org.nd4j.linalg.api.ops.impl.transforms.custom.XwPlusB Maven / Gradle / Ivy
/*******************************************************************************
* Copyright (c) 2015-2018 Skymind, Inc.
*
* This program and the accompanying materials are made available under the
* terms of the Apache License, Version 2.0 which is available at
* https://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.
*
* SPDX-License-Identifier: Apache-2.0
******************************************************************************/
package org.nd4j.linalg.api.ops.impl.transforms.custom;
import lombok.NoArgsConstructor;
import lombok.val;
import org.apache.commons.lang3.ArrayUtils;
import org.nd4j.autodiff.samediff.SDVariable;
import org.nd4j.autodiff.samediff.SameDiff;
import org.nd4j.base.Preconditions;
import org.nd4j.imports.NoOpNameFoundException;
import org.nd4j.imports.graphmapper.tf.TFGraphMapper;
import org.nd4j.linalg.api.blas.params.MMulTranspose;
import org.nd4j.linalg.api.buffer.DataType;
import org.nd4j.linalg.api.ndarray.INDArray;
import org.nd4j.linalg.api.ops.DynamicCustomOp;
import org.tensorflow.framework.AttrValue;
import org.tensorflow.framework.GraphDef;
import org.tensorflow.framework.NodeDef;
import java.util.*;
/**
* Composed op: mmul (X, W) + b
*
* @author Max Pumperla
*/
@NoArgsConstructor
public class XwPlusB extends DynamicCustomOp {
public XwPlusB(SameDiff sameDiff, SDVariable input, SDVariable weights, SDVariable bias) {
super(null, sameDiff, new SDVariable[] {input, weights, bias}, false);
}
@Override
public String opName() {
return "xw_plus_b";
}
@Override
public String tensorflowName() {
throw new NoOpNameFoundException("No tensorflow name found for shape " + opName());
}
@Override
public String onnxName() {
throw new NoOpNameFoundException("No onnx name found for shape " + opName());
}
@Override
public List doDiff(List gradient) {
SDVariable in = arg(0);
SDVariable w = arg(1);
SDVariable dLdOut = gradient.get(0);
SDVariable dLdb = dLdOut.sum(0);
SDVariable dLdIn = sameDiff.mmul(dLdOut, w, MMulTranspose.builder()
.transposeB(true)
.build());
SDVariable dLdW = sameDiff.mmul(in, dLdOut, MMulTranspose.builder()
.transposeA(true)
.build());
return Arrays.asList(dLdIn, dLdW, dLdb);
}
@Override
public List calculateOutputDataTypes(List dataTypes){
Preconditions.checkState(dataTypes != null && dataTypes.size() == 3, "Expected exactly 3 input datatypes, got %s", dataTypes);
DataType first = dataTypes.get(0);
for( int i=0; i<3; i++ ) {
Preconditions.checkState(dataTypes.get(i).isFPType(), "Input %s datatype must be a floating point type, got datypes %s", dataTypes);
if(i > 0){
Preconditions.checkState(first == dataTypes.get(i), "All datatypes must be same type, got input datatypes %s", dataTypes);
}
}
return Collections.singletonList(first);
}
}