All Downloads are FREE. Search and download functionalities are using the official Maven repository.

ai.djl.training.loss.SigmoidBinaryCrossEntropyLoss Maven / Gradle / Ivy

There is a newer version: 0.30.0
Show newest version
/*
 * 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.NDArrays;
import ai.djl.ndarray.NDList;
import ai.djl.nn.Activation;

/**
 * {@code SigmoidBinaryCrossEntropyLoss} is a type of {@link Loss}.
 *
 * 

Sigmoid binary cross-entropy loss is defined by: \(L = -\sum_i {label_i * log(prob_i) * * posWeight + (1 - label_i) * log(1 - prob_i)}\) where \(prob = \frac{1}{1 + e^{-pred}}\) */ public class SigmoidBinaryCrossEntropyLoss extends Loss { private float weight; private boolean fromSigmoid; /** Performs Sigmoid cross-entropy loss for binary classification. */ public SigmoidBinaryCrossEntropyLoss() { this("SigmoidBinaryCrossEntropyLoss"); } /** * Performs Sigmoid cross-entropy loss for binary classification. * * @param name the name of the loss */ public SigmoidBinaryCrossEntropyLoss(String name) { this(name, 1, false); } /** * Performs Sigmoid cross-entropy loss for binary classification. * * @param name the name of the loss * @param weight the weight to apply on the loss value, default 1 * @param fromSigmoid whether the input is from the output of sigmoid, default false */ public SigmoidBinaryCrossEntropyLoss(String name, float weight, boolean fromSigmoid) { super(name); this.weight = weight; this.fromSigmoid = fromSigmoid; } /** {@inheritDoc} */ @Override public NDArray evaluate(NDList label, NDList prediction) { NDArray pred = prediction.singletonOrThrow(); NDArray lab = label.singletonOrThrow(); lab = lab.reshape(pred.getShape()); NDArray loss; if (!fromSigmoid) { // TODO: Add Position weight option loss = Activation.relu(pred) .sub(pred.mul(lab)) .add(Activation.softPlus(pred.abs().neg())); } else { loss = epsLog(pred) .mul(lab) .add(epsLog(NDArrays.sub(1., pred)).mul(NDArrays.sub(1., lab))); } if (weight != 1f) { loss = loss.mul(weight); } return loss.mean(); } /** * Computes a log with added epsilon to avoid errors. * * @param a the input array * @return the computed value */ private NDArray epsLog(NDArray a) { double eps = 1e-12; return a.add(eps).log(); } }





© 2015 - 2024 Weber Informatics LLC | Privacy Policy