void train_network()
{
var mnist = new Datasets.MNIST();
var trainSteps = mnist.train_images.Length / batch_size;
var validationSteps = mnist.test_images.Length / batch_size;
var wrapped_trainGenerator = new InMemoryMiniBatchGenerator(
new float[][][] { mnist.train_images, mnist.train_labels },
new C.Variable[] { imageVariable, categoricalVariable },
batch_size, true, false, "training");
var trainGenerator = new MultiThreadedGenerator(4, wrapped_trainGenerator);
var wrapped_testGenerator = new InMemoryMiniBatchGenerator(
new float[][][] { mnist.test_images, mnist.test_labels },
new C.Variable[] { imageVariable, categoricalVariable },
batch_size, true, false, "testing");
var validationGenerator = new MultiThreadedGenerator(4, wrapped_testGenerator);
var epochs = 6;
Model.fit_generator(
network,
learner,
trainer,
evaluator,
batch_size,
epochs,
trainGenerator, trainSteps,
validationGenerator, validationSteps,
computeDevice, "mnist_");
trainGenerator.Dispose(); trainGenerator = null;
validationGenerator.Dispose(); validationGenerator = null;
}
void train_network()
{
var mnist = new Datasets.MNIST();
var trainSteps = mnist.train_images.Length / batch_size;
var validationSteps = mnist.test_images.Length / batch_size;
var trainGenerator = new InMemoryMiniBatchGenerator(
new float[][][] { mnist.train_images, mnist.train_labels },
new C.Variable[] { imageVariable, categoricalLabel },
batch_size, shuffle: true, only_once: false, name: "train");
var mtTrainGenerator = new MultiThreadedGenerator(workers: 4, generator: trainGenerator);
var validationGenerator = new InMemoryMiniBatchGenerator(
new float[][][] { mnist.test_images, mnist.test_labels },
new C.Variable[] { imageVariable, categoricalLabel },
batch_size, shuffle: false, only_once: false, name: "validation");
var mtValidationGenerator = new MultiThreadedGenerator(workers: 4, generator: validationGenerator);
var epochs = 6;
Model.fit_generator(
network,
learner,
trainer,
evaluator,
batch_size,
epochs,
mtTrainGenerator, trainSteps,
mtValidationGenerator, validationSteps,
computeDevice,
prefix: "capsules_",
trainingLossMetricName: "Training Loss",
trainingEvaluationMetricName: "Training Error",
validationMetricName: "Validation Error");
mtTrainGenerator.Dispose();
mtValidationGenerator.Dispose();
}
void train_network(Data train_data, Data test_data)
{
// For a discussion on the MNIST training/validation sets,
// see https://github.com/keras-team/keras/issues/1753
var trainSteps = train_data.numSteps(batch_size);
var validationSteps = test_data.numSteps(batch_size);
var inner_trainGenerator = new InMemoryMiniBatchGenerator(
train_data.All(), allInputVariables(),
batch_size, true, false, "training");
var trainGenerator = new MultiThreadedGenerator(4, inner_trainGenerator);
var inner_validationGenerator = new InMemoryMiniBatchGenerator(
test_data.All(), allInputVariables(),
batch_size, true, false, "validation");
var validationGenerator = new MultiThreadedGenerator(4, inner_validationGenerator);
Model.fit_generator(network, learner, trainer, evaluator, batch_size, epochs,
trainGenerator, trainSteps, validationGenerator, validationSteps, computeDevice, prefix: "ta_");
trainGenerator.Dispose(); trainGenerator = null;
validationGenerator.Dispose(); validationGenerator = null;
}
