143 lines
7.2 KiB
Plaintext
143 lines
7.2 KiB
Plaintext
/*******************************************************************************
|
|
* Copyright (c) 2019 Konduit K.K.
|
|
*
|
|
* 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 Oleh Semeniv (oleg.semeniv@gmail.com)
|
|
//
|
|
|
|
#include <system/op_boilerplate.h>
|
|
#include <ops/declarable/helpers/updatersHelpers.h>
|
|
#include <helpers/PointersManager.h>
|
|
#include <math/platformmath.h>
|
|
#include <math/templatemath.h>
|
|
|
|
namespace sd {
|
|
namespace ops {
|
|
namespace helpers {
|
|
|
|
///////////////////////////////////////////////////////////////////
|
|
template<typename T>
|
|
__global__ void adaMaxUpdaterCuda(const void* vx, const Nd4jLong* xShapeInfo, const void* vinv, const Nd4jLong* invShapeInfo,
|
|
const void* vinm, const Nd4jLong* inmShapeInfo, void* vz, const Nd4jLong* zShapeInfo,
|
|
void* vstV, const Nd4jLong* stvShapeInfo, void* vstM, const Nd4jLong* stmShapeInfo,
|
|
const T lr, const T beta1, const T beta2, const T epsilon, const T iteration) {
|
|
|
|
const auto grad = reinterpret_cast<const T*>(vx);
|
|
const auto initU = reinterpret_cast<const T*>(vinv);
|
|
const auto initM = reinterpret_cast<const T*>(vinm);
|
|
|
|
auto up = reinterpret_cast<T*>(vz);
|
|
auto stU = reinterpret_cast<T*>(vstV);
|
|
auto stM = reinterpret_cast<T*>(vstM);
|
|
|
|
__shared__ Nd4jLong xLen;
|
|
__shared__ T beta1T, epsilonT;
|
|
__shared__ bool bEWS, bOrdering, bXZsame, bXInUSame, bXStUSame, bXInMSame, bXStMSame;
|
|
|
|
if (threadIdx.x == 0) {
|
|
xLen = shape::length(xShapeInfo);
|
|
beta1T = sd::math::nd4j_pow<T,T,T>(beta1, (iteration + 1) );
|
|
|
|
epsilonT = lr / (1.0 - beta1T);
|
|
if (sd::math::nd4j_isnan(epsilonT) || 0 == epsilonT || sd::math::nd4j_isinf(epsilonT))
|
|
epsilonT = epsilon;
|
|
|
|
bEWS = 1 == shape::elementWiseStride(xShapeInfo) && 1 == shape::elementWiseStride(zShapeInfo) &&
|
|
1 == shape::elementWiseStride(stmShapeInfo) && 1 == shape::elementWiseStride(inmShapeInfo) &&
|
|
1 == shape::elementWiseStride(stvShapeInfo) && 1 == shape::elementWiseStride(invShapeInfo);
|
|
bOrdering = shape::order(xShapeInfo) == shape::order(zShapeInfo) && shape::order(xShapeInfo) == shape::order(stmShapeInfo) &&
|
|
shape::order(xShapeInfo) == shape::order(inmShapeInfo) && shape::order(xShapeInfo) == shape::order(invShapeInfo) &&
|
|
shape::order(xShapeInfo) == shape::order(stvShapeInfo);
|
|
|
|
bXZsame = shape::haveSameShapeAndStrides(xShapeInfo, zShapeInfo);
|
|
bXInUSame = shape::haveSameShapeAndStrides(xShapeInfo, invShapeInfo);
|
|
bXStUSame = shape::haveSameShapeAndStrides(xShapeInfo, stvShapeInfo);
|
|
bXInMSame = shape::haveSameShapeAndStrides(xShapeInfo, inmShapeInfo);
|
|
bXStMSame = shape::haveSameShapeAndStrides(xShapeInfo, stmShapeInfo);
|
|
}
|
|
__syncthreads();
|
|
|
|
int coords[MAX_RANK];
|
|
|
|
for (Nd4jLong i = blockIdx.x * blockDim.x + threadIdx.x; i < xLen; i += gridDim.x * blockDim.x) {
|
|
|
|
|
|
auto xOffset = i, zOffset = i, initMOffset = i, initUOffset = i, stMOffset = i, stUOffset = i;
|
|
|
|
if (!bEWS || !bOrdering) {
|
|
|
|
shape::index2coords(i, xShapeInfo, coords);
|
|
xOffset = shape::getOffset(xShapeInfo, coords);
|
|
