2019-06-06 14:21:15 +02:00
/*******************************************************************************
* 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
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
//
// @author Yurii Shyrma, created on 05.04.2018
//
# include <ops/declarable/CustomOperations.h>
2020-03-02 10:49:41 +01:00
namespace sd {
2019-06-06 14:21:15 +02:00
namespace ops {
//////////////////////////////////////////////////////////////////////////
CUSTOM_OP_IMPL ( dynamic_bidirectional_rnn , 7 , 4 , false , 0 , 0 ) {
auto x = INPUT_VARIABLE ( 0 ) ; // input [time x bS x inSize] or [bS x time x inSize], shape depends on timeMajor parameter
auto WxFW = INPUT_VARIABLE ( 1 ) ; // input-to-hidden weights for forward RNN, [inSize x numUnitsFW]
auto WhFW = INPUT_VARIABLE ( 2 ) ; // hidden-to-hidden weights for forward RNN, [numUnitsFW x numUnitsFW]
auto bFW = INPUT_VARIABLE ( 3 ) ; // biases for forward RNN, [2*numUnitsFW]
auto WxBW = INPUT_VARIABLE ( 4 ) ; // input-to-hidden weights for backward RNN, [inSize x numUnitsBW]
auto WhBW = INPUT_VARIABLE ( 5 ) ; // hidden-to-hidden weights for backward RNN, [numUnitsBW x numUnitsBW]
auto bBW = INPUT_VARIABLE ( 6 ) ; // biases for backward RNN, [2*v]
NDArray * h0FW = nullptr ; // initial cell output for forward RNN (at time step = 0) [bS x numUnitsFW]
NDArray * h0BW = nullptr ; // initial cell output for backward RNN (at time step = 0) [bS x numUnitsBW]
NDArray * maxTimeStep = nullptr ; // vector [bS] containing integer values within [0,time), each element of this vector set max time step per each input in batch, this means there are no calculations for time >= maxTimeStep
const int timeMajor = block . getIArguments ( ) - > size ( ) > 0 ? INT_ARG ( 0 ) : 0 ; // if non zero then [time, bS, ...], else [bS, time, ...]
switch ( block . width ( ) ) {
case 8 :
maxTimeStep = INPUT_VARIABLE ( 7 ) ;
break ;
case 9 :
h0FW = INPUT_VARIABLE ( 7 ) ;
h0BW = INPUT_VARIABLE ( 8 ) ;
break ;
case 10 :
h0FW = INPUT_VARIABLE ( 7 ) ;
h0BW = INPUT_VARIABLE ( 8 ) ;
maxTimeStep = INPUT_VARIABLE ( 9 ) ;
break ;
}
2020-03-03 05:32:37 +01:00
2019-06-06 14:21:15 +02:00
auto hFW = OUTPUT_VARIABLE ( 0 ) ; // cell outputs for forward RNN [time x bS x numUnitsFW] or [bS x time x numUnitsFW], shape depends on timeMajor parameter
auto hBW = OUTPUT_VARIABLE ( 1 ) ; // cell outputs for backward RNN [time x bS x numUnitsBW] or [bS x time x numUnitsBW], shape depends on timeMajor parameter
auto hFWFinal = OUTPUT_VARIABLE ( 2 ) ; // final cell out for forward RNN [bS x numUnitsFW]
auto hBWFinal = OUTPUT_VARIABLE ( 3 ) ; // final cell out for backward RNN [bS x numUnitsBF]
REQUIRE_TRUE ( x - > rankOf ( ) = = 3 , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: input array must have rank = 3, but got %i instead ! " , x - > rankOf ( ) ) ;
2020-03-03 05:32:37 +01:00
REQUIRE_TRUE ( WxFW - > rankOf ( ) = = 2 , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: input-to-hidden weights array (for forward RNN) must have rank = 2, but got %i instead ! " , WxFW - > rankOf ( ) ) ;
REQUIRE_TRUE ( WxBW - > rankOf ( ) = = 2 , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: input-to-hidden weights array (for backward RNN) must have rank = 2, but got %i instead ! " , WxBW - > rankOf ( ) ) ;
2019-06-06 14:21:15 +02:00
const int inRank = x - > rankOf ( ) ;
