m53297601
/
Paddle
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

follow comments

Browse Source
feature/design_of_v2_layer_converter
liaogang 8 years ago
parent 8cde2d119f
commit f27fd9dc28
46 changed files with 222 additions and 251 deletions
Show Stats Download Patch File Download Diff File

3
cmake/cblas.cmake
Unescape View File

@ -44,7 +44,6 @@ if(MKL_INC_DIR AND MKL_CORE_LIB AND MKL_SEQUENTIAL_LIB AND MKL_INTEL_LP64)
message(STATUS "Found MKL (include: ${CBLAS_INC_DIR}, library: ${CBLAS_LIBRARIES})")
set(CBLAS_FOUND ON)
if(${MKL_LAPACK_INC_DIR})
add_definitions(-DPADDLE_USE_LAPACK)
message(STATUS "Found lapack in MKL (include: ${MKL_LAPACK_INC_DIR})")
endif()
return() # return file.
@ -80,7 +79,6 @@ if(ATLAS_INC_DIR AND ATLAS_CBLAS_LIB AND ATLAS_LIB AND NOT CBLAS_FOUND)
message(STATUS "Found ATLAS (include: ${CBLAS_INC_DIR}, library: ${CBLAS_LIBRARIES})")
set(CBLAS_FOUND ON)
if(ATLAS_CLAPACK_INC_DIR)
add_definitions(-DPADDLE_USE_LAPACK)
message(STATUS "Found lapack in ATLAS (include: ${ATLAS_CLAPACK_INC_DIR})")
endif()
return()
@ -114,7 +112,6 @@ if(OPENBLAS_INC_DIR AND OPENBLAS_LIB)
message(STATUS "Found OpenBLAS (include: ${CBLAS_INC_DIR}, library: ${CBLAS_LIBRARIES})")
set(CBLAS_FOUND ON)
if(OPENBLAS_LAPACKE_INC_DIR)
add_definitions(-DPADDLE_USE_LAPACK)
message(STATUS "Found lapack in OpenBLAS (include: ${OPENBLAS_LAPACKE_INC_DIR})")
endif()
return()

2
cmake/external/openblas.cmake vendored
Unescape View File

@ -27,8 +27,6 @@ IF(NOT ${CBLAS_FOUND})
SET(CBLAS_LIBRARIES "${CBLAS_INSTALL_DIR}/lib/libopenblas.a" CACHE FILEPATH "openblas library" FORCE)
ENDIF(WIN32)
ADD_DEFINITIONS(-DPADDLE_USE_LAPACK)
ExternalProject_Add(
openblas
${EXTERNAL_PROJECT_LOG_ARGS}

6
paddle/cuda/include/hl_activation_functions.h
Unescape View File

@ -40,18 +40,18 @@ public:
namespace gpu {
static __device__ Active<real>::forward forward[] = HPPL_ACTIVE_FUNCTION;
static __device__ Active<real>::backward backward[] = HPPL_ACTIVE_FUNCTION;
}
} // namespace gpu
#else
namespace cpu {
static Active<real>::forward forward[] = HPPL_ACTIVE_FUNCTION;
static Active<real>::backward backward[] = HPPL_ACTIVE_FUNCTION;
}
} // namespace cpu
#ifdef __AVX__
namespace avx {
static Active<__m256>::forward forward[] = HPPL_ACTIVE_FUNCTION;
static Active<__m256>::backward backward[] = HPPL_ACTIVE_FUNCTION;
}
} // namespace avx
#endif
#endif

34
paddle/cuda/include/hl_cnn.h
Unescape View File

@ -273,23 +273,23 @@ extern void hl_bilinear_forward(const real* inData,
const real ratioW);
/**
* @brief Bilinear interpolation backward.
*
* @param[out] inGrad input gradient.
* @param[in] inImgH input image height.
* @param[in] inImgW input image width.
* @param[in] inputH input batchSize.
* @param[in] inputW input image data dim.
* @param[in] outGrad output gradient.
* @param[in] outImgH output image height.
* @param[in] outImgW output image width.
* @param[in] outputH output batchSize.
* @param[in] outputW output image data dim.
* @param[in] numChannels number of channels.
* @param[in] ratioH inImgH / outImgH.
* @param[in] ratioW inImgW / outImgW.
*
*/
* @brief Bilinear interpolation backward.
*
* @param[out] inGrad input gradient.
* @param[in] inImgH input image height.
* @param[in] inImgW input image width.
* @param[in] inputH input batchSize.
* @param[in] inputW input image data dim.
* @param[in] outGrad output gradient.
* @param[in] outImgH output image height.
* @param[in] outImgW output image width.
* @param[in] outputH output batchSize.
* @param[in] outputW output image data dim.
* @param[in] numChannels number of channels.
* @param[in] ratioH inImgH / outImgH.
* @param[in] ratioW inImgW / outImgW.
*
*/
extern void hl_bilinear_backward(real* inGrad,
const size_t inImgH,
const size_t inImgW,

