2019-06-06 15:21:15 +03:00

325 lines
12 KiB
Java

/*******************************************************************************
* Copyright (c) 2015-2018 Skymind, Inc.
*
* This program and the accompanying materials are made available under the
* terms of the Apache License, Version 2.0 which is available at
* https://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.
*
* SPDX-License-Identifier: Apache-2.0
******************************************************************************/
package org.deeplearning4j.optimizer.listener;
import lombok.Data;
import lombok.extern.slf4j.Slf4j;
import org.deeplearning4j.BaseDL4JTest;
import org.deeplearning4j.api.storage.StatsStorageRouter;
import org.deeplearning4j.api.storage.listener.RoutingIterationListener;
import org.deeplearning4j.datasets.iterator.impl.IrisDataSetIterator;
import org.deeplearning4j.nn.api.Layer;
import org.deeplearning4j.nn.api.Model;
import org.deeplearning4j.nn.conf.ComputationGraphConfiguration;
import org.deeplearning4j.nn.conf.MultiLayerConfiguration;
import org.deeplearning4j.nn.conf.NeuralNetConfiguration;
import org.deeplearning4j.nn.conf.layers.AutoEncoder;
import org.deeplearning4j.nn.conf.layers.OutputLayer;
import org.deeplearning4j.nn.conf.layers.variational.VariationalAutoencoder;
import org.deeplearning4j.nn.graph.ComputationGraph;
import org.deeplearning4j.nn.multilayer.MultiLayerNetwork;
import org.deeplearning4j.optimize.api.BaseTrainingListener;
import org.deeplearning4j.optimize.api.TrainingListener;
import org.deeplearning4j.optimize.listeners.ComposableIterationListener;
import org.deeplearning4j.optimize.listeners.PerformanceListener;
import org.deeplearning4j.optimize.listeners.ScoreIterationListener;
import org.deeplearning4j.optimize.listeners.TimeIterationListener;
import org.deeplearning4j.optimize.listeners.CheckpointListener;
import org.deeplearning4j.optimize.solvers.BaseOptimizer;
import org.junit.Rule;
import org.junit.Test;
import org.junit.rules.TemporaryFolder;
import org.nd4j.linalg.activations.Activation;
import org.nd4j.linalg.api.ndarray.INDArray;
import org.nd4j.linalg.dataset.api.iterator.DataSetIterator;
import org.nd4j.linalg.lossfunctions.LossFunctions;
import org.nd4j.linalg.primitives.Triple;
import java.io.ByteArrayInputStream;
import java.io.ByteArrayOutputStream;
import java.io.ObjectInputStream;
import java.io.ObjectOutputStream;
import java.util.ArrayList;
import java.util.Collection;
import java.util.List;
import java.util.Map;
import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertTrue;
/**
* Created by Alex on 01/01/2017.
*/
@Slf4j
public class TestListeners extends BaseDL4JTest {
@Rule
public TemporaryFolder tempDir = new TemporaryFolder();
@Test
public void testSettingListenersUnsupervised() {
//Pretrain layers should get copies of the listeners, in addition to the
MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder().list()
.layer(0, new AutoEncoder.Builder().nIn(10).nOut(10).build())
.layer(1, new VariationalAutoencoder.Builder().nIn(10).nOut(10).build()).build();
MultiLayerNetwork net = new MultiLayerNetwork(conf);
net.init();
net.setListeners(new ScoreIterationListener(), new TestRoutingListener());
for (Layer l : net.getLayers()) {
Collection<TrainingListener> layerListeners = l.getListeners();
assertEquals(l.getClass().toString(), 2, layerListeners.size());
TrainingListener[] lArr = layerListeners.toArray(new TrainingListener[2]);
