Brutex
/
cavis
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Shyrma mkl test (#211) 

* - provide nhwc format in mkl conv ops

Signed-off-by: Yurii <iuriish@yahoo.com>

* - corrections in mkl conv3d

Signed-off-by: Yurii <iuriish@yahoo.com>

* - corrections in mkl batchnorm

Signed-off-by: Yurii <iuriish@yahoo.com>

* - corrections in mkl maxpooling2d

Signed-off-by: Yurii <iuriish@yahoo.com>

* - add format format_tag::any to outputs in mkl conv ops

Signed-off-by: Yurii <iuriish@yahoo.com>

* - complete corrections in mkl conv ops

Signed-off-by: Yurii <iuriish@yahoo.com>

* - add test for comparison of execution speeds of mkl conv2d op with different weights format

Signed-off-by: Yurii <iuriish@yahoo.com>

* - take into account order f in mkl conv ops

Signed-off-by: Yurii <iuriish@yahoo.com>
master
Yurii Shyrma 2020-02-06 20:12:54 +02:00 committed by GitHub
parent 5ae40f6e38
commit 948646b32d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
19 changed files with 1737 additions and 1442 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
libnd4j/include/ops/declarable/generic/nn/pooling/maxpool3d.cpp
View File

@ -169,8 +169,8 @@ CUSTOM_OP_IMPL(maxpool3dnew_bp, 2, 1, false, 0, 14) {
// int extraParam0 = INT_ARG(13); // unnecessary for max case, required only for avg and pnorm cases
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
REQUIRE_TRUE(input->rankOf() == 5, 0, "MAXPOOL3D_BP op: input should have rank of 5, but got %i instead", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0, "MAXPOOL3DNEW op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
REQUIRE_TRUE(input->rankOf() == 5, 0, "MAXPOOL3DNEW_BP op: input should have rank of 5, but got %i instead", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0, "MAXPOOL3DNEW_BP op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
@ -178,8 +178,8 @@ CUSTOM_OP_IMPL(maxpool3dnew_bp, 2, 1, false, 0, 14) {
std::string expectedGradOShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx({bS,iC,oD,oH,oW, 0,indIOioC,indIOioD,indIOioD+1,indIOioD+2}));
std::string expectedGradIShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx({bS,iC,iD,iH,iW, 0,indIOioC,indIOioD,indIOioD+1,indIOioD+2}));
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0, "MAXPOOL3D_BP op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !", expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedGradIShape == ShapeUtils::shapeAsString(gradI), 0, "MAXPOOL3D_BP op: wrong shape of input's gradients array (epsilon), expected is %s, but got %s instead !", expectedGradIShape.c_str(), ShapeUtils::shapeAsString(gradI).c_str());
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0, "MAXPOOL3DNEW_BP op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !", expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedGradIShape == ShapeUtils::shapeAsString(gradI), 0, "MAXPOOL3DNEW_BP op: wrong shape of input's gradients array (epsilon), expected is %s, but got %s instead !", expectedGradIShape.c_str(), ShapeUtils::shapeAsString(gradI).c_str());
if(!isNCDHW) {
input = new NDArray(input->permute({0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]

2
libnd4j/include/ops/declarable/platform/cudnn/cudnnUtils.cu
View File

@ -250,7 +250,7 @@ void pooling3dCUDNN(const LaunchContext* context,
auto handle = reinterpret_cast<cudnnHandle_t *>(context->getCuDnnHandle());
cudnnStatus_t err = cudnnSetStream(*handle, *context->getCudaStream());
if (err != 0) throw nd4j::cuda_exception::build("pooling3dCUDNN: can't set stream for cuDNN", err);
printf("fffffffffff\n");
const int numDims = 5;
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;