const int time = timeMajor ? x - > sizeAt ( 0 ) : x - > sizeAt ( 1 ) ;
const int bS = timeMajor ? x - > sizeAt ( 1 ) : x - > sizeAt ( 0 ) ;
const int numUnitsFW = WxFW - > sizeAt ( 1 ) ;
const int numUnitsBW = WxBW - > sizeAt ( 1 ) ;
2020-03-03 05:32:37 +01:00
std : : vector < Nd4jLong > expectedWhFWshape = { numUnitsFW , numUnitsFW } ;
std : : vector < Nd4jLong > expectedWhBWshape = { numUnitsBW , numUnitsBW } ;
std : : vector < Nd4jLong > expectedbFWshape = { 2 * numUnitsFW } ;
std : : vector < Nd4jLong > expectedbBWshape = { 2 * numUnitsBW } ;
REQUIRE_TRUE ( WhFW - > isSameShape ( expectedWhFWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of hidden-to-hidden weights array (for forward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedWhFWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( WhFW ) . c_str ( ) ) ;
REQUIRE_TRUE ( WhBW - > isSameShape ( expectedWhBWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of hidden-to-hidden weights array (for backward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedWhBWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( WhBW ) . c_str ( ) ) ;
REQUIRE_TRUE ( bFW - > isSameShape ( expectedbFWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of biases array (for forward RNN), expected is %s, but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedbFWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( bFW ) . c_str ( ) ) ;
REQUIRE_TRUE ( bBW - > isSameShape ( expectedbBWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of biases array (for backward RNN), expected is %s, but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedbBWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( bBW ) . c_str ( ) ) ;
if ( h0FW ) {
std : : vector < Nd4jLong > expectedh0FWshape = { bS , numUnitsFW } ;
REQUIRE_TRUE ( h0FW - > isSameShape ( expectedh0FWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of initial cell output array (for forward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedh0FWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( h0FW ) . c_str ( ) ) ;
}
if ( h0BW ) {
std : : vector < Nd4jLong > expectedh0BWshape = { bS , numUnitsBW } ;
REQUIRE_TRUE ( h0BW - > isSameShape ( expectedh0BWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of initial cell output array (for backward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedh0BWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( h0BW ) . c_str ( ) ) ;
}
if ( maxTimeStep ) {
std : : vector < Nd4jLong > expectedmaxTimeStepshape = { bS } ;
REQUIRE_TRUE ( maxTimeStep - > isSameShape ( expectedmaxTimeStepshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of maxTimeStep array, expected is [%i], but got %s instead ! " , bS , ShapeUtils : : shapeAsString ( maxTimeStep ) . c_str ( ) ) ;
}
2019-06-06 14:21:15 +02:00
// forward steps
2020-03-02 10:49:41 +01:00
sd : : ops : : dynamic_rnn dynamicRnn ;
2020-01-30 08:07:24 +01:00
auto resultsFW = dynamicRnn . evaluate ( { x , WxFW , WhFW , bFW , h0FW , maxTimeStep } , { timeMajor } ) ;
2020-03-10 05:42:50 +01:00
hFW - > assign ( resultsFW . at ( 0 ) ) ; // [time x bS x numUnitsFW] or [bS x time x numUnitsFW]
hFWFinal - > assign ( resultsFW . at ( 1 ) ) ;
2019-06-06 14:21:15 +02:00
auto seqLen = maxTimeStep ;
if ( seqLen = = nullptr ) {
// FIXME: which datatype should be used here?