2
paddle/cuda/src/hl_cuda_cublas.cc
Unescape View File

@ -16,7 +16,7 @@ limitations under the License. */
#include <sys/time.h>
#include "hl_cuda.h"
#include "hl_thread.ph"
#include "paddle/utils/DynamicLoad.h"
#include "paddle/utils/DynamicLoader.h"
#include "paddle/utils/Logging.h"
namespace dynload {

2
paddle/cuda/src/hl_cuda_cudnn.cc
Unescape View File

@ -17,7 +17,7 @@ limitations under the License. */
#include <gflags/gflags.h>
#include "hl_cuda_cudnn.ph"
#include "hl_thread.ph"
#include "paddle/utils/DynamicLoad.h"
#include "paddle/utils/DynamicLoader.h"
#include "paddle/utils/Logging.h"
DEFINE_int32(cudnn_conv_workspace_limit_in_mb,

2
paddle/cuda/src/hl_cuda_device.cc
Unescape View File

@ -24,7 +24,7 @@ limitations under the License. */
#include "hl_cuda.ph"
#include "hl_thread.ph"
#include "paddle/utils/Logging.h"
#include "paddle/utils/DynamicLoad.h"
#include "paddle/utils/DynamicLoader.h"
// clang-format on
namespace dynload {

2
paddle/cuda/src/hl_warpctc_wrap.cc
Unescape View File

@ -14,7 +14,7 @@ limitations under the License. */
#include "hl_warpctc_wrap.h"
#include <mutex>
#include "paddle/utils/DynamicLoad.h"
#include "paddle/utils/DynamicLoader.h"
#include "paddle/utils/Logging.h"
namespace dynload {

2
paddle/function/BufferArgTest.cpp
Unescape View File

@ -12,8 +12,8 @@ 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. */
#include "BufferArg.h"
#include <gtest/gtest.h>
#include "BufferArg.h"
#include "paddle/math/MemoryHandle.h"
namespace paddle {

12
paddle/function/CosSimOp.cpp
Unescape View File

@ -165,12 +165,12 @@ void CosSimBackward<DEVICE_TYPE_CPU>(const CpuMatrix& out_grad,
real reciprocal_square_sum_x = 1.0f / square_sum_x;
real reciprocal_square_sum_y = 1.0f / square_sum_y;
for (size_t j = 0; j < dim; ++j) {
prev_grad_x[j] +=
out[i] * grad[i] * (prev_out_y[j] * reciprocal_xy -
prev_out_x[j] * reciprocal_square_sum_x);
prev_grad_y[j] +=
out[i] * grad[i] * (prev_out_x[j] * reciprocal_xy -
prev_out_y[j] * reciprocal_square_sum_y);
prev_grad_x[j] += out[i] * grad[i] *
(prev_out_y[j] * reciprocal_xy -
prev_out_x[j] * reciprocal_square_sum_x);
prev_grad_y[j] += out[i] * grad[i] *
(prev_out_x[j] * reciprocal_xy -
prev_out_y[j] * reciprocal_square_sum_y);
}
}
}

2
paddle/function/FunctionTest.cpp
Unescape View File

@ -12,8 +12,8 @@ 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. */
#include "Function.h"
#include <gtest/gtest.h>
#include "Function.h"
#include "paddle/math/SparseMatrix.h"
namespace paddle {

18
paddle/function/MulOpTest.cpp
Unescape View File

@ -74,9 +74,9 @@ TEST(MulOp, DDDMatrixMul) {
}
/**
* C += A * B, B, C dense, A sparse
* dense = sparse * dense
*/
* C += A * B, B, C dense, A sparse
* dense = sparse * dense
*/
void testFuncDSparseDMatrix(
size_t dimM, size_t dimN, size_t dimK, size_t nnz, SparseFormat FORMAT) {
real scaleT = 1.0;
@ -119,9 +119,9 @@ TEST(MuLOp, DSparseDMul) {
}
/**
* C += A * B, A, C dense, B sparse
* dense = dense * sparse
*/
* C += A * B, A, C dense, B sparse
* dense = dense * sparse
*/
void testFuncDDSparseMatrix(
size_t dimM, size_t dimN, size_t dimK, size_t nnz, SparseFormat FORMAT) {
real scaleT = 1.0;
@ -165,9 +165,9 @@ TEST(MulOp, DDSparseMul) {
}
/**
* C += A * B, A sparse, B, C dense
* sparse = dense * dense
*/
* C += A * B, A sparse, B, C dense
* sparse = dense * dense
*/
void testFuncSparseDDMatrix(
size_t dimM, size_t dimN, size_t dimK, size_t nnz, SparseFormat FORMAT) {
real scaleT = 1.0;

