Mateusz Bugaj :
I was want to load a training data set form file with RecodrReader and DataSetIterator but I get an error java.lang.ExceptionInInitializerError while trying and it doesn't tell me anything.
Here is main logic and where error is occuring:
public void trainNN() throws IOException, InterruptedException {
String setName = "trainingDataCSV";
createDataSet(TRAINING_SET_SIZE, setName);
DataSet allData;
RecordReader recordReader = new CSVRecordReader(0, ',');
recordReader.initialize(new FileSplit(new File(setName))); // error here
int numPixels = IMAGE_SIZE * IMAGE_SIZE;
DataSetIterator iterator = new RecordReaderDataSetIterator(recordReader, TRAINING_SET_SIZE, numPixels, numPixels + 4);
allData = iterator.next();
DataNormalization normalizer = new NormalizerStandardize();
normalizer.fit(allData);
normalizer.transform(allData);
SplitTestAndTrain testAndTrain = allData.splitTestAndTrain(0.65);
DataSet trainingData = testAndTrain.getTrain();
DataSet testData = testAndTrain.getTest();
MultiLayerConfiguration configuration = new NeuralNetConfiguration.Builder()
.activation(Activation.TANH)
.weightInit(WeightInit.XAVIER)
// .learningRate(0.1)
.l2(0.0001)
.list()
.layer(0, new DenseLayer.Builder().nIn(numPixels).nOut(50).build())
.layer(1, new DenseLayer.Builder().nIn(50).nOut(50).build())
.layer(2, new OutputLayer.Builder(
LossFunctions.LossFunction.NEGATIVELOGLIKELIHOOD)
.activation(Activation.SOFTMAX)
.nIn(50).nOut(4).build())
.build();
MultiLayerNetwork neuralNetwork = new MultiLayerNetwork(configuration);
INDArray output = neuralNetwork.output(testData.getFeatures());
Evaluation eval = new Evaluation(3);
eval.eval(testData.getLabels(), output);
System.out.println(eval.stats());
}
Full error stack trace
java.lang.ExceptionInInitializerError
at org.datavec.api.util.ndarray.RecordConverter.toMinibatchArray(RecordConverter.java:198)
at org.deeplearning4j.datasets.datavec.RecordReaderMultiDataSetIterator.next(RecordReaderMultiDataSetIterator.java:160)
at org.deeplearning4j.datasets.datavec.RecordReaderDataSetIterator.next(RecordReaderDataSetIterator.java:377)
at org.deeplearning4j.datasets.datavec.RecordReaderDataSetIterator.next(RecordReaderDataSetIterator.java:452)
at org.deeplearning4j.datasets.datavec.RecordReaderDataSetIterator.next(RecordReaderDataSetIterator.java:85)
at App.trainNN(App.java:116)
at App.setup(App.java:59)
at processing.core.PApplet.handleDraw(PApplet.java:2412)
at processing.awt.PSurfaceAWT$12.callDraw(PSurfaceAWT.java:1557)
at processing.core.PSurfaceNone$AnimationThread.run(PSurfaceNone.java:316)
Caused by: java.lang.RuntimeException: org.nd4j.linalg.factory.Nd4jBackend$NoAvailableBackendException: Please ensure that you have an nd4j backend on your classpath. Please see: http://nd4j.org/getstarted.html
at org.nd4j.linalg.factory.Nd4j.initContext(Nd4j.java:5131)
at org.nd4j.linalg.factory.Nd4j.<clinit>(Nd4j.java:226)
... 10 more
Caused by: org.nd4j.linalg.factory.Nd4jBackend$NoAvailableBackendException: Please ensure that you have an nd4j backend on your classpath. Please see: http://nd4j.org/getstarted.html
at org.nd4j.linalg.factory.Nd4jBackend.load(Nd4jBackend.java:218)
at org.nd4j.linalg.factory.Nd4j.initContext(Nd4j.java:5128)
... 11 more
And here is how data file is created just in case.
It's 400 values containing 1.0 or 0.0 separated with semicolons and 4 additional values for position and dimentions of the shape on image
public void createDataSet(int amount, String name) throws IOException {
BufferedWriter writer = new BufferedWriter(new FileWriter(name));
writer.write("");
for (int i = 0; i < amount; i++) {
Image image = new Image(IMAGE_SIZE, MIN_OBJECT_SIZE, MAX_OBJECT_SIZE, OBJECTS_AMOUNT);
int positionX = image.getPosition()[0];
int positionY = image.getPosition()[1];
int w = image.getW();
int h = image.getH();
for (int pixel = 0; pixel < image.getPixelValues().length; pixel++) {
writer.append(image.getPixelValues()[pixel] + ",");
}
writer.append(positionX + ",");
writer.append(positionY + ",");
writer.append(w + ",");
writer.append(String.valueOf(h) + '\n');
}
writer.close();
}
Paul Dubs :
The important part of the exception is this:
Caused by: org.nd4j.linalg.factory.Nd4jBackend$NoAvailableBackendException: Please ensure that you have an nd4j backend on your classpath. Please see: http://nd4j.org/getstarted.html
And you will likely get the same error with just something that initializes ND4J (like Nd4j.zeros(1)).
Make sure that you've got a proper ND4J backend in your dependencies, like shown in the documentation: https://deeplearning4j.konduit.ai/config/maven#add-a-backend
Guess you like
Origin http://43.154.161.224:23101/article/api/json?id=352873&siteId=1