186
libnd4j/include/ops/declarable/platform/mkldnn/avgpooling2d.cpp
View File

@ -17,6 +17,7 @@
//
// @author saudet
// @author raver119@gmail.com
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <ops/declarable/PlatformHelper.h>
@ -36,103 +37,44 @@ namespace platforms {
//////////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(avgpool2d, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
REQUIRE_TRUE(input->rankOf() == 4, 0, "Input should have rank of 4, but got %i instead",
input->rankOf());
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
// 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
auto argI = *(block.getIArguments());
auto output = OUTPUT_VARIABLE(0);
const auto kH = INT_ARG(0);
const auto kW = INT_ARG(1);
const auto sH = INT_ARG(2);
const auto sW = INT_ARG(3);
int pH = INT_ARG(4);
int pW = INT_ARG(5);
auto pH = INT_ARG(4);
auto pW = INT_ARG(5);
const auto dH = INT_ARG(6);
const auto dW = INT_ARG(7);
const auto isSameMode = static_cast<bool>(INT_ARG(8));
const auto paddingMode = INT_ARG(8);
const auto extraParam0 = INT_ARG(9);
const int isNCHW = block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
REQUIRE_TRUE(dH != 0 && dW != 0, 0, "AVGPOOL2D op: dilation must not be zero, but got instead {%i, %i}",
dH, dW);
REQUIRE_TRUE(input->rankOf() == 4, 0, "AVGPOOL2D MKLDNN op: input should have rank of 4, but got %i instead", input->rankOf());
REQUIRE_TRUE(dH != 0 && dW != 0, 0, "AVGPOOL2D MKLDNN op: dilation must not be zero, but got instead {%i, %i}", dH, dW);
int oH = 0;
int oW = 0;
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
int isNCHW = block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
const int iH = static_cast<int>(isNCHW ? input->sizeAt(2) : input->sizeAt(1));
const int iW = static_cast<int>(isNCHW ? input->sizeAt(3) : input->sizeAt(2));
if (!isNCHW) {
input = new NDArray(
input->permute({0, 3, 1, 2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
output = new NDArray(
output->permute({0, 3, 1, 2})); // [bS, oH, oW, iC] -> [bS, iC, oH, oW]
}
ConvolutionUtils::calcOutSizePool2D(oH, oW, kH, kW, sH, sW, pH, pW, dH, dW, iH, iW, isSameMode);
if (isSameMode)
if (paddingMode)
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW);
const int bS = input->sizeAt(0);
const int iC = input->sizeAt(1);
const int oC = output->sizeAt(1);
auto mode = (extraParam0 == 0) ? algorithm::pooling_avg_exclude_padding : algorithm::pooling_avg_include_padding;
auto poolingMode = PoolingType::AVG_POOL;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool2d(kH, kW, sH, sW, pH, pW, dH, dW, poolingMode, extraParam0,
true,
bS, iC, iH, iW, oC, oH, oW, input, nullptr, output,
algorithm,
&pool_src_md, nullptr, &pool_dst_md, &user_src_md, nullptr,
&user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto pool_desc = pooling_forward::desc(prop_kind::forward_inference, algorithm, pool_src_md,
pool_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto user_src_memory = dnnl::memory(user_src_md, engine, input->buffer());
auto user_dst_memory = dnnl::memory(user_dst_md, engine, output->buffer());
auto pool_src_memory = user_src_memory;
dnnl::stream stream(engine);
if (pool_prim_desc.src_desc() != user_src_memory.get_desc()) {
pool_src_memory = dnnl::memory(pool_prim_desc.src_desc(), engine);
reorder(user_src_memory, pool_src_memory).execute(stream, user_src_memory, pool_src_memory);
}
auto pool_dst_memory = user_dst_memory;
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
pool_dst_memory = dnnl::memory(pool_prim_desc.dst_desc(), engine);
}
pooling_forward(pool_prim_desc).execute(stream, {{DNNL_ARG_SRC, pool_src_memory},
{DNNL_ARG_DST, pool_dst_memory}});
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
reorder(pool_dst_memory, user_dst_memory).execute(stream, pool_dst_memory, user_dst_memory);
}
stream.wait();
//streams[0].submitAndWait();
if (!isNCHW) {
delete input;
delete output;
}
mkldnnUtils::poolingMKLDNN(input, output, 0,kH,kW, 0,sH,sW, 0,pH,pW, isNCHW, mode);
return Status::OK();
}
//////////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(avgpool2d, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
@ -141,12 +83,10 @@ PLATFORM_CHECK(avgpool2d, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(avgpool2d_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(
0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto gradO = INPUT_VARIABLE(
1); // [bS, oH, oW, oC] (NHWC) or [bS, oC, oH, oW] (NCHW), epsilon_next
auto gradI = OUTPUT_VARIABLE(
0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW), epsilon
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto gradO = INPUT_VARIABLE(1); // [bS, oH, oW, oC] (NHWC) or [bS, oC, oH, oW] (NCHW), epsilon_next
auto gradI = OUTPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW), epsilon