2
paddle/function/TensorShapeTest.cpp
Unescape View File

@ -12,8 +12,8 @@ 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. */
#include "TensorShape.h"
#include <gtest/gtest.h>
#include "TensorShape.h"
namespace paddle {

2
paddle/function/TensorTypeTest.cpp
Unescape View File

@ -12,8 +12,8 @@ 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. */
#include "TensorType.h"
#include <gtest/gtest.h>
#include "TensorType.h"
namespace paddle {

4
paddle/gserver/dataproviders/PyDataProvider.cpp
Unescape View File

@ -194,8 +194,8 @@ void PyDataProvider::fillSlotsByStr(const std::string& samples) {
auto& slot = slots_[j];
CHECK(SlotDef::INDEX >= slot.type || SlotDef::STRING == slot.type)
<< " Slot type:" << slot.type << " is out of range.";
CHECK_GE(slot.type, SlotDef::VECTOR_DENSE) << " Slot type:" << slot.type
<< " is out of range.";
CHECK_GE(slot.type, SlotDef::VECTOR_DENSE)
<< " Slot type:" << slot.type << " is out of range.";
switch (slot.type) {
case SlotDef::VECTOR_DENSE:
fillDenseSlot(slot, data, dataEnd);

6
paddle/gserver/evaluators/Evaluator.cpp
Unescape View File

@ -446,9 +446,9 @@ real AucEvaluator::evalImp(std::vector<Argument>& arguments) {
for (size_t i = 0; i < insNum; ++i) {
real value = outputD[pos];
uint32_t binIdx = static_cast<uint32_t>(value * kBinNum_);
CHECK(binIdx <= kBinNum_) << "bin index [" << binIdx
<< "] out of range, predict value[" << value
<< "]";
CHECK(binIdx <= kBinNum_)
<< "bin index [" << binIdx << "] out of range, predict value[" << value
<< "]";
real w = supportWeight ? weightD[i] : 1.0;
if (labelD[i] == kNegativeLabel_) {
statNeg_[binIdx] += w;

1
paddle/gserver/gradientmachines/GradientMachine.cpp
Unescape View File

@ -21,7 +21,6 @@ limitations under the License. */
#include "MultiGradientMachine.h"
#include "MultiNetwork.h"
#include "NeuralNetwork.h"
#include "NeuralNetwork.h"
#include "ParallelNeuralNetwork.h"
#include "hl_gpu.h"

2
paddle/gserver/gradientmachines/RecurrentGradientMachine.cpp
Unescape View File

@ -637,7 +637,7 @@ void RecurrentGradientMachine::removeBeamSearchStatisticsCallbacks() {
/* create scattered id infomation for all realLayer of inFrameLines one time.
* If hasSubseq, will also create scattered sequenceStartPositions infomation
* for all realLayer of inFrameLines one time.
*/
*/
void RecurrentGradientMachine::createInFrameInfo(int inlinkId,
const Argument& input,

30
paddle/gserver/gradientmachines/RecurrentGradientMachine.h
Unescape View File

@ -107,18 +107,18 @@ public:
DropCallback;
/**
* @brief NormOrDropNodeCallback
*
* Normalize a path's probabilities or just drop it by modifying path.logProb
*
* The first parameter is sequence index in a batch
*
* The second parameter is path.ids
*
* The third parameter is probabilites for each node in this path.
*
* The fourth parameter is the probability of the whole path.
*/
* @brief NormOrDropNodeCallback
*
* Normalize a path's probabilities or just drop it by modifying path.logProb
*
* The first parameter is sequence index in a batch
*
* The second parameter is path.ids
*
* The third parameter is probabilites for each node in this path.
*
* The fourth parameter is the probability of the whole path.
*/
typedef std::function<void(
int seqId, const std::vector<int>&, std::vector<real>&, real*)>
NormOrDropNodeCallback;
@ -348,9 +348,9 @@ protected:
int targetInfoInlinkId_;
/* create scattered id infomation for all realLayer of inFrameLines one time.
* If hasSubseq, will also create scattered sequenceStartPositions infomation
* for all realLayer of inFrameLines one time.
*/
* If hasSubseq, will also create scattered sequenceStartPositions infomation
* for all realLayer of inFrameLines one time.
*/
void createInFrameInfo(int inlinks_id,
const Argument& input,
PassType passType);

