ai.djl.training.loss.L2Loss Maven / Gradle / Ivy
/*
* Copyright 2019 Amazon.com, Inc. or its affiliates. All Rights Reserved.
*
* Licensed under the Apache License, Version 2.0 (the "License"). You may not use this file except in compliance
* with the License. A copy of the License is located at
*
* http://aws.amazon.com/apache2.0/
*
* or in the "license" file accompanying this file. This file 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 ai.djl.training.loss;
import ai.djl.ndarray.NDArray;
import ai.djl.ndarray.NDList;
/**
* Calculates L2Loss between label and prediction, a.k.a. MSE(Mean Square Error).
*
* L2 loss is defined by \(L = \frac{1}{2} \sum_i \vert {label}_i - {prediction}_i \vert^2\)
*/
public class L2Loss extends Loss {
private float weight;
/** Calculate L2Loss between the label and prediction, a.k.a. MSE(Mean Square Error). */
public L2Loss() {
this("L2Loss");
}
/**
* Calculate L2Loss between the label and prediction, a.k.a. MSE(Mean Square Error).
*
* @param name the name of the loss
*/
public L2Loss(String name) {
this(name, 1.f / 2);
}
/**
* Calculates L2Loss between the label and prediction, a.k.a. MSE(Mean Square Error).
*
* @param name the name of the loss
* @param weight the weight to apply on loss value, default 1/2
*/
public L2Loss(String name, float weight) {
super(name);
this.weight = weight;
}
/** {@inheritDoc} */
@Override
public NDArray evaluate(NDList label, NDList prediction) {
NDArray pred = prediction.singletonOrThrow();
NDArray labelReshaped = label.singletonOrThrow().reshape(pred.getShape());
NDArray loss = labelReshaped.sub(pred).square().mul(weight);
return loss.mean();
}
}