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Add Spark VAE reconstruction error (for VAE + loss function models)
- Loading branch information
1 parent
e684335
commit c12e989
Showing
7 changed files
with
284 additions
and
98 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
116 changes: 116 additions & 0 deletions
116
...src/main/java/org/deeplearning4j/spark/impl/common/score/BaseVaeScoreWithKeyFunction.java
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,116 @@ | ||
| /* | ||
| * | ||
| * * Copyright 2016 Skymind,Inc. | ||
| * * | ||
| * * Licensed under the Apache License, Version 2.0 (the "License"); | ||
| * * you may not use this file except in compliance with the License. | ||
| * * You may obtain a copy of the License at | ||
| * * | ||
| * * http://www.apache.org/licenses/LICENSE-2.0 | ||
| * * | ||
| * * Unless required by applicable law or agreed to in writing, software | ||
| * * distributed under the License is distributed on an "AS IS" BASIS, | ||
| * * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| * * See the License for the specific language governing permissions and | ||
| * * limitations under the License. | ||
| * | ||
| */ | ||
|
|
||
| package org.deeplearning4j.spark.impl.common.score; | ||
|
|
||
| import lombok.extern.slf4j.Slf4j; | ||
| import org.apache.spark.api.java.function.PairFlatMapFunction; | ||
| import org.apache.spark.broadcast.Broadcast; | ||
| import org.deeplearning4j.nn.layers.variational.VariationalAutoencoder; | ||
| import org.nd4j.linalg.api.ndarray.INDArray; | ||
| import org.nd4j.linalg.api.ops.executioner.GridExecutioner; | ||
| import org.nd4j.linalg.factory.Nd4j; | ||
| import org.slf4j.Logger; | ||
| import org.slf4j.LoggerFactory; | ||
| import scala.Tuple2; | ||
|
|
||
| import java.util.ArrayList; | ||
| import java.util.Collections; | ||
| import java.util.Iterator; | ||
| import java.util.List; | ||
|
|
||
| /** | ||
| * Function to calculate the scores (reconstruction probability, reconstruction error) for a variational autoencoder.<br> | ||
| * Note that scoring is batched for computational efficiency.<br> | ||
| * | ||
| * @param <K> Type of key, associated with each example. Used to keep track of which score belongs to which example | ||
| * @author Alex Black | ||
| */ | ||
| @Slf4j | ||
| public abstract class BaseVaeScoreWithKeyFunction<K> implements PairFlatMapFunction<Iterator<Tuple2<K, INDArray>>, K, Double> { | ||
|
|
||
| protected final Broadcast<INDArray> params; | ||
| protected final Broadcast<String> jsonConfig; | ||
| private final int batchSize; | ||
|
|
||
|
|
||
| /** | ||
| * @param params MultiLayerNetwork parameters | ||
| * @param jsonConfig MultiLayerConfiguration, as json | ||
| * @param batchSize Batch size to use when scoring | ||
| */ | ||
| public BaseVaeScoreWithKeyFunction(Broadcast<INDArray> params, Broadcast<String> jsonConfig, int batchSize) { | ||
| this.params = params; | ||
| this.jsonConfig = jsonConfig; | ||
| this.batchSize = batchSize; | ||
| } | ||
|
|
||
| public abstract VariationalAutoencoder getVaeLayer(); | ||
|
|
||
| public abstract INDArray computeScore(VariationalAutoencoder vae, INDArray toScore); | ||
|
|
||
|
|
||
| @Override | ||
| public Iterable<Tuple2<K, Double>> call(Iterator<Tuple2<K, INDArray>> iterator) throws Exception { | ||
| if (!iterator.hasNext()) { | ||
| return Collections.emptyList(); | ||
| } | ||
|
|
||
| VariationalAutoencoder vae = getVaeLayer(); | ||
|
|
||
| List<Tuple2<K, Double>> ret = new ArrayList<>(); | ||