int kH = INT_ARG(0); // filter(kernel) height
int kW = INT_ARG(1); // filter(kernel) width
@ -156,92 +96,26 @@ PLATFORM_IMPL(avgpool2d_bp, ENGINE_CPU) {
int pW = INT_ARG(5); // paddings width
int dH = INT_ARG(6); // dilations height
int dW = INT_ARG(7); // dilations width
int isSameMode = INT_ARG(8); // 0-VALID, 1-SAME
int paddingMode = INT_ARG(8); // 0-VALID, 1-SAME
int extraParam0 = INT_ARG(9);
int isNCHW =
block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
int isNCHW = block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
REQUIRE_TRUE(input->rankOf() == 4, 0,
"AVGPOOL2D_BP op: input should have rank of 4, but got %i instead", input->rankOf());
REQUIRE_TRUE(dH != 0 && dW != 0, 0,
"AVGPOOL2D_BP op: dilation must not be zero, but got instead {%i, %i}", dH, dW);
REQUIRE_TRUE(input->rankOf() == 4, 0, "AVGPOOL2D_BP MKLDNN op: input should have rank of 4, but got %i instead", input->rankOf());
REQUIRE_TRUE(dH != 0 && dW != 0, 0, "AVGPOOL2D_BP MKLDNN op: dilation must not be zero, but got instead {%i, %i}", dH, dW);
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC,
indIiH, indWiC, indWoC, indWkH, indOoH);
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
std::string expectedGradOShape = ShapeUtils::shapeAsString(
ShapeUtils::composeShapeUsingDimsAndIdx({bS, iC, oH, oW, 0, indIOioC, indIiH, indIiH + 1}));
std::string expectedGradIShape = ShapeUtils::shapeAsString(
ShapeUtils::composeShapeUsingDimsAndIdx({bS, iC, iH, iW, 0, indIOioC, indIiH, indIiH + 1}));
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0,
"AVGPOOL2D_BP op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !",
expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedGradIShape == ShapeUtils::shapeAsString(gradI), 0,
"AVGPOOL2D_BP op: wrong shape of input's gradients array (epsilon), expected is %s, but got %s instead !",
expectedGradIShape.c_str(), ShapeUtils::shapeAsString(gradI).c_str());
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx({bS,iC,oH,oW, 0,indIOioC,indIiH,indIiH+1});
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "AVGPOOL2D_BP MKLDNN op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
if (!isNCHW) {
input = new NDArray(input->permute(
{0, 3, 1, 2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradI = new NDArray(gradI->permute(
{0, 3, 1, 2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradO = new NDArray(gradO->permute(
{0, 3, 1, 2})); // [bS, oH, oW, iC] -> [bS, iC, oH, oW]
}
if (isSameMode) // SAME
if(paddingMode) // SAME
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW);
auto poolingMode = PoolingType::AVG_POOL;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_diff_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_diff_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool2d(kH, kW, sH, sW, pH, pW, dH, dW, poolingMode, extraParam0,
true,
bS, iC, iH, iW, oC, oH, oW, input, gradI, gradO, algorithm,
&pool_src_md, &pool_diff_src_md, &pool_dst_md, &user_src_md,
&user_diff_src_md, &user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto pool_desc = pooling_forward::desc(prop_kind::forward, algorithm,
input->buffer() != nullptr ? pool_src_md : pool_diff_src_md,
pool_dst_md, pool_strides, pool_kernel, pool_padding,
pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto poolB_desc = pooling_backward::desc(algorithm, pool_diff_src_md, pool_dst_md, pool_strides,
pool_kernel, pool_padding, pool_padding_r);
auto poolB_prim_desc = pooling_backward::primitive_desc(poolB_desc, engine, pool_prim_desc);
auto userB_src_memory = dnnl::memory(user_src_md, engine, gradI->buffer());
auto userB_dst_memory = dnnl::memory(user_dst_md, engine, gradO->buffer());
auto poolB_src_memory = userB_src_memory;
dnnl::stream stream(engine);
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
poolB_src_memory = dnnl::memory(poolB_prim_desc.diff_src_desc(), engine);
}
auto poolB_dst_memory = userB_dst_memory;
if (poolB_prim_desc.diff_dst_desc() != userB_dst_memory.get_desc()) {
poolB_dst_memory = dnnl::memory(poolB_prim_desc.diff_dst_desc(), engine);
reorder(userB_dst_memory, poolB_dst_memory).execute(stream, userB_dst_memory, poolB_dst_memory);
}
pooling_backward(poolB_prim_desc).execute(stream, {{DNNL_ARG_DIFF_DST, poolB_dst_memory},
{DNNL_ARG_DIFF_SRC, poolB_src_memory}});
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
reorder(poolB_src_memory, userB_src_memory).execute(stream, poolB_src_memory, userB_src_memory);
}
stream.wait();
if (!isNCHW) {
delete input;
delete gradI;
delete gradO;
}
auto mode = (extraParam0 == 0) ? algorithm::pooling_avg_exclude_padding : algorithm::pooling_avg_include_padding;
mkldnnUtils::poolingBpMKLDNN(input, gradO, gradI, 0,kH,kW, 0,sH,sW, 0,pH,pW, isNCHW, mode);
return Status::OK();
}