2020-03-02 10:49:41 +01:00
seqLen = new NDArray ( x - > ordering ( ) , { bS } , sd : : DataType : : INT64 , block . launchContext ( ) ) ;
2019-06-06 14:21:15 +02:00
seqLen - > assign ( time ) ; // set each element of seqLen to be equal to time
}
// reverse x
2020-03-02 10:49:41 +01:00
sd : : ops : : reverse_sequence reverse ;
2020-01-30 08:07:24 +01:00
auto resultsIn = timeMajor ? reverse . evaluate ( { x , seqLen } , { 0 , 1 } ) : reverse . evaluate ( { x , seqLen } , { 1 , 0 } ) ;
2020-03-10 05:42:50 +01:00
REQUIRE_TRUE ( resultsIn . status ( ) = = ND4J_STATUS_OK , 0 , " dynamic_bidirectional_rnn: there is a problem with reverse on the sequence. " ) ;
auto revInput = resultsIn . at ( 0 ) ;
2019-06-06 14:21:15 +02:00
2020-03-03 05:32:37 +01:00
// backward steps
2020-01-30 08:07:24 +01:00
auto resultsBW = dynamicRnn . evaluate ( { revInput , WxBW , WhBW , bBW , h0BW , maxTimeStep } , { timeMajor } ) ;
2020-03-10 05:42:50 +01:00
auto hBWtemp = resultsBW . at ( 0 ) ; // [time x bS x numUnitsBW] or [ bS x time xnumUnitsBW]
hBWFinal - > assign ( resultsBW . at ( 1 ) ) ;
2019-06-06 14:21:15 +02:00
2020-03-03 05:32:37 +01:00
// reverse hBWtemp
2020-01-30 08:07:24 +01:00
auto resultsOut = timeMajor ? reverse . evaluate ( { hBWtemp , seqLen } , { 0 , 1 } ) : reverse . evaluate ( { hBWtemp , seqLen } , { 1 , 0 } ) ;
2020-03-10 05:42:50 +01:00
hBW - > assign ( resultsOut . at ( 0 ) ) ;
2019-06-06 14:21:15 +02:00
if ( seqLen ! = maxTimeStep )
delete seqLen ;
return Status : : OK ( ) ;
}
DECLARE_TYPES ( dynamic_bidirectional_rnn ) {
getOpDescriptor ( )
2020-03-02 10:49:41 +01:00
- > setAllowedInputTypes ( sd : : DataType : : ANY )
2019-06-06 14:21:15 +02:00
- > setAllowedOutputTypes ( { ALL_FLOATS } ) ;
}
2020-03-03 05:32:37 +01:00
DECLARE_SHAPE_FN ( dynamic_bidirectional_rnn ) {
2019-06-06 14:21:15 +02:00
auto x = INPUT_VARIABLE ( 0 ) ; // input [time x bS x inSize] or [bS x time x inSize], shape depends on timeMajor parameter
auto WxFW = INPUT_VARIABLE ( 1 ) ; // input-to-hidden weights for forward RNN, [inSize x numUnitsFW]
auto WhFW = INPUT_VARIABLE ( 2 ) ; // hidden-to-hidden weights for forward RNN, [numUnitsFW x numUnitsFW]
auto bFW = INPUT_VARIABLE ( 3 ) ; // biases for forward RNN, [2*numUnitsFW]
auto WxBW = INPUT_VARIABLE ( 4 ) ; // input-to-hidden weights for backward RNN, [inSize x numUnitsBW]
auto WhBW = INPUT_VARIABLE ( 5 ) ; // hidden-to-hidden weights for backward RNN, [numUnitsBW x numUnitsBW]
auto bBW = INPUT_VARIABLE ( 6 ) ; // biases for backward RNN, [2*numUnitsBW]
NDArray * h0FW = nullptr ; // initial cell output for forward RNN (at time step = 0) [bS x numUnitsFW]
NDArray * h0BW = nullptr ; // initial cell output for backward RNN (at time step = 0) [bS x numUnitsBW]
NDArray * maxTimeStep = nullptr ; // vector [bS] containing integer values within [0,time), each element of this vector set max time step per each input in batch, this means there are no calculations for time >= maxTimeStep
const int timeMajor = block . getIArguments ( ) - > size ( ) > 0 ? INT_ARG ( 0 ) : 0 ; // if true then [time, bS, ...], else [bS, time, ...]