5
paddle/gserver/layers/Layer.cpp
Unescape View File

@ -263,8 +263,9 @@ void Layer::zeroGrad() {
}
void Layer::initNeedFlags() {
auto initFlag = [this](
bool& flag, bool (Layer::*flagQueryFunc)() const, ParameterType type) {
auto initFlag = [this](bool& flag,
bool (Layer::*flagQueryFunc)() const,
ParameterType type) {
flag = false;
if (biasParameter_ && biasParameter_->hasType(type)) {
flag = true;

12
paddle/gserver/layers/Layer.h
Unescape View File

@ -106,9 +106,9 @@ protected:
public:
/**
* Wait until all input value ready.
* Called before Layer::forward() function.
*/
* Wait until all input value ready.
* Called before Layer::forward() function.
*/
virtual void waitInputValue();
/**
@ -118,9 +118,9 @@ public:
virtual void copyOutputToOtherDevice();
/**
* Wait until all output grad ready and merge them to output_.grad.
* Called before Layer::backward() function.
*/
* Wait until all output grad ready and merge them to output_.grad.
* Called before Layer::backward() function.
*/
virtual void waitAndMergeOutputGrad();
/**

2
paddle/gserver/layers/RotateLayer.h
Unescape View File

@ -29,7 +29,7 @@ namespace paddle {
*
* The config file api is rotate_layer
*
*/
*/
class RotateLayer : public Layer {
public:

2
paddle/gserver/layers/SequencePoolLayer.cpp
Unescape View File

@ -60,7 +60,7 @@ void SequencePoolLayer::forward(PassType passType) {
* thus, in this case, output_ has no sequenceStartPositions.
* If type_ = kSeq, seq has sub-seq degrades to a seq, thus, only in this
* case, we should compute the new sequenceStartPositions.
*/
*/
if (type_) {
CHECK(input.subSequenceStartPositions)
<< "when trans_type = seq, input must hasSubseq";

39
paddle/gserver/tests/test_RecurrentLayer.cpp
Unescape View File

@ -292,26 +292,27 @@ void checkRecurrentLayer(LayerConfig layerConfig,
TestRecurrentLayer<T> testGpu(layerConfig, true, gpuBatch);
testCpu.init(batchSize);
testGpu.init(batchSize);
auto checkError = [](
MatrixPtr cpu, MatrixPtr gpu, int numSequences, const char* str) {
CpuMatrix check(gpu->getHeight(), gpu->getWidth());
check.copyFrom(*gpu);
int height = cpu->getHeight();
int width = cpu->getWidth();
const real* data1 = cpu->getData();
const real* data2 = check.getData();
int count = 0;
for (int i = 0; i < height; i++) {
for (int j = 0; j < width; j++) {
if (fabs(data1[i * width + j] - data2[i * width + j]) / numSequences >
1e-4) {
count++;
auto checkError =
[](MatrixPtr cpu, MatrixPtr gpu, int numSequences, const char* str) {
CpuMatrix check(gpu->getHeight(), gpu->getWidth());
check.copyFrom(*gpu);
int height = cpu->getHeight();
int width = cpu->getWidth();
const real* data1 = cpu->getData();
const real* data2 = check.getData();
int count = 0;
for (int i = 0; i < height; i++) {
for (int j = 0; j < width; j++) {
if (fabs(data1[i * width + j] - data2[i * width + j]) /
numSequences >
1e-4) {
count++;
}
}
}
}
}
EXPECT_EQ(count, 0) << "[" << str << "]"
<< "There are " << count << " different element.";
};
EXPECT_EQ(count, 0) << "[" << str << "]"
<< "There are " << count << " different element.";
};
T* cpuLayer = dynamic_cast<T*>(testCpu.testLayer_.get());
T* gpuLayer = dynamic_cast<T*>(testGpu.testLayer_.get());