200
libnd4j/include/ops/declarable/platform/mkldnn/avgpooling3d.cpp
View File

@ -17,6 +17,7 @@
//
// @author saudet
// @author raver119@gmail.com
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <ops/declarable/PlatformHelper.h>
@ -29,113 +30,110 @@
using namespace dnnl;
namespace nd4j {
namespace ops {
namespace platforms {
PLATFORM_IMPL(avgpool3dnew, ENGINE_CPU) {
auto input = INPUT_VARIABLE(
0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto output = OUTPUT_VARIABLE(
0); // [bS, oD, oH, oW, iC] (NDHWC) or [bS, iC, oD, oH, oW] (NCDHW)
namespace nd4j {
namespace ops {
namespace platforms {
int kD = INT_ARG(0); // filter(kernel) depth
int kH = INT_ARG(1); // filter(kernel) height
int kW = INT_ARG(2); // filter(kernel) width
int sD = INT_ARG(3); // strides depth
int sH = INT_ARG(4); // strides height
int sW = INT_ARG(5); // strides width
int pD = INT_ARG(6); // paddings depth
int pH = INT_ARG(7); // paddings height
int pW = INT_ARG(8); // paddings width
int dD = INT_ARG(9); // dilations depth
int dH = INT_ARG(10); // dilations height
int dW = INT_ARG(11); // dilations width
int isSameMode = INT_ARG(12); // 1-SAME, 0-VALID
int extraParam0 = INT_ARG(13); // unnecessary for max case, required only for avg and pnorm cases
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
//////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(avgpool3dnew, ENGINE_CPU) {
REQUIRE_TRUE(input->rankOf() == 5, 0,
"MAXPOOL3DNEW OP: rank of input array must be equal to 5, but got %i instead !",
input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0,
"MAXPOOL3DNEW op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
auto input = INPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto output = OUTPUT_VARIABLE(0); // [bS, oD, oH, oW, iC] (NDHWC) or [bS, iC, oD, oH, oW] (NCDHW)
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW,
indIOioC, indIOioD, indWiC, indWoC, indWkD);
int kD = INT_ARG(0); // filter(kernel) depth
int kH = INT_ARG(1); // filter(kernel) height
int kW = INT_ARG(2); // filter(kernel) width
int sD = INT_ARG(3); // strides depth
int sH = INT_ARG(4); // strides height
int sW = INT_ARG(5); // strides width
int pD = INT_ARG(6); // paddings depth
int pH = INT_ARG(7); // paddings height
int pW = INT_ARG(8); // paddings width
int dD = INT_ARG(9); // dilations depth
int dH = INT_ARG(10); // dilations height
int dW = INT_ARG(11); // dilations width
int paddingMode = INT_ARG(12); // 1-SAME, 0-VALID
int extraParam0 = INT_ARG(13);
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 0-NCDHW, 1-NDHWC
std::string expectedOutputShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx(
{bS, iC, oD, oH, oW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
REQUIRE_TRUE(expectedOutputShape == ShapeUtils::shapeAsString(output), 0,
"MAXPOOL3D op: wrong shape of output array, expected is %s, but got %s instead !",
expectedOutputShape.c_str(), ShapeUtils::shapeAsString(output).c_str());
// REQUIRE_TRUE(iD >= kD && iH >= kH && iW >= kW, 0, "MAXPOOL3D OP: the input depth/height/width must be greater or equal to kernel(filter) depth/height/width, but got [%i, %i, %i] and [%i, %i, %i] correspondingly !", iD,iH,iW, kD,kH,kW);
// REQUIRE_TRUE(kD/2 >= pD && kH/2 >= pH && kW/2 >= pW, 0, "MAXPOOL3D OP: pad depth/height/width must not be greater than half of kernel depth/height/width, but got [%i, %i, %i] and [%i, %i, %i] correspondingly !", pD,pH,pW, kD,kH,kW);
REQUIRE_TRUE(input->rankOf() == 5, 0, "AVGPOOL3DNEW MKLDNN OP: rank of input array must be equal to 5, but got %i instead !", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0, "AVGPOOL3DNEW MKLDNN OP: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
if (!isNCDHW) {
input = new NDArray(
input->permute({0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
output = new NDArray(
output->permute({0, 4, 1, 2, 3})); // [bS, oD, oH, oW, iC] -> [bS, iC, oD, oH, oW]
}
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
if (isSameMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
if(paddingMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
auto mode = (extraParam0 == 0) ? algorithm::pooling_avg_exclude_padding : algorithm::pooling_avg_include_padding;
mkldnnUtils::poolingMKLDNN(input, output, kD,kH,kW, sD,sH,sW, pD,pH,pW, isNCDHW, mode);
return Status::OK();
}
//////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(avgpool3dnew, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
return block.isUseMKLDNN() && nd4j::MKLDNNStream::isSupported({input, output});
}
//////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(avgpool3dnew_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto gradO = INPUT_VARIABLE(1); // [bS, oD, oH, oW, oC] (NDHWC) or [bS, oC, oD, oH, oW] (NCDHW), epsilon_next
auto gradI = OUTPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW), epsilon
const int kD = INT_ARG(0); // filter(kernel) depth
const int kH = INT_ARG(1); // filter(kernel) height
const int kW = INT_ARG(2); // filter(kernel) width
const int sD = INT_ARG(3); // strides depth
const int sH = INT_ARG(4); // strides height
const int sW = INT_ARG(5); // strides width
int pD = INT_ARG(6); // paddings depth
int pH = INT_ARG(7); // paddings height
int pW = INT_ARG(8); // paddings width
const int dD = INT_ARG(9); // dilations depth
const int dH = INT_ARG(10); // dilations height
const int dW = INT_ARG(11); // dilations width
const int paddingMode = INT_ARG(12); // 1-SAME, 0-VALID
const int extraParam0 = INT_ARG(13); // define what divisor to use while averaging
const int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 0-NCDHW, 1-NDHWC
REQUIRE_TRUE(input->rankOf() == 5, 0, "AVGPOOL3DNEW_BP MKLDNN op: input should have rank of 5, but got %i instead", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0, "AVGPOOL3DNEW_BP MKLDNN op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx({bS,iC,oD,oH,oW, 0,indIOioC,indIOioD,indIOioD+1,indIOioD+2});
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "AVGPOOL3DNEW_BP MKLDNN op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
if(paddingMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
auto mode = (extraParam0 == 0) ? algorithm::pooling_avg_exclude_padding : algorithm::pooling_avg_include_padding;
mkldnnUtils::poolingBpMKLDNN(input, gradO, gradI, kD,kH,kW, sD,sH,sW, pD,pH,pW, isNCDHW, mode);
return Status::OK();
}
//////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(avgpool3dnew_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
return block.isUseMKLDNN() && nd4j::MKLDNNStream::isSupported({input, output});
}
auto poolingMode = PoolingType::AVG_POOL;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool3d(kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, poolingMode,
extraParam0, true,
bS, iC, iD, iH, iW, oC, oD, oH, oW, input, nullptr, output,
algorithm,
&pool_src_md, nullptr, &pool_dst_md, &user_src_md, nullptr,
&user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto pool_desc = pooling_forward::desc(prop_kind::forward_inference, algorithm, pool_src_md,
pool_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto user_src_memory = dnnl::memory(user_src_md, engine, input->buffer());
auto user_dst_memory = dnnl::memory(user_dst_md, engine, output->buffer());
auto pool_src_memory = user_src_memory;
if (pool_prim_desc.src_desc() != user_src_memory.get_desc()) {
pool_src_memory = dnnl::memory(pool_prim_desc.src_desc(), engine);
reorder(user_src_memory, pool_src_memory).execute(stream, user_src_memory, pool_src_memory);
}
auto pool_dst_memory = user_dst_memory;
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
pool_dst_memory = dnnl::memory(pool_prim_desc.dst_desc(), engine);
}
pooling_forward(pool_prim_desc).execute(stream, {{DNNL_ARG_SRC, pool_src_memory},
{DNNL_ARG_DST, pool_dst_memory}});
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
reorder(pool_dst_memory, user_dst_memory).execute(stream, pool_dst_memory, user_dst_memory);
}
stream.wait();
if (!isNCDHW) {
delete input;
delete output;
}
return Status::OK();
}
PLATFORM_CHECK(avgpool3dnew, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
return block.isUseMKLDNN() && nd4j::MKLDNNStream::isSupported({input, output});
}
}
}
}
}
}