2020-03-03 05:32:37 +01:00
2019-06-06 14:21:15 +02:00
switch ( block . width ( ) ) {
case 8 :
maxTimeStep = INPUT_VARIABLE ( 7 ) ;
break ;
case 9 :
h0FW = INPUT_VARIABLE ( 7 ) ;
h0BW = INPUT_VARIABLE ( 8 ) ;
break ;
case 10 :
h0FW = INPUT_VARIABLE ( 7 ) ;
h0BW = INPUT_VARIABLE ( 8 ) ;
maxTimeStep = INPUT_VARIABLE ( 9 ) ;
break ;
}
REQUIRE_TRUE ( x - > rankOf ( ) = = 3 , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: input array must have rank = 3, but got %i instead ! " , x - > rankOf ( ) ) ;
2020-03-03 05:32:37 +01:00
REQUIRE_TRUE ( WxFW - > rankOf ( ) = = 2 , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: input-to-hidden weights array (for forward RNN) must have rank = 2, but got %i instead ! " , WxFW - > rankOf ( ) ) ;
REQUIRE_TRUE ( WxBW - > rankOf ( ) = = 2 , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: input-to-hidden weights array (for backward RNN) must have rank = 2, but got %i instead ! " , WxBW - > rankOf ( ) ) ;
2019-06-06 14:21:15 +02:00
const int inRank = x - > rankOf ( ) ;
const int time = timeMajor ? x - > sizeAt ( 0 ) : x - > sizeAt ( 1 ) ;
const int bS = timeMajor ? x - > sizeAt ( 1 ) : x - > sizeAt ( 0 ) ;
const int numUnitsFW = WxFW - > sizeAt ( 1 ) ;
const int numUnitsBW = WxBW - > sizeAt ( 1 ) ;
2020-03-03 05:32:37 +01:00
std : : vector < Nd4jLong > expectedWhFWshape = { numUnitsFW , numUnitsFW } ;
std : : vector < Nd4jLong > expectedWhBWshape = { numUnitsBW , numUnitsBW } ;
std : : vector < Nd4jLong > expectedbFWshape = { 2 * numUnitsFW } ;
std : : vector < Nd4jLong > expectedbBWshape = { 2 * numUnitsBW } ;
REQUIRE_TRUE ( WhFW - > isSameShape ( expectedWhFWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of hidden-to-hidden weights array (for forward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedWhFWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( WhFW ) . c_str ( ) ) ;
REQUIRE_TRUE ( WhBW - > isSameShape ( expectedWhBWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of hidden-to-hidden weights array (for backward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedWhBWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( WhBW ) . c_str ( ) ) ;
REQUIRE_TRUE ( bFW - > isSameShape ( expectedbFWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of biases array (for forward RNN), expected is %s, but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedbFWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( bFW ) . c_str ( ) ) ;
REQUIRE_TRUE ( bBW - > isSameShape ( expectedbBWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of biases array (for backward RNN), expected is %s, but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedbBWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( bBW ) . c_str ( ) ) ;
if ( h0FW ) {
std : : vector < Nd4jLong > expectedh0FWshape = { bS , numUnitsFW } ;
