Java example source code file (ParameterAveragingElementAddFunction.java)

This example Java source code file (ParameterAveragingElementAddFunction.java) is included in the alvinalexander.com "Java Source Code Warehouse" project. The intent of this project is to help you "Learn Java by Example" ^TM.

Learn more about this Java project at its project page.

Java - Java tags/keywords

exception, function2, indarray, override, parameteraveragingaggregationtuple, parameteraveragingelementaddfunction, parameteraveragingtrainingresult, sparktrainingstats, updateraggregator

The ParameterAveragingElementAddFunction.java Java example source code

package org.deeplearning4j.spark.impl.paramavg.aggregator;

import org.apache.spark.api.java.function.Function2;
import org.deeplearning4j.nn.updater.aggregate.UpdaterAggregator;
import org.deeplearning4j.nn.updater.graph.ComputationGraphUpdater;
import org.deeplearning4j.spark.api.stats.SparkTrainingStats;
import org.deeplearning4j.spark.impl.paramavg.ParameterAveragingTrainingResult;
import org.nd4j.linalg.api.ndarray.INDArray;

/**
 * Add function for parameter averaging
 *
 * @author Alex Black
 */
public class ParameterAveragingElementAddFunction implements Function2<ParameterAveragingAggregationTuple, ParameterAveragingTrainingResult, ParameterAveragingAggregationTuple> {

    @Override
    public ParameterAveragingAggregationTuple call(ParameterAveragingAggregationTuple tuple, ParameterAveragingTrainingResult result) throws Exception {

        if (tuple == null) {
            UpdaterAggregator ua = (result.getUpdater() != null ? result.getUpdater().getAggregator(true) : null);
            ComputationGraphUpdater.Aggregator uaGraph = (result.getGraphUpdater() != null ? result.getGraphUpdater().getAggregator(true) : null);
            return new ParameterAveragingAggregationTuple(result.getParameters(), ua, uaGraph, result.getScore(), 1, result.getSparkTrainingStats());
        }

        INDArray params = tuple.getParametersSum().addi(result.getParameters());
        UpdaterAggregator aggregator;
        if (tuple.getUpdaterAggregator() == null) {
            if (result.getUpdater() == null) aggregator = null;
            else aggregator = result.getUpdater().getAggregator(true);
        } else {
            aggregator = tuple.getUpdaterAggregator();
            if (result.getUpdater() != null) aggregator.aggregate(result.getUpdater());
        }

        ComputationGraphUpdater.Aggregator aggregatorGraph;
        if (tuple.getUpdaterAggregatorGraph() == null) {
            if (result.getGraphUpdater() == null) aggregatorGraph = null;
            else aggregatorGraph = result.getGraphUpdater().getAggregator(true);
        } else {
            aggregatorGraph = tuple.getUpdaterAggregatorGraph();
            if (result.getGraphUpdater() != null) aggregatorGraph.aggregate(result.getGraphUpdater());
        }

        double scoreSum = tuple.getScoreSum() + result.getScore();
        SparkTrainingStats stats = tuple.getSparkTrainingStats();
        if (result.getSparkTrainingStats() != null) {
            if (stats == null) stats = result.getSparkTrainingStats();
            else stats.addOtherTrainingStats(result.getSparkTrainingStats());
        }
        return new ParameterAveragingAggregationTuple(params, aggregator, aggregatorGraph, scoreSum, tuple.getAggregationsCount() + 1, stats);
    }
}