154
libnd4j/include/ops/declarable/platform/mkldnn/avgpooling3d_bp.cpp
View File

@ -1,154 +0,0 @@
/*******************************************************************************
* 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 raver119@gmail.com
//
#include <ops/declarable/PlatformHelper.h>
#include <ops/declarable/OpRegistrator.h>
#include <platform_boilerplate.h>
#include <helpers/MKLDNNStream.h>
#include "mkldnnUtils.h"
#include <ops/declarable/helpers/convolutions.h>
using namespace dnnl;
namespace nd4j {
namespace ops {
namespace platforms {
PLATFORM_IMPL(avgpool3dnew_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(
0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto gradO = INPUT_VARIABLE(
1); // [bS, oD, oH, oW, oC] (NDHWC) or [bS, oC, oD, oH, oW] (NCDHW), epsilon_next
auto gradI = OUTPUT_VARIABLE(
0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW), epsilon
const int kD = INT_ARG(0); // filter(kernel) depth
const int kH = INT_ARG(1); // filter(kernel) height
const int kW = INT_ARG(2); // filter(kernel) width
const int sD = INT_ARG(3); // strides depth
const int sH = INT_ARG(4); // strides height
const int sW = INT_ARG(5); // strides width
int pD = INT_ARG(6); // paddings depth
int pH = INT_ARG(7); // paddings height
int pW = INT_ARG(8); // paddings width
const int dD = INT_ARG(9); // dilations depth
const int dH = INT_ARG(10); // dilations height
const int dW = INT_ARG(11); // dilations width
const int isSameMode = INT_ARG(12); // 1-SAME, 0-VALID
int extraParam0 = INT_ARG(13); // unnecessary for max case, required only for avg and pnorm cases
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
REQUIRE_TRUE(input->rankOf() == 5, 0,
"MAXPOOL3D_BP op: input should have rank of 5, but got %i instead", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0,
"MAXPOOL3DNEW op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW,
indIOioC, indIOioD, indWiC, indWoC, indWkD);
std::string expectedGradOShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx(
{bS, iC, oD, oH, oW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
std::string expectedGradIShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx(
{bS, iC, iD, iH, iW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0,
"MAXPOOL3D_BP op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !",
expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedGradIShape == ShapeUtils::shapeAsString(gradI), 0,
"MAXPOOL3D_BP op: wrong shape of input's gradients array (epsilon), expected is %s, but got %s instead !",
expectedGradIShape.c_str(), ShapeUtils::shapeAsString(gradI).c_str());
if (!isNCDHW) {
input = new NDArray(input->permute(
{0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
gradI = new NDArray(gradI->permute(
{0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
gradO = new NDArray(gradO->permute(
{0, 4, 1, 2, 3})); // [bS, oD, oH, oW, iC] -> [bS, iC, oD, oH, oW]
}
if (isSameMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
auto poolingMode = PoolingType::AVG_POOL;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_diff_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_diff_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool3d(kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, poolingMode,
extraParam0, true,
bS, iC, iD, iH, iW, oC, oD, oH, oW, input, gradI, gradO,
algorithm,
&pool_src_md, &pool_diff_src_md, &pool_dst_md, &user_src_md,
&user_diff_src_md, &user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
if (input->buffer() == nullptr) {
pool_src_md = pool_diff_src_md;
user_src_md = user_diff_src_md;
}
auto pool_desc = pooling_forward::desc(prop_kind::forward, algorithm, pool_src_md, pool_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto poolB_desc = pooling_backward::desc(algorithm, pool_diff_src_md, pool_dst_md, pool_strides,
pool_kernel, pool_padding, pool_padding_r);
auto poolB_prim_desc = pooling_backward::primitive_desc(poolB_desc, engine, pool_prim_desc);
auto userB_src_memory = dnnl::memory(user_diff_src_md, engine, gradI->buffer());
auto userB_dst_memory = dnnl::memory(user_dst_md, engine, gradO->buffer());
auto poolB_src_memory = userB_src_memory;
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
poolB_src_memory = dnnl::memory(poolB_prim_desc.diff_src_desc(), engine);
}
auto poolB_dst_memory = userB_dst_memory;
if (poolB_prim_desc.diff_dst_desc() != userB_dst_memory.get_desc()) {
poolB_dst_memory = dnnl::memory(poolB_prim_desc.diff_dst_desc(), engine);
reorder(userB_dst_memory, poolB_dst_memory).execute(stream, userB_dst_memory, poolB_dst_memory);
}
pooling_backward(poolB_prim_desc).execute(stream, {{DNNL_ARG_DIFF_DST, poolB_dst_memory},
{DNNL_ARG_DIFF_SRC, poolB_src_memory}});
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
reorder(poolB_src_memory, userB_src_memory).execute(stream, poolB_src_memory, userB_src_memory);
}
stream.wait();
if (!isNCDHW) {
delete input;
delete gradI;
delete gradO;
}
return Status::OK();
}
PLATFORM_CHECK(avgpool3dnew_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto output = OUTPUT_VARIABLE(0);
return block.isUseMKLDNN() && nd4j::MKLDNNStream::isSupported({input, output});
}
}
}
}