REQUIRE_TRUE ( h0FW - > isSameShape ( expectedh0FWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of initial cell output array (for forward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedh0FWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( h0FW ) . c_str ( ) ) ;
}
if ( h0BW ) {
std : : vector < Nd4jLong > expectedh0BWshape = { bS , numUnitsBW } ;
REQUIRE_TRUE ( h0BW - > isSameShape ( expectedh0BWshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of initial cell output array (for backward RNN), expected is %s but got %s instead ! " , ShapeUtils : : shapeAsString ( expectedh0BWshape ) . c_str ( ) , ShapeUtils : : shapeAsString ( h0BW ) . c_str ( ) ) ;
}
if ( maxTimeStep ) {
std : : vector < Nd4jLong > expectedmaxTimeStepshape = { bS } ;
REQUIRE_TRUE ( maxTimeStep - > isSameShape ( expectedmaxTimeStepshape ) , 0 , " DYNAMIC_BIDIRECTIONAL_RNN custom operation: wrong shape of maxTimeStep array, expected is [%i], but got %s instead ! " , bS , ShapeUtils : : shapeAsString ( maxTimeStep ) . c_str ( ) ) ;
}
2019-06-06 14:21:15 +02:00
// evaluate output shapeInfos
Nd4jLong * hFWShapeInfo ( nullptr ) , * hBWShapeInfo ( nullptr ) , * hFWFinalPrevShapeInfo ( nullptr ) , * hBWFinalPrevShapeInfo ( nullptr ) ;
ALLOCATE ( hFWShapeInfo , block . getWorkspace ( ) , shape : : shapeInfoLength ( inRank ) , Nd4jLong ) ;
ALLOCATE ( hBWShapeInfo , block . getWorkspace ( ) , shape : : shapeInfoLength ( inRank ) , Nd4jLong ) ;
ALLOCATE ( hFWFinalPrevShapeInfo , block . getWorkspace ( ) , shape : : shapeInfoLength ( inRank - 1 ) , Nd4jLong ) ;
ALLOCATE ( hBWFinalPrevShapeInfo , block . getWorkspace ( ) , shape : : shapeInfoLength ( inRank - 1 ) , Nd4jLong ) ;
2020-03-03 05:32:37 +01:00
hFWShapeInfo [ 0 ] = hBWShapeInfo [ 0 ] = inRank ;
2019-06-06 14:21:15 +02:00
hFWShapeInfo [ 1 ] = hBWShapeInfo [ 1 ] = timeMajor ? time : bS ;
hFWShapeInfo [ 2 ] = hBWShapeInfo [ 2 ] = timeMajor ? bS : time ;
hFWShapeInfo [ 3 ] = numUnitsFW ;
hBWShapeInfo [ 3 ] = numUnitsBW ;
hFWFinalPrevShapeInfo [ 0 ] = hBWFinalPrevShapeInfo [ 0 ] = inRank - 1 ;
2020-03-03 05:32:37 +01:00
hFWFinalPrevShapeInfo [ 1 ] = hBWFinalPrevShapeInfo [ 1 ] = bS ;
2019-06-06 14:21:15 +02:00
hFWFinalPrevShapeInfo [ 2 ] = numUnitsFW ;
hBWFinalPrevShapeInfo [ 2 ] = numUnitsBW ;
2020-05-09 07:06:14 +02:00
ShapeUtils : : updateStridesAndType ( hFWShapeInfo , x - > shapeInfo ( ) , x - > ordering ( ) ) ;
ShapeUtils : : updateStridesAndType ( hBWShapeInfo , x - > shapeInfo ( ) , x - > ordering ( ) ) ;
ShapeUtils : : updateStridesAndType ( hFWFinalPrevShapeInfo , x - > shapeInfo ( ) , x - > ordering ( ) ) ;
ShapeUtils : : updateStridesAndType ( hBWFinalPrevShapeInfo , x - > shapeInfo ( ) , x - > ordering ( ) ) ;
2020-03-03 05:32:37 +01:00
2019-06-06 14:21:15 +02:00
return SHAPELIST ( CONSTANT ( hFWShapeInfo ) , CONSTANT ( hBWShapeInfo ) , CONSTANT ( hFWFinalPrevShapeInfo ) , CONSTANT ( hBWFinalPrevShapeInfo ) ) ;
2020-03-03 05:32:37 +01:00
}
2019-06-06 14:21:15 +02:00
}
}