zOffset = bXZsame ? xOffset : shape::getOffset(zShapeInfo, coords);
|
|
initUOffset = bXInUSame ? xOffset : shape::getOffset(invShapeInfo, coords);
|
|
stUOffset = bXStUSame ? xOffset : shape::getOffset(stvShapeInfo, coords);
|
|
initMOffset = bXInMSame ? xOffset : shape::getOffset(inmShapeInfo, coords);
|
|
stMOffset = bXStMSame ? xOffset : shape::getOffset(stmShapeInfo, coords);
|
|
}
|
|
|
|
//m = B_1 * m + (1-B_1)*grad
|
|
stM[stMOffset] = beta1 * initM[initMOffset] + grad[xOffset] * (1 - beta1);
|
|
//u = max(B_2 * u, |grad|)
|
|
stU[stUOffset] = sd::math::nd4j_max( (beta2* initU[initUOffset]), sd::math::nd4j_abs(grad[xOffset])) + 1e-32;
|
|
|
|
up[zOffset] = (stM[stMOffset] * epsilonT) / stU[stUOffset];
|
|
}
|
|
}
|
|
|
|
///////////////////////////////////////////////////////////////////
|
|
template<typename T>
|
|
linkage void adaMaxUpdaterCudaLauncher(const int blocksPerGrid, const int threadsPerBlock, const cudaStream_t* stream, const void* vx, const Nd4jLong* xShapeInfo,
|
|
const void* vinv, const Nd4jLong* invShapeInfo, const void* vinm, const Nd4jLong* inmShapeInfo,
|
|
void* vz, const Nd4jLong* zShapeInfo, void* vstV, const Nd4jLong* stvShapeInfo,
|
|
void* vstM, const Nd4jLong* stmShapeInfo, const double dLr,
|
|
const double dBeta1, const double dBeta2, const double dEpsilon, const int nIteration) {
|
|
|
|
const T lr = static_cast<T>(dLr);
|
|
const T beta1 = static_cast<T>(dBeta1);
|
|
const T beta2 = static_cast<T>(dBeta2);
|
|
const T epsilon = static_cast<T>(dEpsilon);
|
|
const T iteration = static_cast<T>(nIteration);
|
|
|
|
adaMaxUpdaterCuda<T> << <blocksPerGrid, threadsPerBlock, 256, * stream >> > (vx, xShapeInfo, vinv, invShapeInfo, vinm, inmShapeInfo, vz,
|
|
zShapeInfo, vstV, stvShapeInfo, vstM, stmShapeInfo, lr, beta1, beta2, epsilon, iteration);
|
|
}
|
|
|
|
///////////////////////////////////////////////////////////////////
|
|
void updaterAdaMax(sd::LaunchContext* context, const NDArray& gradient, const NDArray& initStateU, const NDArray& initStateM,
|
|
NDArray& update, NDArray& stateU, NDArray& stateM, const double dLr, const double dBeta1,
|
|
const double dBeta2, const double dEpsilon, const int nIteration) {
|
|
|
|
PointersManager manager(context, "adaMaxUpdater");
|
|
|
|
const int threadsPerBlock = MAX_NUM_THREADS / 4;
|
|
const int blocksPerGrid = (gradient.lengthOf() + threadsPerBlock - 1) / threadsPerBlock;
|
|
|
|
NDArray::prepareSpecialUse({ &update, &stateU, &stateM }, { &gradient, &initStateU, &initStateM });
|
|
BUILD_SINGLE_SELECTOR(gradient.dataType(), adaMaxUpdaterCudaLauncher, (blocksPerGrid, threadsPerBlock, context->getCudaStream(),
|
|
gradient.getSpecialBuffer(), gradient.getSpecialShapeInfo(), initStateU.getSpecialBuffer(),
|
|
initStateU.getSpecialShapeInfo(), initStateM.getSpecialBuffer(), initStateM.getSpecialShapeInfo(),
|
|
update.getSpecialBuffer(), update.getSpecialShapeInfo(), stateU.getSpecialBuffer(),
|
|
stateU.getSpecialShapeInfo(), stateM.getSpecialBuffer(), stateM.getSpecialShapeInfo(),
|
|
dLr, dBeta1, dBeta2, dEpsilon, nIteration ), FLOAT_TYPES);
|
|
NDArray::registerSpecialUse({ &update, &stateU, &stateM }, { &gradient, &initStateU, &initStateM });
|
|
|
|
manager.synchronize();
|
|
}
|
|
|
|
}
|
|
}
|
|
}
|