Skip to content

P
Paddle

BaiXuePrincess / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
“273598ef3498158fa0d0c09bace7e2a8af4e19d8”上不存在“examples/C++/PaddleClas/imagenet/resnet50_rpc_client.py”
提交 b52039bd 编写于 作者: H hedaoyuan 提交者： qingqing01
浏览文件
操作

some bug fix for sparse matrix (#133) 

* some bug fix for sparse matrix

* a minor bug fix
上级 0276f15a
显示空白变更内容
内联 并排
Showing with 98 addition and 88 deletion
paddle/cuda/include/hl_sparse.h
浏览文件 @ b52039bd
...@@ -223,6 +223,7 @@ extern void hl_matrix_csc2dense(hl_sparse_matrix_s A_d, ...@@ -223,6 +223,7 @@ extern void hl_matrix_csc2dense(hl_sparse_matrix_s A_d,
* @param[in] dimK width of op(A) & height of op(B) * @param[in] dimK width of op(A) & height of op(B)
* @param[in] alpha scalar used for multiplication. * @param[in] alpha scalar used for multiplication.
* @param[in] beta scalar used for multiplication. * @param[in] beta scalar used for multiplication.
* If beta is zero, C does not have to be a valid input.
* *
* @note transb is not support HPPL_OP_T. * @note transb is not support HPPL_OP_T.
* *
...@@ -251,6 +252,7 @@ extern void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, ...@@ -251,6 +252,7 @@ extern void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d,
* @param[in] dimK width of op(A) & height of op(B) * @param[in] dimK width of op(A) & height of op(B)
* @param[in] alpha scalar used for multiplication. * @param[in] alpha scalar used for multiplication.
* @param[in] beta scalar used for multiplication. * @param[in] beta scalar used for multiplication.
* If beta is zero, C does not have to be a valid input.
* *
* @note transb is not support HPPL_OP_T. * @note transb is not support HPPL_OP_T.
* *
...@@ -275,6 +277,7 @@ extern void hl_matrix_csc_mul_dense(hl_sparse_matrix_s A_d, ...@@ -275,6 +277,7 @@ extern void hl_matrix_csc_mul_dense(hl_sparse_matrix_s A_d,
* @param[in] dimK width of op(A) & height of op(B) * @param[in] dimK width of op(A) & height of op(B)
* @param[in] alpha scalar used for multiplication. * @param[in] alpha scalar used for multiplication.
* @param[in] beta scalar used for multiplication. * @param[in] beta scalar used for multiplication.
* If beta is zero, C does not have to be a valid input.
* *
* @note transa is not support HPPL_OP_T. * @note transa is not support HPPL_OP_T.
* *
...@@ -327,6 +330,7 @@ extern void hl_sparse_matrix_mul(real* A_d, hl_trans_op_t transa, ...@@ -327,6 +330,7 @@ extern void hl_sparse_matrix_mul(real* A_d, hl_trans_op_t transa,
* @param[in] dimK width of op(A) & height of op(B) * @param[in] dimK width of op(A) & height of op(B)
* @param[in] alpha scalar used for multiplication. * @param[in] alpha scalar used for multiplication.
* @param[in] beta scalar used for multiplication. * @param[in] beta scalar used for multiplication.
* If beta is zero, C does not have to be a valid input.
* *
* *
* @note transa is not support HPPL_OP_T. * @note transa is not support HPPL_OP_T.
......
paddle/cuda/src/hl_cuda_sparse.cu
浏览文件 @ b52039bd
...@@ -562,6 +562,22 @@ void hl_memcpy_sparse_matrix(hl_sparse_matrix_s dst, ...@@ -562,6 +562,22 @@ void hl_memcpy_sparse_matrix(hl_sparse_matrix_s dst,
} }
} }
/**
* Calculate beta * C, if beta is zero, C does not have to be a valid input.
