wehub-resource-sync
199 lines
7.9 KiB
C++
199 lines
7.9 KiB
C++
/*
|
|
* ******************************************************************************
|
|
* *
|
|
* *
|
|
* * 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
|
|
* *****************************************************************************
|
|
*/
|
|
|
|
//
|
|
// @author Oleh Semeniv (oleg.semeniv@gmail.com)
|
|
//
|
|
#include <execution/Threads.h>
|
|
#include <math/platformmath.h>
|
|
#include <math/templatemath.h>
|
|
#include <ops/declarable/helpers/updatersHelpers.h>
|
|
#if NOT_EXCLUDED(OP_amsgrad_updater)
|
|
namespace sd {
|
|
namespace ops {
|
|
namespace helpers {
|
|
|
|
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
|
|
template <typename T>
|
|
static void amsGradUpdater_(NDArray& gradient, NDArray& initStateV, NDArray& initStateM,
|
|
NDArray& initStateH, NDArray& update, NDArray& stateV, NDArray& stateM,
|
|
NDArray& stateH, const double dLr, const double dBeta1, const double dBeta2,
|
|
const double dEpsilon, const int nIteration) {
|
|
const T* grad = gradient.bufferAsT<T>();
|
|
const T* initV = initStateV.bufferAsT<T>();
|
|
const T* initM = initStateM.bufferAsT<T>();
|
|
const T* initH = initStateH.bufferAsT<T>();
|
|
|
|
T* up = update.bufferAsT<T>();
|
|
T* stV = stateV.bufferAsT<T>();
|
|
T* stM = stateM.bufferAsT<T>();
|
|
T* stH = stateH.bufferAsT<T>();
|
|
|
|
const T lr = static_cast<T>(dLr);
|
|
const T beta1 = static_cast<T>(dBeta1);
|
|
const T beta2 = static_cast<T>(dBeta2);
|
|
T epsilon = static_cast<T>(dEpsilon);
|
|
//fp16 to prevent underflow
|
|
if(epsilon == 0.0) {
|
|
epsilon = static_cast<T>(1e-7);
|
|
}
|
|
const T iteration = static_cast<T>(nIteration);
|
|
|
|
T epsilonT = lr * sd::math::sd_sqrt<T, T>(1.0 - sd::math::sd_pow<T, T, T>(beta2, (iteration + 1))) /
|
|
(1.0 - sd::math::sd_pow<T, T, T>(beta1, (iteration + 1)));
|
|
|
|
if (sd::math::sd_isnan(epsilonT) || 0 == epsilonT || sd::math::sd_isinf(epsilonT)) epsilonT = epsilon;
|
|
|
|
const T mbeta1 = (1 - beta1);
|
|
const T mbeta2 = (1 - beta2);
|
|
|
|
bool bEws1 = 1 == gradient.ews() && 1 == update.ews() && 1 == stateM.ews() && 1 == initStateM.ews() &&
|
|
1 == stateV.ews() && 1 == initStateV.ews() && 1 == stateH.ews() && 1 == initStateH.ews();
|
|
bool bSameOrdering = gradient.ordering() == update.ordering() && update.ordering() == stateV.ordering() &&
|
|
stateV.ordering() == initStateV.ordering() && stateV.ordering() == initStateM.ordering() &&
|
|
stateM.ordering() == initStateM.ordering() && stateM.ordering() == initStateH.ordering() &&
|
|
stateH.ordering() == initStateH.ordering();
|
|
|
|
if (bEws1 && bSameOrdering) {
|
|
auto func = PRAGMA_THREADS_FOR {
|
|
for (auto i = start; i < stop; i++) {
|
|
stM[i] = beta1 * initM[i] + grad[i] * mbeta1;
|
|
stV[i] = beta2 * initV[i] + grad[i] * grad[i] * mbeta2;
|
|
stH[i] = sd::math::sd_max(initH[i], stV[i]);
|
|
|
|
up[i] = epsilonT * stM[i] / (sd::math::sd_sqrt<T, T>(stH[i]) + epsilon);
|
|
}
|
|
};
|
|
|
|
samediff::Threads::parallel_for(func, 0, gradient.lengthOf(), 1);
|
|
return;
|
|
}
|
|
|
|
// Cache shape and stride information
|
|
sd::LongType gradRank = gradient.rankOf();
|
|
sd::LongType updateRank = update.rankOf();
|
|
sd::LongType initStateVRank = initStateV.rankOf();
|
|
sd::LongType stateVRank = stateV.rankOf();
|
|
sd::LongType initStateMRank = initStateM.rankOf();
|
|
sd::LongType stateMRank = stateM.rankOf();
|
|
sd::LongType initStateHRank = initStateH.rankOf();
|
|
sd::LongType stateHRank = stateH.rankOf();
|
|
|
|
sd::LongType *gradShape = shape::shapeOf(gradient.shapeInfo());
|
|
sd::LongType *gradStride = shape::stride(gradient.shapeInfo());
|
|