199
libnd4j/include/ops/declarable/platform/mkldnn/batchnorm.cpp
View File

@ -37,12 +37,12 @@ namespace platforms {
//////////////////////////////////////////////////////////////////////////
static void batchnormMKLDNN(const NDArray* x, const NDArray* mean, const NDArray* variance, const NDArray* weights, const float epsilon, NDArray* z) {
static void batchnormMKLDNN(const NDArray* x, const NDArray* mean, const NDArray* variance, const NDArray* weights, NDArray* z,
const float epsilon, const bool isNCHW) {
// unfortunately mkl dnn doesn't support any format (dnnl::memory::format_tag::any)
// also it gives wrong results for formats nhwc and ndhwc
// unfortunately mkl dnn doesn't support any format (dnnl::memory::format_tag::any) for x
// x -> 2D:nc, 4D:nchw, 5D:ncdhw
// x -> 2D:nc, 4D:nchw/nhwc, 5D:ncdhw/ndhwc
// mean -> 1D [c]
// variance -> 1D [c]
// weights 2D [2, c], weights({0,1, 0,0}) contains gamma and weights({1,2, 0,0}) contains beta
@ -50,8 +50,6 @@ static void batchnormMKLDNN(const NDArray* x, const NDArray* mean, const NDArray
const int xRank = x->rankOf();
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// input type
dnnl::memory::data_type type = dnnl::memory::data_type::f32;
@ -63,17 +61,28 @@ static void batchnormMKLDNN(const NDArray* x, const NDArray* mean, const NDArray
dnnl::memory::dims dims;
dnnl::memory::format_tag format;
const int indHW = isNCHW ? 2 : 1;
const int bS = x->sizeAt(0);
const int iC = isNCHW ? x->sizeAt(1) : x->sizeAt(-1);
int iD, iH, iW;
if(xRank == 2) {
dims = {x->sizeAt(0), x->sizeAt(1)};
dims = {bS, iC};
format = dnnl::memory::format_tag::nc;
}
else if(xRank == 4) {
dims = {x->sizeAt(0), x->sizeAt(1), x->sizeAt(2), x->sizeAt(3)};
format = dnnl::memory::format_tag::nchw;
iH = x->sizeAt(indHW);
iW = x->sizeAt(indHW + 1);
dims = {bS, iC, iH, iW};
format = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
}
else { // xRank = 5
dims = {x->sizeAt(0), x->sizeAt(1), x->sizeAt(2), x->sizeAt(3), x->sizeAt(4)};
format = dnnl::memory::format_tag::ncdhw;
iD = x->sizeAt(indHW);
iH = x->sizeAt(indHW + 1);
iW = x->sizeAt(indHW + 2);
dims = {bS, iC, iD, iH, iW};
format = isNCHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
}
// memory descriptors for arrays
@ -81,29 +90,34 @@ static void batchnormMKLDNN(const NDArray* x, const NDArray* mean, const NDArray
// x
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(dims, type, format);
dnnl::memory::desc x_user_md = dnnl::memory::desc(dims, type, format);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = x->stridesOf()[0];
x_user_md.data.format_desc.blocking.strides[1] = x->stridesOf()[1];
if(xRank > 2) {
x_user_md.data.format_desc.blocking.strides[2] = x->stridesOf()[2];
x_user_md.data.format_desc.blocking.strides[3] = x->stridesOf()[3];
if(x->ews() != 1 || x->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = x->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = x->strideAt(1);
if(xRank > 2) {
x_user_md.data.format_desc.blocking.strides[2] = x->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = x->strideAt(3);
}
if(xRank > 4)
x_user_md.data.format_desc.blocking.strides[4] = x->strideAt(4);
}
if(xRank > 4)
x_user_md.data.format_desc.blocking.strides[4] = x->stridesOf()[4];
// z, output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(dims, type, format);
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(dims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(dims, type, format);
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = z->stridesOf()[0];
z_user_md.data.format_desc.blocking.strides[1] = z->stridesOf()[1];
if(xRank > 2) {
z_user_md.data.format_desc.blocking.strides[2] = z->stridesOf()[2];
z_user_md.data.format_desc.blocking.strides[3] = z->stridesOf()[3];
if(z->ews() != 1 || z->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = z->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = z->strideAt(1);
if(xRank > 2) {
z_user_md.data.format_desc.blocking.strides[2] = z->strideAt(2);
z_user_md.data.format_desc.blocking.strides[3] = z->strideAt(3);
}
if(xRank > 4)
z_user_md.data.format_desc.blocking.strides[4] = z->strideAt(4);
}
if(xRank > 4)
z_user_md.data.format_desc.blocking.strides[4] = z->stridesOf()[4];
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// batchnorm forward description
dnnl::batch_normalization_forward::desc op_ff_desc(dnnl::prop_kind::forward_inference, x_mkl_md, epsilon, flags);
@ -162,12 +176,11 @@ static void batchnormMKLDNN(const NDArray* x, const NDArray* mean, const NDArray
//////////////////////////////////////////////////////////////////////////
static void batchnormBackPropMKLDNN(const NDArray* x, const NDArray* mean, const NDArray* variance, const NDArray* dLdO, const NDArray* weights,
const float epsilon, NDArray* dLdI, NDArray* dLdW) {
NDArray* dLdI, NDArray* dLdW, const float epsilon, const bool isNCHW) {
// unfortunately mkl dnn doesn't support any format (dnnl::memory::format_tag::any)
// also it gives wrong results for formats nhwc and ndhwc
// unfortunately mkl dnn doesn't support any format (dnnl::memory::format_tag::any) for x
// x -> 2D:nc, 4D:nchw, 5D:ncdhw
// x -> 2D:nc, 4D:nchw/nhwc, 5D:ncdhw/ndhwc
// mean -> 1D [c]
// variance -> 1D [c]
// dLdO - same shape as x
@ -177,8 +190,6 @@ static void batchnormBackPropMKLDNN(const NDArray* x, const NDArray* mean, const
const int xRank = x->rankOf();
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// input type
dnnl::memory::data_type type = dnnl::memory::data_type::f32;
@ -190,17 +201,28 @@ static void batchnormBackPropMKLDNN(const NDArray* x, const NDArray* mean, const
dnnl::memory::dims dims;
dnnl::memory::format_tag format;
const int indHW = isNCHW ? 2 : 1;
const int bS = x->sizeAt(0);
const int iC = isNCHW ? x->sizeAt(1) : x->sizeAt(-1);
int iD, iH, iW;
if(xRank == 2) {
dims = {x->sizeAt(0), x->sizeAt(1)};
dims = {bS, iC};
format = dnnl::memory::format_tag::nc;
}
else if(xRank == 4) {
dims = {x->sizeAt(0), x->sizeAt(1), x->sizeAt(2), x->sizeAt(3)};
format = dnnl::memory::format_tag::nchw;
iH = x->sizeAt(indHW);
iW = x->sizeAt(indHW + 1);
dims = {bS, iC, iH, iW};
format = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
}
else { // xRank = 5
dims = {x->sizeAt(0), x->sizeAt(1), x->sizeAt(2), x->sizeAt(3), x->sizeAt(4)};
format = dnnl::memory::format_tag::ncdhw;
iD = x->sizeAt(indHW);
iH = x->sizeAt(indHW + 1);
iW = x->sizeAt(indHW + 2);
dims = {bS, iC, iD, iH, iW};
format = isNCHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
}
// memory descriptors for arrays
@ -208,41 +230,49 @@ static void batchnormBackPropMKLDNN(const NDArray* x, const NDArray* mean, const
// x
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(dims, type, format);