*/
static void _beta_mul_c(real *c, int dimM, int dimN, real beta) {
if (beta == 0.0) {
hl_gpu_apply_unary_op(unary::Zero<real>(), c, dimM, dimN, dimN);
} else {
if (beta != 1.0){
hl_gpu_apply_unary_op(
unary::mul_scalar<real>(beta), c, dimM, dimN, dimN);
}
}
return;
}
void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa, void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa,
real *B_d, hl_trans_op_t transb, real *B_d, hl_trans_op_t transb,
real *C_d, real *C_d,
...@@ -580,16 +596,9 @@ void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa, ...@@ -580,16 +596,9 @@ void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa,
} }
if (A_d->nnz == 0) { if (A_d->nnz == 0) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
} else {
return; return;
} }
}
/* nnz != 0 */ /* nnz != 0 */
hl_csr_matrix A_d2 = (hl_csr_matrix)(A_d->matrix); hl_csr_matrix A_d2 = (hl_csr_matrix)(A_d->matrix);
...@@ -633,13 +642,7 @@ void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa, ...@@ -633,13 +642,7 @@ void hl_matrix_csr_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa,
beta); beta);
} }
} else if (HPPL_OP_T == transa) { } else if (HPPL_OP_T == transa) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
}
int blocksX = (dimN + CU_CSC_MUL_DENSE_BLOCK_N - 1) / int blocksX = (dimN + CU_CSC_MUL_DENSE_BLOCK_N - 1) /
CU_CSC_MUL_DENSE_BLOCK_N; CU_CSC_MUL_DENSE_BLOCK_N;
...@@ -699,16 +702,9 @@ void hl_matrix_dense_mul_csc(real *A_d, hl_trans_op_t transa, ...@@ -699,16 +702,9 @@ void hl_matrix_dense_mul_csc(real *A_d, hl_trans_op_t transa,
<< "matrix format error!"; << "matrix format error!";
if (B_d->nnz == 0) { if (B_d->nnz == 0) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
} else {
return; return;
} }
}
/* nnz != 0 */ /* nnz != 0 */
hl_csc_matrix B_d2 = (hl_csc_matrix)(B_d->matrix); hl_csc_matrix B_d2 = (hl_csc_matrix)(B_d->matrix);
...@@ -750,13 +746,7 @@ void hl_matrix_dense_mul_csc(real *A_d, hl_trans_op_t transa, ...@@ -750,13 +746,7 @@ void hl_matrix_dense_mul_csc(real *A_d, hl_trans_op_t transa,
beta); beta);
} }
} else if (transb == HPPL_OP_T) { } else if (transb == HPPL_OP_T) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
}
int blocksX = 1 + (dimK-1)/CU_DM_CSR_THREAD_X; int blocksX = 1 + (dimK-1)/CU_DM_CSR_THREAD_X;
int blocksY = 1 + (dimM-1)/CU_DM_CSR_BLOCK_M; int blocksY = 1 + (dimM-1)/CU_DM_CSR_BLOCK_M;
dim3 threads(CU_DM_CSR_THREAD_X, CU_DM_CSR_THREAD_Y); dim3 threads(CU_DM_CSR_THREAD_X, CU_DM_CSR_THREAD_Y);
...@@ -813,16 +803,9 @@ void hl_matrix_dense_mul_csr(real *A_d, hl_trans_op_t transa, ...@@ -813,16 +803,9 @@ void hl_matrix_dense_mul_csr(real *A_d, hl_trans_op_t transa,
<< "matrix format error!"; << "matrix format error!";
if (B_d->nnz == 0) { if (B_d->nnz == 0) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
} else {
return; return;
} }
}
/* nnz != 0 */ /* nnz != 0 */
hl_csr_matrix B_d2 = (hl_csr_matrix)(B_d->matrix); hl_csr_matrix B_d2 = (hl_csr_matrix)(B_d->matrix);
...@@ -833,14 +816,7 @@ void hl_matrix_dense_mul_csr(real *A_d, hl_trans_op_t transa, ...@@ -833,14 +816,7 @@ void hl_matrix_dense_mul_csr(real *A_d, hl_trans_op_t transa,
} }