78
paddle/math/MathFunctions.cpp
Unescape View File

@ -15,7 +15,7 @@ limitations under the License. */
#include "MathFunctions.h"
#include "hl_matrix_apply.cuh"
#include "hl_matrix_ops.cuh"
#include "paddle/utils/DynamicLoad.h"
#include "paddle/utils/DynamicLoader.h"
namespace dynload {
@ -32,7 +32,7 @@ void* lapack_dso_handle = nullptr;
#define DYNAMIC_LOAD_LAPACK_WRAP(__name) \
struct DynLoad__##__name { \
template <typename... Args> \
auto operator()(Args... args)->decltype(__name(args...)) { \
auto operator()(Args... args) -> decltype(__name(args...)) { \
using lapack_func = decltype(__name(args...)) (*)(Args...); \
std::call_once(lapack_dso_flag, GetLapackDsoHandle, &lapack_dso_handle); \
void* p_##__name = dlsym(lapack_dso_handle, #__name); \
@ -41,24 +41,27 @@ void* lapack_dso_handle = nullptr;
} __name; // struct DynLoad__##__name
// clang-format off
#ifdef PADDLE_USE_LAPACK
#ifdef PADDLE_USE_ATLAS
#define LAPACK_ROUTINE_EACH(__macro) \
__macro(clapack_sgetrf) \
__macro(clapack_dgetrf) \
__macro(clapack_sgetri) \
__macro(clapack_dgetri)
#define PADDLE_SGETRF clapack_sgetrf
#define PADDLE_DGETRF clapack_dgetrf
#define PADDLE_SGETRI clapack_sgetri
#define PADDLE_DGETRI clapack_dgetri
#else
#define LAPACK_ROUTINE_EACH(__macro) \
__macro(LAPACKE_sgetrf) \
__macro(LAPACKE_dgetrf) \
__macro(LAPACKE_sgetri) \
__macro(LAPACKE_dgetri)
#endif
#define PADDLE_SGETRF LAPACKE_sgetrf
#define PADDLE_DGETRF LAPACKE_dgetrf
#define PADDLE_SGETRI LAPACKE_sgetri
#define PADDLE_DGETRI LAPACKE_dgetri
#endif
#define LAPACK_ROUTINE_EACH(__macro) \
__macro(PADDLE_SGETRF) \
__macro(PADDLE_DGETRF) \
__macro(PADDLE_SGETRI) \
__macro(PADDLE_DGETRI)
// clang-format on
LAPACK_ROUTINE_EACH(DYNAMIC_LOAD_LAPACK_WRAP)
#endif
// clang-format on
} // namespace dynload
namespace paddle {
@ -130,16 +133,7 @@ int getrf<float>(const CBLAS_ORDER order,
float* A,
const int lda,
int* ipiv) {
#ifdef PADDLE_USE_LAPACK
#ifdef PADDLE_USE_ATLAS
return dynload::clapack_sgetrf(order, M, N, A, lda, ipiv);
#else
return dynload::LAPACKE_sgetrf(order, M, N, A, lda, ipiv);
#endif
#else
LOG(FATAL) << "Not implemented";
#endif
return 0;
return dynload::PADDLE_SGETRF(order, M, N, A, lda, ipiv);
}
template <>
@ -149,16 +143,7 @@ int getrf<double>(const CBLAS_ORDER order,
double* A,
const int lda,
int* ipiv) {
#ifdef PADDLE_USE_LAPACK
#ifdef PADDLE_USE_ATLAS
return dynload::clapack_dgetrf(order, M, N, A, lda, ipiv);
#else
return dynload::LAPACKE_dgetrf(order, M, N, A, lda, ipiv);
#endif
#else
LOG(FATAL) << "Not implemented";
#endif
return 0;
return dynload::PADDLE_DGETRF(order, M, N, A, lda, ipiv);
}
template <>
@ -167,16 +152,7 @@ int getri<float>(const CBLAS_ORDER order,
float* A,
const int lda,
const int* ipiv) {
#ifdef PADDLE_USE_LAPACK
#ifdef PADDLE_USE_ATLAS
return dynload::clapack_sgetri(order, N, A, lda, ipiv);
#else
return dynload::LAPACKE_sgetri(order, N, A, lda, ipiv);
#endif
#else
LOG(FATAL) << "Not implemented";
#endif
return 0;
return dynload::PADDLE_SGETRI(order, N, A, lda, ipiv);
}
template <>
@ -185,15 +161,7 @@ int getri<double>(const CBLAS_ORDER order,
double* A,
const int lda,
const int* ipiv) {
#ifdef PADDLE_USE_LAPACK
#ifdef PADDLE_USE_ATLAS
return dynload::clapack_dgetri(order, N, A, lda, ipiv);
#else
return dynload::LAPACKE_dgetri(order, N, A, lda, ipiv);
#endif
#else
LOG(FATAL) << "Not implemented";
#endif
return dynload::PADDLE_DGETRI(order, N, A, lda, ipiv);
return 0;
}

Some files were not shown because too many files have changed in this diff Show More

Write Preview
Loading…
Cancel
Save