dnnl::memory::desc x_user_md = dnnl::memory::desc(dims, type, format);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = x->stridesOf()[0];
x_user_md.data.format_desc.blocking.strides[1] = x->stridesOf()[1];
if(xRank > 2) {
x_user_md.data.format_desc.blocking.strides[2] = x->stridesOf()[2];
x_user_md.data.format_desc.blocking.strides[3] = x->stridesOf()[3];
if(x->ews() != 1 || x->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = x->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = x->strideAt(1);
if(xRank > 2) {
x_user_md.data.format_desc.blocking.strides[2] = x->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = x->strideAt(3);
}
if(xRank > 4)
x_user_md.data.format_desc.blocking.strides[4] = x->strideAt(4);
}
if(xRank > 4)
x_user_md.data.format_desc.blocking.strides[4] = x->stridesOf()[4];
// dLdO
dnnl::memory::desc dLdO_mkl_md = dnnl::memory::desc(dims, type, format);
dnnl::memory::desc dLdO_mkl_md = dnnl::memory::desc(dims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc dLdO_user_md = dnnl::memory::desc(dims, type, format);
dLdO_user_md.data.format_kind = dnnl_blocked; // overrides format
dLdO_user_md.data.format_desc.blocking.strides[0] = dLdO->stridesOf()[0];
dLdO_user_md.data.format_desc.blocking.strides[1] = dLdO->stridesOf()[1];
if(xRank > 2) {
dLdO_user_md.data.format_desc.blocking.strides[2] = dLdO->stridesOf()[2];
dLdO_user_md.data.format_desc.blocking.strides[3] = dLdO->stridesOf()[3];
if(dLdO->ews() != 1 || dLdO->ordering() != 'c') {
dLdO_user_md.data.format_kind = dnnl_blocked; // overrides format
dLdO_user_md.data.format_desc.blocking.strides[0] = dLdO->strideAt(0);
dLdO_user_md.data.format_desc.blocking.strides[1] = dLdO->strideAt(1);
if(xRank > 2) {
dLdO_user_md.data.format_desc.blocking.strides[2] = dLdO->strideAt(2);
dLdO_user_md.data.format_desc.blocking.strides[3] = dLdO->strideAt(3);
}
if(xRank > 4)
dLdO_user_md.data.format_desc.blocking.strides[4] = dLdO->strideAt(4);
}
if(xRank > 4)
dLdO_user_md.data.format_desc.blocking.strides[4] = dLdO->stridesOf()[4];
// dLdI
dnnl::memory::desc dLdI_mkl_md = dnnl::memory::desc(dims, type, format);
dnnl::memory::desc dLdI_mkl_md = dnnl::memory::desc(dims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc dLdI_user_md = dnnl::memory::desc(dims, type, format);
dLdI_user_md.data.format_kind = dnnl_blocked; // overrides format
dLdI_user_md.data.format_desc.blocking.strides[0] = dLdI->stridesOf()[0];
dLdI_user_md.data.format_desc.blocking.strides[1] = dLdI->stridesOf()[1];
if(xRank > 2) {
dLdI_user_md.data.format_desc.blocking.strides[2] = dLdI->stridesOf()[2];
dLdI_user_md.data.format_desc.blocking.strides[3] = dLdI->stridesOf()[3];
if(dLdI->ews() != 1 || dLdI->ordering() != 'c') {
dLdI_user_md.data.format_kind = dnnl_blocked; // overrides format
dLdI_user_md.data.format_desc.blocking.strides[0] = dLdI->strideAt(0);
dLdI_user_md.data.format_desc.blocking.strides[1] = dLdI->strideAt(1);
if(xRank > 2) {
dLdI_user_md.data.format_desc.blocking.strides[2] = dLdI->strideAt(2);
dLdI_user_md.data.format_desc.blocking.strides[3] = dLdI->strideAt(3);
}
if(xRank > 4)
dLdI_user_md.data.format_desc.blocking.strides[4] = dLdI->strideAt(4);
}
if(xRank > 4)
dLdI_user_md.data.format_desc.blocking.strides[4] = dLdI->stridesOf()[4];
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// batchnorm forward description
dnnl::batch_normalization_forward::desc op_ff_desc(dnnl::prop_kind::forward_inference, x_mkl_md, epsilon, flags);
@ -331,7 +361,7 @@ static void batchnormBackPropMKLDNN(const NDArray* x, const NDArray* mean, const
// dLdI = dfdm / N + (2/N) * dfdv * (dvdm/2 + (x - m))
// dLdI = gamma * ( stdInv * -g_sum/N + (2/N) * dfdv * (dvdm/2 + (x - m)) )
std::vector<int> axes = {1};
std::vector<int> axes = isNCHW ? std::vector<int>{1} : std::vector<int>{xRank - 1};
const auto excludedAxes = ShapeUtils::evalDimsToExclude(x->rankOf(), axes);
// inversed batch size 1 / N
@ -377,7 +407,7 @@ static void batchnormBackPropMKLDNN(const NDArray* x, const NDArray* mean, const
PLATFORM_IMPL(batchnorm, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw, 5D:ncdhw
auto input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw/nhwc, 5D:ncdhw/ndhwc
auto mean = INPUT_VARIABLE(1); // [c]
auto variance = INPUT_VARIABLE(2); // [c]
NDArray* gamma = nullptr; // [c]
@ -436,27 +466,19 @@ PLATFORM_IMPL(batchnorm, ENGINE_CPU) {
(*weights)({1,2, 0,0}).assign(0);
}
if(axes[0] == inRank - 1 && inRank > 2) { // if nhwc or ndhwc
std::vector<int> permut = inRank == 4 ? std::vector<int>({0,3,1,2}) : std::vector<int>({0,4,1,2,3});
input = new NDArray(input->permute(permut));
output = new NDArray(output->permute(permut));
}
const bool isNCHW = !(axes[0] == inRank - 1 && inRank > 2);
batchnormMKLDNN(input, mean, variance, weights, epsilon, output);
batchnormMKLDNN(input, mean, variance, weights, output, epsilon, isNCHW);
delete weights;
if(axes[0] == inRank - 1 && inRank > 2) {
delete input;
delete output;
}
return Status::OK();
}
//////////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(batchnorm, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw, 5D:ncdhw
auto input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw/nhwc, 5D:ncdhw/ndhwc
auto mean = INPUT_VARIABLE(1); // [c]
auto variance = INPUT_VARIABLE(2); // [c]
NDArray* gamma = nullptr; // [c]
@ -630,7 +652,7 @@ PLATFORM_CHECK(batchnorm, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(batchnorm_bp, ENGINE_CPU) {
NDArray* input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw, 5D:ncdhw
NDArray* input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw/nhwc, 5D:ncdhw/ndhwc
NDArray* mean = INPUT_VARIABLE(1); // [c]
NDArray* variance = INPUT_VARIABLE(2); // [c]
NDArray* gamma = nullptr; // [c]
@ -698,15 +720,9 @@ PLATFORM_IMPL(batchnorm_bp, ENGINE_CPU) {
(*weights)({1,2, 0,0}).assign(0);
}
const bool isNCHW = !(axes[0] == inRank - 1 && inRank > 2);
if(axes[0] == inRank - 1 && inRank > 2) { // if nhwc or ndhwc
std::vector<int> permut = inRank == 4 ? std::vector<int>({0,3,1,2}) : std::vector<int>({0,4,1,2,3});
input = new NDArray(input->permute(permut));
dLdO = new NDArray(dLdO->permute(permut));
dLdI = new NDArray(dLdI->permute(permut));
}
batchnormBackPropMKLDNN(input, mean, variance, dLdO, weights, epsilon, dLdI, dLdW);
batchnormBackPropMKLDNN(input, mean, variance, dLdO, weights, dLdI, dLdW, epsilon, isNCHW);
*dLdM = 0;
*dLdV = 0;
@ -721,17 +737,12 @@ PLATFORM_IMPL(batchnorm_bp, ENGINE_CPU) {
delete dLdW;
}
if(axes[0] == inRank - 1 && inRank > 2) {
delete input;
delete dLdO;
delete dLdI;
}
return Status::OK();
}
//////////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(batchnorm_bp, ENGINE_CPU) {
NDArray* input = INPUT_VARIABLE(0); // 2D:nc, 4D:nchw, 5D:ncdhw
NDArray* mean = INPUT_VARIABLE(1); // [c]
NDArray* variance = INPUT_VARIABLE(2); // [c]