if (transb == HPPL_OP_N) { if (transb == HPPL_OP_N) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
}
int blocksX = 1 + (dimK-1)/CU_DM_CSR_THREAD_X; int blocksX = 1 + (dimK-1)/CU_DM_CSR_THREAD_X;
int blocksY = 1 + (dimM-1)/CU_DM_CSR_BLOCK_M; int blocksY = 1 + (dimM-1)/CU_DM_CSR_BLOCK_M;
dim3 threads(CU_DM_CSR_THREAD_X, CU_DM_CSR_THREAD_Y); dim3 threads(CU_DM_CSR_THREAD_X, CU_DM_CSR_THREAD_Y);
...@@ -925,16 +901,9 @@ void hl_matrix_csc_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa, ...@@ -925,16 +901,9 @@ void hl_matrix_csc_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa,
} }
if (A_d->nnz == 0) { if (A_d->nnz == 0) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
} else {
return; return;
} }
}
/* nnz != 0 */ /* nnz != 0 */
hl_csc_matrix A_d2 = (hl_csc_matrix)(A_d->matrix); hl_csc_matrix A_d2 = (hl_csc_matrix)(A_d->matrix);
...@@ -945,13 +914,7 @@ void hl_matrix_csc_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa, ...@@ -945,13 +914,7 @@ void hl_matrix_csc_mul_dense(hl_sparse_matrix_s A_d, hl_trans_op_t transa,
} }
if (HPPL_OP_N == transa) { if (HPPL_OP_N == transa) {
if (beta != 1.0) { _beta_mul_c(C_d, dimM, dimN, beta);
hl_gpu_apply_unary_op(unary::mul_scalar<real>(beta),
C_d,
dimM,
dimN,
dimN);
}
int blocksX = (dimN + CU_CSC_MUL_DENSE_BLOCK_N -1)/CU_CSC_MUL_DENSE_BLOCK_N; int blocksX = (dimN + CU_CSC_MUL_DENSE_BLOCK_N -1)/CU_CSC_MUL_DENSE_BLOCK_N;
int blocksY = (dimK + CU_CSC_MUL_DENSE_BLOCK_K -1)/CU_CSC_MUL_DENSE_BLOCK_K; int blocksY = (dimK + CU_CSC_MUL_DENSE_BLOCK_K -1)/CU_CSC_MUL_DENSE_BLOCK_K;
...@@ -1113,7 +1076,7 @@ void hl_sparse_matrix_mul(real *A_d, hl_trans_op_t transa, ...@@ -1113,7 +1076,7 @@ void hl_sparse_matrix_mul(real *A_d, hl_trans_op_t transa,
CHECK(!transA) << "Not supported A is trans and B is not trans!"; CHECK(!transA) << "Not supported A is trans and B is not trans!";
dim3 block(CU_BLOCK_SIZE, 1); dim3 block(CU_BLOCK_SIZE, 1);
int avgNnzPerRow = C_d2->nnz_s / dimM; int avgNnzPerRow = C_d->nnz / dimM;
avgNnzPerRow = avgNnzPerRow > 0 ? avgNnzPerRow : 1; avgNnzPerRow = avgNnzPerRow > 0 ? avgNnzPerRow : 1;
int gridx = DIVUP(avgNnzPerRow, CU_BLOCK_SIZE); int gridx = DIVUP(avgNnzPerRow, CU_BLOCK_SIZE);
dim3 grid(gridx, dimM); dim3 grid(gridx, dimM);
...@@ -1242,9 +1205,9 @@ void hl_matrix_csr_column_sum(real* A_d, hl_sparse_matrix_s B_d, ...@@ -1242,9 +1205,9 @@ void hl_matrix_csr_column_sum(real* A_d, hl_sparse_matrix_s B_d,
LOG(FATAL) << "parameter B is null!"; LOG(FATAL) << "parameter B is null!";
} }
if (B_d2->nnz_s == 0) return; if (B_d->nnz == 0) return;
int nnz = B_d2->nnz_s; int nnz = B_d->nnz;
int block = 512; int block = 512;
int grid = DIVUP(nnz, 512); int grid = DIVUP(nnz, 512);
KeSMatrixCsrColumnSum<<<grid, block, 0, STREAM_DEFAULT>>>( KeSMatrixCsrColumnSum<<<grid, block, 0, STREAM_DEFAULT>>>(
...@@ -1273,9 +1236,9 @@ void hl_matrix_csr_add_bias(hl_sparse_matrix_s A_d, real* B_d, ...@@ -1273,9 +1236,9 @@ void hl_matrix_csr_add_bias(hl_sparse_matrix_s A_d, real* B_d,
LOG(FATAL) << "parameter A_d is null!"; LOG(FATAL) << "parameter A_d is null!";
} }
if (A_d2->nnz_s == 0) return; if (A_d->nnz == 0) return;
int nnz = A_d2->nnz_s; int nnz = A_d->nnz;