assertTrue(lArr[0] instanceof ScoreIterationListener);
assertTrue(lArr[1] instanceof TestRoutingListener);
}
Collection<TrainingListener> netListeners = net.getListeners();
assertEquals(2, netListeners.size());
TrainingListener[] lArr = netListeners.toArray(new TrainingListener[2]);
assertTrue(lArr[0] instanceof ScoreIterationListener);
assertTrue(lArr[1] instanceof TestRoutingListener);
ComputationGraphConfiguration gConf = new NeuralNetConfiguration.Builder().graphBuilder().addInputs("in")
.addLayer("0", new AutoEncoder.Builder().nIn(10).nOut(10).build(), "in")
.addLayer("1", new VariationalAutoencoder.Builder().nIn(10).nOut(10).build(), "0")
.setOutputs("1").build();
ComputationGraph cg = new ComputationGraph(gConf);
cg.init();
cg.setListeners(new ScoreIterationListener(), new TestRoutingListener());
for (Layer l : cg.getLayers()) {
Collection<TrainingListener> layerListeners = l.getListeners();
assertEquals(2, layerListeners.size());
lArr = layerListeners.toArray(new TrainingListener[2]);
assertTrue(lArr[0] instanceof ScoreIterationListener);
assertTrue(lArr[1] instanceof TestRoutingListener);
}
netListeners = cg.getListeners();
assertEquals(2, netListeners.size());
lArr = netListeners.toArray(new TrainingListener[2]);
assertTrue(lArr[0] instanceof ScoreIterationListener);
assertTrue(lArr[1] instanceof TestRoutingListener);
}
private static class TestRoutingListener extends BaseTrainingListener implements RoutingIterationListener {
@Override
public void setStorageRouter(StatsStorageRouter router) {}
@Override
public StatsStorageRouter getStorageRouter() {
return null;
}
@Override
public void setWorkerID(String workerID) {}
@Override
public String getWorkerID() {
return null;
}
@Override
public void setSessionID(String sessionID) {}
@Override
public String getSessionID() {
return null;
}
@Override
public RoutingIterationListener clone() {
return null;
}
@Override
public void iterationDone(Model model, int iteration, int epoch) {}
}
@Test
public void testListenerSerialization() throws Exception {
//Note: not all listeners are (or should be) serializable. But some should be - for Spark etc
List<TrainingListener> listeners = new ArrayList<>();
listeners.add(new ScoreIterationListener());
listeners.add(new PerformanceListener(1, true, true));
listeners.add(new TimeIterationListener(10000));
listeners.add(new ComposableIterationListener(new ScoreIterationListener(), new PerformanceListener(1, true, true)));
listeners.add(new CheckpointListener.Builder(tempDir.newFolder()).keepAll().saveEveryNIterations(3).build()); //Doesn't usually need to be serialized, but no reason it can't be...
DataSetIterator iter = new IrisDataSetIterator(10, 150);
MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
.list()
.layer(new OutputLayer.Builder().nIn(4).nOut(3)
.activation(Activation.SOFTMAX)
.lossFunction(LossFunctions.LossFunction.MCXENT).build())
.build();
MultiLayerNetwork net = new MultiLayerNetwork(conf);
net.init();
net.setListeners(listeners);
net.fit(iter);
List<TrainingListener> listeners2 = new ArrayList<>();
for(TrainingListener il : listeners){
log.info("------------------");
log.info("Testing listener: {}", il);
ByteArrayOutputStream baos = new ByteArrayOutputStream();
ObjectOutputStream oos = new ObjectOutputStream(baos);
oos.writeObject(il);
byte[] bytes = baos.toByteArray();
ObjectInputStream ois = new ObjectInputStream(new ByteArrayInputStream(bytes));
TrainingListener il2 = (TrainingListener) ois.readObject();
listeners2.add(il2);
