org.nd4j.linalg.api.ops.impl.transforms.custom.DotProductAttention Maven / Gradle / Ivy
The newest version!
/*
* ******************************************************************************
* *
* *
* * 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.
* *
* * See the NOTICE file distributed with this work for additional
* * information regarding copyright ownership.
* * 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.NonNull;
import org.nd4j.autodiff.samediff.SDVariable;
import org.nd4j.autodiff.samediff.SameDiff;
import org.nd4j.common.base.Preconditions;
import org.nd4j.linalg.api.buffer.DataType;
import org.nd4j.linalg.api.ndarray.INDArray;
import org.nd4j.linalg.api.ops.DynamicCustomOp;
import java.util.Arrays;
import java.util.Collections;
import java.util.List;
@NoArgsConstructor
public class DotProductAttention extends DynamicCustomOp {
private boolean withWeights;
private boolean scaled;
public DotProductAttention(SameDiff sameDiff, SDVariable queries, SDVariable keys, SDVariable values, SDVariable mask, boolean scaled, boolean withWeights) {
super(null, sameDiff, mask == null ? new SDVariable[] {queries, keys, values} : new SDVariable[] {queries, keys, values, mask}, false);
this.scaled = scaled;
this.withWeights = withWeights;
addIArgument(scaled ? 1 : 0);
addIArgument(withWeights ? 1 : 0);
}
public DotProductAttention(@NonNull INDArray queries, @NonNull INDArray keys, @NonNull INDArray values, INDArray mask, boolean scaled){
this(queries, keys, values, mask, scaled, false);
}
public DotProductAttention(@NonNull INDArray queries, @NonNull INDArray keys, @NonNull INDArray values, INDArray mask, boolean scaled, boolean withWeights){
super(wrapFilterNull(queries, keys, values, mask), null);
this.scaled = scaled;
this.withWeights = withWeights;
addIArgument(scaled ? 1 : 0);
addIArgument(withWeights ? 1 : 0);
}
@Override
public String opName() {
return "dot_product_attention";
}
@Override
public List doDiff(List gradient) {
SDVariable mask = args().length == 4 ? arg(3) : null;
if (iArguments.size() > 0) {
this.scaled = iArguments.get(0).intValue() == 1 ? true : false;
this.withWeights = iArguments.size() > 0 && iArguments.get(1).intValue() == 1 ? true : false ;
}
return Arrays.asList(new DotProductAttentionBp(sameDiff, arg(0), arg(1), arg(2), gradient.get(0), mask, scaled).outputVariables());
}
@Override
public List calculateOutputDataTypes(List dataTypes){
Preconditions.checkState(dataTypes != null && (dataTypes.size() == 3 || dataTypes.size() == 4), "Expected exactly 3 or 4 input datatypes, got %s", dataTypes);
DataType first = dataTypes.get(0);
for( int i=0; i 0){
Preconditions.checkState(first == dataTypes.get(i), "All datatypes must be same type, got input datatypes %s", dataTypes);
}
}
if(withWeights){
return Arrays.asList(first, first);
}else{
return Collections.singletonList(first);
}
}
@Override
public int getNumOutputs() {
if(withWeights){
return 2;
}else{
return 1;
}
}
}