int block = 512; int block = 512;
int grid = DIVUP(nnz, 512); int grid = DIVUP(nnz, 512);
KeSMatrixCsrAddBias<<<grid, block, 0, STREAM_DEFAULT>>>( KeSMatrixCsrAddBias<<<grid, block, 0, STREAM_DEFAULT>>>(
...@@ -1308,9 +1271,9 @@ void hl_matrix_csr_add_dense(hl_sparse_matrix_s A_d, real* B_d, int dimM, ...@@ -1308,9 +1271,9 @@ void hl_matrix_csr_add_dense(hl_sparse_matrix_s A_d, real* B_d, int dimM,
LOG(FATAL) << "parameter A_d is null!"; LOG(FATAL) << "parameter A_d is null!";
} }
if (A_d2->nnz_s == 0) return; if (A_d->nnz == 0) return;
int gridX = DIVUP((A_d2->nnz_s / dimM), 512); int gridX = DIVUP((A_d->nnz / dimM), 512);
gridX = gridX > 0 ? gridX : 1; gridX = gridX > 0 ? gridX : 1;
dim3 block(512, 1); dim3 block(512, 1);
dim3 grid(gridX, dimM); dim3 grid(gridX, dimM);
......
paddle/cuda/src/hl_cuda_sparse.cuh
浏览文件 @ b52039bd
...@@ -85,6 +85,15 @@ __global__ void KeSMatrixCsc2Dense(real * csc_val, ...@@ -85,6 +85,15 @@ __global__ void KeSMatrixCsc2Dense(real * csc_val,
C_d[row*dimN + col] = sum; C_d[row*dimN + col] = sum;
} }
__device__ __forceinline__
void _calculate_c(real &c, real sum) {
c = sum;
}
__device__ __forceinline__
void _calculate_c(real &c, real sum, real beta) {
c = sum + beta * c;
}
#define CU_CSRMM_N 4 #define CU_CSRMM_N 4
#define CU_CSRMM_THREAD_X 32 #define CU_CSRMM_THREAD_X 32
#define CU_CSRMM_THREAD_Y 32 #define CU_CSRMM_THREAD_Y 32
...@@ -191,13 +200,21 @@ __global__ void KeSMatrixCsrMulDense(real *C_d, ...@@ -191,13 +200,21 @@ __global__ void KeSMatrixCsrMulDense(real *C_d,
} }
C_d += __mul24(index_m, dimN); C_d += __mul24(index_m, dimN);
#pragma unroll if (beta == 0.0) {
for (int n = 0; n < CU_CSRMM_N; n++) {
if (index_n < dimN) {
_calculate_c(C_d[index_n], alpha * sum[n]);
index_n += CU_CSRMM_THREAD_X;
}
}
} else {
for (int n = 0; n < CU_CSRMM_N; n++) { for (int n = 0; n < CU_CSRMM_N; n++) {
if (index_n < dimN) { if (index_n < dimN) {
C_d[index_n] = alpha*sum[n] + beta*C_d[index_n]; _calculate_c(C_d[index_n], alpha * sum[n], beta);
index_n += CU_CSRMM_THREAD_X; index_n += CU_CSRMM_THREAD_X;
} }
} }
}
} }
#define CU_CSC_MUL_DENSE_THREAD_N 1 #define CU_CSC_MUL_DENSE_THREAD_N 1
...@@ -544,14 +561,23 @@ TEMP_TEST: ...@@ -544,14 +561,23 @@ TEMP_TEST:
int index_m_c = ibx + idy; int index_m_c = ibx + idy;
int index_n_c = blockIdx.y*CU_CSCMM_BLOCK_N_BEST + idx; int index_n_c = blockIdx.y*CU_CSCMM_BLOCK_N_BEST + idx;
C_d += index_n_c + __mul24(index_m_c, dimN); C_d += index_n_c + __mul24(index_m_c, dimN);
#pragma unroll if (beta == 0.0) {
for (int m = 0; m < CU_CSCMM_THREAD_M_BEST; m++) { for (int m = 0; m < CU_CSCMM_THREAD_M_BEST; m++) {
if (index_m_c < dimM && index_n_c < dimN) { if (index_m_c < dimM && index_n_c < dimN) {
C_d[0] = A_s[idy+m*32][idx] + beta*C_d[0]; _calculate_c(C_d[0], A_s[idy + m * 32][idx]);
} }
index_m_c += 32; index_m_c += 32;
C_d += dimN*32; C_d += dimN*32;
} }
} else {
for (int m = 0; m < CU_CSCMM_THREAD_M_BEST; m++) {
if (index_m_c < dimM && index_n_c < dimN) {
_calculate_c(C_d[0], A_s[idy + m * 32][idx], beta);
}
index_m_c += 32;
C_d += dimN*32;
}
}
} }
#define CU_DM_CSR_THREAD_X 32 #define CU_DM_CSR_THREAD_X 32
......