}
net.setListeners(listeners2);
net.fit(iter);
}
@Test
public void testListenerCalls(){
MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
.list()
.layer(new OutputLayer.Builder().nIn(4).nOut(3).activation(Activation.SOFTMAX).lossFunction(LossFunctions.LossFunction.MCXENT).build())
.build();
MultiLayerNetwork net = new MultiLayerNetwork(conf);
net.init();
TestListener tl = new TestListener();
net.setListeners(tl);
DataSetIterator irisIter = new IrisDataSetIterator(50, 150);
net.fit(irisIter, 2);
List<Triple<Call,Integer,Integer>> exp = new ArrayList<>();
exp.add(new Triple<>(Call.EPOCH_START, 0, 0));
exp.add(new Triple<>(Call.ON_FWD, 0, 0));
exp.add(new Triple<>(Call.ON_BWD, 0, 0));
exp.add(new Triple<>(Call.ON_GRAD, 0, 0));
exp.add(new Triple<>(Call.ITER_DONE, 0, 0));
exp.add(new Triple<>(Call.ON_FWD, 1, 0));
exp.add(new Triple<>(Call.ON_BWD, 1, 0));
exp.add(new Triple<>(Call.ON_GRAD, 1, 0));
exp.add(new Triple<>(Call.ITER_DONE, 1, 0));
exp.add(new Triple<>(Call.ON_FWD, 2, 0));
exp.add(new Triple<>(Call.ON_BWD, 2, 0));
exp.add(new Triple<>(Call.ON_GRAD, 2, 0));
exp.add(new Triple<>(Call.ITER_DONE, 2, 0));
exp.add(new Triple<>(Call.EPOCH_END, 3, 0)); //Post updating iter count, pre update epoch count
exp.add(new Triple<>(Call.EPOCH_START, 3, 1));
exp.add(new Triple<>(Call.ON_FWD, 3, 1));
exp.add(new Triple<>(Call.ON_BWD, 3, 1));
exp.add(new Triple<>(Call.ON_GRAD, 3, 1));
exp.add(new Triple<>(Call.ITER_DONE, 3, 1));
exp.add(new Triple<>(Call.ON_FWD, 4, 1));
exp.add(new Triple<>(Call.ON_BWD, 4, 1));
exp.add(new Triple<>(Call.ON_GRAD, 4, 1));
exp.add(new Triple<>(Call.ITER_DONE, 4, 1));
exp.add(new Triple<>(Call.ON_FWD, 5, 1));
exp.add(new Triple<>(Call.ON_BWD, 5, 1));
exp.add(new Triple<>(Call.ON_GRAD, 5, 1));
exp.add(new Triple<>(Call.ITER_DONE, 5, 1));
exp.add(new Triple<>(Call.EPOCH_END, 6, 1));
assertEquals(exp, tl.getCalls());
tl = new TestListener();
ComputationGraph cg = net.toComputationGraph();
cg.setListeners(tl);
cg.fit(irisIter, 2);
assertEquals(exp, tl.getCalls());
}
private static enum Call {
ITER_DONE,
EPOCH_START,
EPOCH_END,
ON_FWD,
ON_GRAD,
ON_BWD
}
@Data
private static class TestListener implements TrainingListener {
private List<Triple<Call,Integer,Integer>> calls = new ArrayList<>();
@Override
public void iterationDone(Model model, int iteration, int epoch) {
calls.add(new Triple<>(Call.ITER_DONE, iteration, epoch));
}
@Override
public void onEpochStart(Model model) {
calls.add(new Triple<>(Call.EPOCH_START, BaseOptimizer.getIterationCount(model), BaseOptimizer.getEpochCount(model)));
}
@Override
public void onEpochEnd(Model model) {
calls.add(new Triple<>(Call.EPOCH_END, BaseOptimizer.getIterationCount(model), BaseOptimizer.getEpochCount(model)));
}
@Override
public void onForwardPass(Model model, List<INDArray> activations) {
calls.add(new Triple<>(Call.ON_FWD, BaseOptimizer.getIterationCount(model), BaseOptimizer.getEpochCount(model)));
}
@Override
public void onForwardPass(Model model, Map<String, INDArray> activations) {
calls.add(new Triple<>(Call.ON_FWD, BaseOptimizer.getIterationCount(model), BaseOptimizer.getEpochCount(model)));
}
@Override
public void onGradientCalculation(Model model) {
calls.add(new Triple<>(Call.ON_GRAD, BaseOptimizer.getIterationCount(model), BaseOptimizer.getEpochCount(model)));
}
@Override
public void onBackwardPass(Model model) {
calls.add(new Triple<>(Call.ON_BWD, BaseOptimizer.getIterationCount(model), BaseOptimizer.getEpochCount(model)));
}
}
}