com.simiacryptus.mindseye.opt.line.StaticLearningRate Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of mindseye-core Show documentation
Show all versions of mindseye-core Show documentation
Core Neural Networks Framework
/*
* Copyright (c) 2019 by Andrew Charneski.
*
* The author licenses this file to you under the
* Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance
* with the License. You may obtain a copy
* of the License at
*
* http://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.
*/
package com.simiacryptus.mindseye.opt.line;
import com.simiacryptus.mindseye.lang.PointSample;
import com.simiacryptus.mindseye.opt.TrainingMonitor;
import com.simiacryptus.ref.wrappers.RefString;
import javax.annotation.Nonnull;
import javax.annotation.Nullable;
public class StaticLearningRate implements LineSearchStrategy {
private double minimumRate = 1e-12;
private double rate = 1e-4;
public StaticLearningRate(double rate) {
this.rate = rate;
}
public double getMinimumRate() {
return minimumRate;
}
@Nonnull
public StaticLearningRate setMinimumRate(final double minimumRate) {
this.minimumRate = minimumRate;
return this;
}
public double getRate() {
return rate;
}
@Nonnull
public StaticLearningRate setRate(final double rate) {
this.rate = rate;
return this;
}
@Nullable
@Override
public PointSample step(@Nonnull final LineSearchCursor cursor, @Nonnull final TrainingMonitor monitor) {
double thisRate = rate;
final LineSearchPoint startPoint = cursor.step(0, monitor);
assert startPoint != null;
final double startValue = startPoint.getPointSum(); // theta(0)
@Nullable
LineSearchPoint lastStep = null;
try {
while (true) {
if (null != lastStep) lastStep.freeRef();
lastStep = cursor.step(thisRate, monitor);
assert lastStep != null;
double lastValue = lastStep.getPointSum();
if (!Double.isFinite(lastValue)) {
lastValue = Double.POSITIVE_INFINITY;
}
if (lastValue + startValue * 1e-15 > startValue) {
monitor.log(RefString.format("Non-decreasing runStep. %s > %s at " + thisRate, lastValue, startValue));
thisRate /= 2;
if (thisRate < getMinimumRate()) {
return startPoint.getPoint();
}
} else {
return lastStep.getPoint();
}
}
} finally {
if (null != lastStep) lastStep.freeRef();
startPoint.freeRef();
cursor.freeRef();
}
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy