deepboof.DFunction Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of main Show documentation

Trainer Agnostic Deep Learning

There is a newer version: 0.5.2

/*
 * Copyright (c) 2016, Peter Abeles. All Rights Reserved.
 *
 * This file is part of DeepBoof
 *
 * Licensed 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 deepboof;

import java.util.List;

/**
 * {@link Function Functions} which also implement the backwards step and compute the gradient for all inputs.
 * Functions have two modes for operation, learning and evaluating.  When in learning mode they are free
 * to modify their internal state during the forward step, otherwise, while in evaluation mode, they are not
 * allowed to modify their state.  By default, all functions start in evaluation mode.
 *
 * @author Peter Abeles
 */
public interface DFunction> extends Function {
	/**
	 * Puts the function into learning mode.
	 */
	void learning();

	/**
	 * Puts the function into evaluation mode.
	 */
	void evaluating();

	/**
	 * Computes the derivatives of all the inputs and parameters to this function.  The {@link #forward} function
	 * must be called first before calling this one and the same inputs and parameters must be passed in.
	 *
	 * @param input The same input tensor which was passed in during the forward pass.
	 * @param dout Derivative of output, computed from next layer.
	 * @param gradientInput gradient of input {@link Tensor}
	 * @param gradientParameters  Gradients of all parameter {@link Tensor Tensors}.  Same order as parameters
	 *                            in {@link #forward}
	 */
	void backwards(T input, T dout , T gradientInput , List gradientParameters );

	/**
	 * Is the function in the learning state?
	 *
	 * @return true if in learning state or false if it's in the evaluation state
	 */
	boolean isLearning();
}