paddle/gserver/layers/Layer.cpp
浏览文件 @ b52039bd
...@@ -16,6 +16,7 @@ limitations under the License. */ ...@@ -16,6 +16,7 @@ limitations under the License. */
#include "paddle/utils/Util.h" #include "paddle/utils/Util.h"
#include "paddle/utils/Logging.h" #include "paddle/utils/Logging.h"
#include "paddle/math/SparseMatrix.h"
#include "AddtoLayer.h" #include "AddtoLayer.h"
#include "CosSimLayer.h" #include "CosSimLayer.h"
...@@ -290,14 +291,30 @@ void Layer::showOutputStats() { ...@@ -290,14 +291,30 @@ void Layer::showOutputStats() {
<< " is 0, skip to show the statistics"; << " is 0, skip to show the statistics";
return; return;
} }
real mean = out->getSum() / out->getElementCnt(); MatrixPtr outSquare;
MatrixPtr outSquare = out->clone(); if (dynamic_cast<GpuSparseMatrix*>(out.get())) {
outSquare->copyFrom(*out); GpuSparseMatrix *tmp = dynamic_cast<GpuSparseMatrix*>(out.get());
outSquare = std::make_shared<CpuSparseMatrix>(
tmp->getHeight(), tmp->getWidth(), tmp->getElementCnt(),
tmp->getValueType(), tmp->getFormat());
} else {
outSquare = out->clone();
}
outSquare->copyFrom(*out, HPPL_STREAM_DEFAULT);
hl_stream_synchronize(HPPL_STREAM_DEFAULT);
real mean = outSquare->getSum() / out->getElementCnt();
real min;
real max;
if (dynamic_cast<CpuSparseMatrix*>(outSquare.get())) { if (dynamic_cast<CpuSparseMatrix*>(outSquare.get())) {
auto tmpMat = dynamic_cast<CpuSparseMatrix*>(outSquare.get()); auto tmpMat = dynamic_cast<CpuSparseMatrix*>(outSquare.get());
min = tmpMat->getMin();
max = tmpMat->getMax();
tmpMat->square(); tmpMat->square();
LOG(INFO) << "show statistics of [none zero values] in sparse matrix"; LOG(INFO) << "show statistics of [none zero values] in sparse matrix";
} else { } else {
min = outSquare->getMin();
max = outSquare->getMax();
outSquare->square(); outSquare->square();
} }
real std = (outSquare->getSum() / outSquare->getElementCnt()) - mean * mean; real std = (outSquare->getSum() / outSquare->getElementCnt()) - mean * mean;
...@@ -306,8 +323,8 @@ void Layer::showOutputStats() { ...@@ -306,8 +323,8 @@ void Layer::showOutputStats() {
<< ", " << ", "
<< "std=" << std << "std=" << std
<< ", " << ", "
<< "min=" << out->getMin() << ", " << "min=" << min << ", "
<< "max=" << out->getMax(); << "max=" << max;
} }
void Layer::forwardActivation() { void Layer::forwardActivation() {
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或