|
|
||
| List<INDArray> collect = new ArrayList<>(batchSize); | ||
| List<K> collectKey = new ArrayList<>(batchSize); | ||
| int totalCount = 0; | ||
| while (iterator.hasNext()) { | ||
| collect.clear(); | ||
| collectKey.clear(); | ||
| int nExamples = 0; | ||
| while (iterator.hasNext() && nExamples < batchSize) { | ||
| Tuple2<K, INDArray> t2 = iterator.next(); | ||
| INDArray features = t2._2(); | ||
| int n = features.size(0); | ||
| if (n != 1) throw new IllegalStateException("Cannot score examples with one key per data set if " | ||
| + "data set contains more than 1 example (numExamples: " + n + ")"); | ||
| collect.add(features); | ||
| collectKey.add(t2._1()); | ||
| nExamples += n; | ||
| } | ||
| totalCount += nExamples; | ||
|
|
||
| INDArray toScore = Nd4j.vstack(collect); | ||
| INDArray scores = computeScore(vae, toScore); | ||
|
|
||
| double[] doubleScores = scores.data().asDouble(); | ||
|
|
||
| for (int i = 0; i < doubleScores.length; i++) { | ||
| ret.add(new Tuple2<>(collectKey.get(i), doubleScores[i])); | ||
| } | ||
| } | ||
|
|
||
| if (Nd4j.getExecutioner() instanceof GridExecutioner) | ||
| ((GridExecutioner) Nd4j.getExecutioner()).flushQueueBlocking(); | ||
|
|
||
| if (log.isDebugEnabled()) { | ||
| log.debug("Scored {} examples ", totalCount); | ||
| } | ||
|
|
||
| return ret; | ||
| } | ||
| } |
54 changes: 54 additions & 0 deletions
54
.../org/deeplearning4j/spark/impl/graph/scoring/CGVaeReconstructionErrorWithKeyFunction.java
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,54 @@ | ||
| package org.deeplearning4j.spark.impl.graph.scoring; | ||
|
|
||
| import org.apache.spark.broadcast.Broadcast; | ||
| import org.deeplearning4j.nn.api.Layer; | ||
| import org.deeplearning4j.nn.conf.ComputationGraphConfiguration; | ||
| import org.deeplearning4j.nn.graph.ComputationGraph; | ||
| import org.deeplearning4j.nn.layers.variational.VariationalAutoencoder; | ||
| import org.deeplearning4j.spark.impl.common.score.BaseVaeReconstructionProbWithKeyFunction; | ||
| import org.deeplearning4j.spark.impl.common.score.BaseVaeScoreWithKeyFunction; | ||
| import org.nd4j.linalg.api.ndarray.INDArray; | ||
|
|
||
| /** | ||
| * Function to calculate the reconstruction error for a variational autoencoder, that is the first layer in a | ||
| * ComputationGraph.<br> | ||
| * Note that the VAE must be using a loss function, not a {@link org.deeplearning4j.nn.conf.layers.variational.ReconstructionDistribution}<br> | ||
| * Also note that scoring is batched for computational efficiency.<br> | ||
| * | ||
| * @author Alex Black | ||
| * @see CGVaeReconstructionProbWithKeyFunction | ||
| */ | ||
| public class CGVaeReconstructionErrorWithKeyFunction<K> extends BaseVaeScoreWithKeyFunction<K> { | ||
|
|
||
|
|
||
| /** | ||
| * @param params MultiLayerNetwork parameters | ||
| * @param jsonConfig MultiLayerConfiguration, as json | ||
| * @param batchSize Batch size to use when scoring | ||
| */ | ||
| public CGVaeReconstructionErrorWithKeyFunction(Broadcast<INDArray> params, Broadcast<String> jsonConfig, int batchSize) { | ||
| super(params, jsonConfig, batchSize); | ||
| } | ||
|
|
||
| @Override | ||
| public VariationalAutoencoder getVaeLayer() { | ||
| ComputationGraph network = new ComputationGraph(ComputationGraphConfiguration.fromJson((String)jsonConfig.getValue())); | ||
| network.init(); | ||
| INDArray val = ((INDArray)params.value()).unsafeDuplication(); | ||
| if (val.length() != network.numParams(false)) | ||