sd::LongType *updateStride = shape::stride(update.shapeInfo());
|
|
sd::LongType *initStateVStride = shape::stride(initStateV.shapeInfo());
|
|
sd::LongType *stateVStride = shape::stride(stateV.shapeInfo());
|
|
sd::LongType *initStateMStride = shape::stride(initStateM.shapeInfo());
|
|
sd::LongType *stateMStride = shape::stride(stateM.shapeInfo());
|
|
sd::LongType *initStateHStride = shape::stride(initStateH.shapeInfo());
|
|
sd::LongType *stateHStride = shape::stride(stateH.shapeInfo());
|
|
|
|
bool bXZsame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), update.shapeInfo());
|
|
bool bXInVSame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), initStateV.shapeInfo());
|
|
bool bXStVSame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), stateV.shapeInfo());
|
|
bool bXInMSame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), initStateM.shapeInfo());
|
|
bool bXStMSame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), stateM.shapeInfo());
|
|
bool bXInHSame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), initStateH.shapeInfo());
|
|
bool bXStHSame = shape::haveSameShapeAndStrides(gradient.shapeInfo(), stateH.shapeInfo());
|
|
|
|
auto func = PRAGMA_THREADS_FOR {
|
|
sd::LongType coords[SD_MAX_RANK];
|
|
for (sd::LongType i = start; i < stop; i++) {
|
|
INDEX2COORDS(i, gradRank, gradShape, coords);
|
|
|
|
sd::LongType xOffset;
|
|
COORDS2INDEX(gradRank, gradStride, coords, xOffset);
|
|
|
|
sd::LongType zOffset;
|
|
if (bXZsame) {
|
|
zOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(updateRank, updateStride, coords, zOffset);
|
|
}
|
|
|
|
sd::LongType initVOffset;
|
|
if (bXInVSame) {
|
|
initVOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(initStateVRank, initStateVStride, coords, initVOffset);
|
|
}
|
|
|
|
sd::LongType stVOffset;
|
|
if (bXStVSame) {
|
|
stVOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(stateVRank, stateVStride, coords, stVOffset);
|
|
}
|
|
|
|
sd::LongType initMOffset;
|
|
if (bXInMSame) {
|
|
initMOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(initStateMRank, initStateMStride, coords, initMOffset);
|
|
}
|
|
|
|
sd::LongType stMOffset;
|
|
if (bXStMSame) {
|
|
stMOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(stateMRank, stateMStride, coords, stMOffset);
|
|
}
|
|
|
|
sd::LongType initHOffset;
|
|
if (bXInHSame) {
|
|
initHOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(initStateHRank, initStateHStride, coords, initHOffset);
|
|
}
|
|
|
|
sd::LongType stHOffset;
|
|
if (bXStHSame) {
|
|
stHOffset = xOffset;
|
|
} else {
|
|
COORDS2INDEX(stateHRank, stateHStride, coords, stHOffset);
|
|
}
|
|
|
|
stM[stMOffset] = beta1 * initM[initMOffset] + grad[xOffset] * mbeta1;
|
|
stV[stVOffset] = beta2 * initV[initVOffset] + grad[xOffset] * grad[xOffset] * mbeta2;
|
|
stH[stHOffset] = sd::math::sd_max(initH[initHOffset], stV[stVOffset]);
|
|
|
|
up[zOffset] = epsilonT * stM[stMOffset] / (sd::math::sd_sqrt<T, T>(stH[stHOffset]) + epsilon);
|
|
}
|
|
};
|
|
|
|
samediff::Threads::parallel_for(func, 0, gradient.lengthOf(), 1);
|
|
return;
|
|
}
|
|
|
|
void updaterAmsGrad(sd::LaunchContext* context, NDArray& gradient, NDArray& initStateV,
|
|
NDArray& initStateM, NDArray& initStateH, NDArray& update, NDArray& stateV,
|
|
NDArray& stateM, NDArray& stateH, const double dLr, const double dBeta1, const double dBeta2,
|
|
const double dEpsilon, const int nIteration) {
|
|
BUILD_SINGLE_SELECTOR(gradient.dataType(), amsGradUpdater_,
|
|
(gradient, initStateV, initStateM, initStateH, update, stateV, stateM, stateH, dLr, dBeta1,
|
|
dBeta2, dEpsilon, nIteration),
|
|
SD_FLOAT_TYPES);
|
|
}
|
|
|
|
} // namespace helpers
|
|
} // namespace ops
|
|
} // namespace sd
|
|
#endif |