org.deeplearning4j.earlystopping.trainer.EarlyStoppingTrainer Maven / Gradle / Ivy

Show more of this group Show more artifacts with this name
Show all versions of deeplearning4j-nn Show documentation
There is a newer version: 1.0.0-M2.1
/*******************************************************************************
 * 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.earlystopping.trainer;

import org.deeplearning4j.datasets.iterator.MultiDataSetWrapperIterator;
import org.deeplearning4j.datasets.iterator.impl.SingletonDataSetIterator;
import org.deeplearning4j.datasets.iterator.impl.SingletonMultiDataSetIterator;
import org.deeplearning4j.earlystopping.EarlyStoppingConfiguration;
import org.deeplearning4j.earlystopping.listener.EarlyStoppingListener;
import org.deeplearning4j.nn.conf.MultiLayerConfiguration;
import org.deeplearning4j.nn.graph.ComputationGraph;
import org.deeplearning4j.nn.multilayer.MultiLayerNetwork;
import org.nd4j.linalg.dataset.DataSet;
import org.nd4j.linalg.dataset.api.MultiDataSet;
import org.nd4j.linalg.dataset.api.iterator.DataSetIterator;

/**
 * Class for conducting early stopping training locally (single machine), for training a
 * {@link MultiLayerNetwork}. To train a {@link ComputationGraph}, use {@link EarlyStoppingGraphTrainer}
 */
public class EarlyStoppingTrainer extends BaseEarlyStoppingTrainer {

    private MultiLayerNetwork net;
    private boolean isMultiEpoch = false;


    public EarlyStoppingTrainer(EarlyStoppingConfiguration earlyStoppingConfiguration,
                    MultiLayerConfiguration configuration, DataSetIterator train) {
        this(earlyStoppingConfiguration, new MultiLayerNetwork(configuration), train);
        net.init();
    }

    public EarlyStoppingTrainer(EarlyStoppingConfiguration esConfig, MultiLayerNetwork net,
                    DataSetIterator train) {
        this(esConfig, net, train, null);
    }

    public EarlyStoppingTrainer(EarlyStoppingConfiguration esConfig, MultiLayerNetwork net,
                    DataSetIterator train, EarlyStoppingListener listener) {
        super(esConfig, net, train, null, listener);
        this.net = net;
    }

    @Override
    protected void fit(DataSet ds) {
        net.fit(ds);
    }

    @Override
    protected void fit(MultiDataSet mds) {
        net.fit(mds);
    }

    @Override
    protected void pretrain(DataSet ds) {
        net.pretrain(new SingletonDataSetIterator(ds));
    }

    @Override
    protected void pretrain(MultiDataSet mds) {
        net.pretrain(new MultiDataSetWrapperIterator(new SingletonMultiDataSetIterator(mds)));
    }
}