| throw new IllegalStateException("Network did not have same number of parameters as the broadcasted set parameters"); | ||
| network.setParams(val); | ||
|
|
||
| Layer l = network.getLayer(0); | ||
| if (!(l instanceof VariationalAutoencoder)) { | ||
| throw new RuntimeException("Cannot use CGVaeReconstructionErrorWithKeyFunction on network that doesn't have a VAE " | ||
| + "layer as layer 0. Layer type: " + l.getClass()); | ||
| } | ||
| return (VariationalAutoencoder)l; | ||
| } | ||
|
|
||
| @Override | ||
| public INDArray computeScore(VariationalAutoencoder vae, INDArray toScore) { | ||
| return vae.reconstructionError(toScore); | ||
| } | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
53 changes: 53 additions & 0 deletions
53
...g/deeplearning4j/spark/impl/multilayer/scoring/VaeReconstructionErrorWithKeyFunction.java
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,53 @@ | ||
| package org.deeplearning4j.spark.impl.multilayer.scoring; | ||
|
|
||
| import org.apache.spark.broadcast.Broadcast; | ||
| import org.deeplearning4j.nn.api.Layer; | ||
| import org.deeplearning4j.nn.conf.MultiLayerConfiguration; | ||
| import org.deeplearning4j.nn.layers.variational.VariationalAutoencoder; | ||
| import org.deeplearning4j.nn.multilayer.MultiLayerNetwork; | ||
| import org.deeplearning4j.spark.impl.common.score.BaseVaeReconstructionProbWithKeyFunction; | ||
| import org.deeplearning4j.spark.impl.common.score.BaseVaeScoreWithKeyFunction; | ||
| import org.nd4j.linalg.api.ndarray.INDArray; | ||
|
|
||
| /** | ||
| * Function to calculate the reconstruction error for a variational autoencoder, that is the first layer in a | ||
| * MultiLayerNetwork.<br> | ||
| * Note that the VAE must be using a loss function, not a {@link org.deeplearning4j.nn.conf.layers.variational.ReconstructionDistribution}<br> | ||
| * Also note that scoring is batched for computational efficiency.<br> | ||
| * | ||
| * @author Alex Black | ||
| * @see VaeReconstructionProbWithKeyFunction | ||
| */ | ||
| public class VaeReconstructionErrorWithKeyFunction<K> extends BaseVaeScoreWithKeyFunction<K> { | ||
|
|
||
| /** | ||
| * @param params MultiLayerNetwork parameters | ||
| * @param jsonConfig MultiLayerConfiguration, as json | ||
| * @param batchSize Batch size to use when scoring | ||
| */ | ||
| public VaeReconstructionErrorWithKeyFunction(Broadcast<INDArray> params, Broadcast<String> jsonConfig, int batchSize) { | ||
| super(params, jsonConfig, batchSize); | ||
| } | ||
|
|
||
| @Override | ||
| public VariationalAutoencoder getVaeLayer() { | ||
| MultiLayerNetwork network = new MultiLayerNetwork(MultiLayerConfiguration.fromJson((String)jsonConfig.getValue())); | ||
| network.init(); | ||
| INDArray val = ((INDArray)params.value()).unsafeDuplication(); | ||
| if (val.length() != network.numParams(false)) | ||
| throw new IllegalStateException("Network did not have same number of parameters as the broadcasted set parameters"); | ||
| network.setParameters(val); | ||
|
|
||
| Layer l = network.getLayer(0); | ||
| if (!(l instanceof VariationalAutoencoder)) { | ||
| throw new RuntimeException("Cannot use VaeReconstructionErrorWithKeyFunction on network that doesn't have a VAE " | ||
| + "layer as layer 0. Layer type: " + l.getClass()); | ||
| } | ||
| return (VariationalAutoencoder)l; | ||
| } | ||
|
|
||
| @Override | ||
| public INDArray computeScore(VariationalAutoencoder vae, INDArray toScore) { | ||
| return vae.reconstructionError(toScore); | ||
| } | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.