390
libnd4j/include/ops/declarable/platform/mkldnn/conv2d.cpp
View File

@ -17,6 +17,7 @@
//
// @author saudet
// @author raver119@gmail.com
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <ops/declarable/PlatformHelper.h>
@ -33,6 +34,298 @@ namespace nd4j {
namespace ops {
namespace platforms {
//////////////////////////////////////////////////////////////////////
static void conv2dMKLDNN(const NDArray *input, const NDArray *weights,
const NDArray *bias, NDArray *output,
const int kH, const int kW, const int sH, const int sW, const int pH, const int pW, const int dH, const int dW,
const int paddingMode, const int isNCHW) {
// weights [kH, kW, iC, oC], we'll perform permutation since mkl support [oC, iC, kH, kW]
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
const int pWSame = (paddingMode == 2 && dW > 1) ? ((oW - 1) * sW + (kW - 1) * dW + 1 - iW) / 2 : pW; // dH == 1 for causal mode in conv1d
dnnl::memory::dims strides = { sH, sW };
dnnl::memory::dims padding = { pH, pW };
dnnl::memory::dims padding_r = { (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pWSame };
dnnl::memory::dims dilation = { dH-1, dW-1};
auto xzFrmat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
dnnl::memory::dims wDims = {oC, iC, kH, kW};
dnnl::memory::dims zDims = {bS, oC, oH, oW};
auto type = dnnl::memory::data_type::f32;
// memory descriptors for arrays
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, type, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(3); // permute [kH, kW, iC, oC] -> [oC, iC, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(2);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
// bias
dnnl::memory::desc b_mkl_md;
if(bias != nullptr)
b_mkl_md = dnnl::memory::desc({oC}, type, dnnl::memory::format_tag::x);
// output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(zDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(zDims, type, xzFrmat);
if(output->ews() != 1 || output->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(1);
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(2);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(3);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// operation primitive description
dnnl::convolution_forward::desc op_desc(dnnl::prop_kind::forward_inference, dnnl::algorithm::convolution_auto, x_mkl_md, w_mkl_md, b_mkl_md, z_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_forward::primitive_desc op_prim_desc(op_desc, engine);
// arguments (memory buffers) necessary for calculations
std::unordered_map<int, dnnl::memory> args;
dnnl::stream stream(engine);
// provide memory buffers and check whether reorder is required
// input
auto x_user_mem = dnnl::memory(x_user_md, engine, input->getBuffer());
const bool xReorder = op_prim_desc.src_desc() != x_user_mem.get_desc();
auto x_mkl_mem = xReorder ? dnnl::memory(op_prim_desc.src_desc(), engine) : x_user_mem;
if (xReorder)
dnnl::reorder(x_user_mem, x_mkl_mem).execute(stream, x_user_mem, x_mkl_mem);
args[DNNL_ARG_SRC] = x_mkl_mem;
// weights
auto w_user_mem = dnnl::memory(w_user_md, engine, weights->getBuffer());
const bool wReorder = op_prim_desc.weights_desc() != w_user_mem.get_desc();
auto w_mkl_mem = wReorder ? dnnl::memory(op_prim_desc.weights_desc(), engine) : w_user_mem;
if (wReorder)
dnnl::reorder(w_user_mem, w_mkl_mem).execute(stream, w_user_mem, w_mkl_mem);
args[DNNL_ARG_WEIGHTS] = w_mkl_mem;
// bias
if(bias != nullptr) {
auto b_mkl_mem = dnnl::memory(b_mkl_md, engine, bias->getBuffer());
args[DNNL_ARG_BIAS] = b_mkl_mem;
}
// output
auto z_user_mem = dnnl::memory(z_user_md, engine, output->getBuffer());
const bool zReorder = op_prim_desc.dst_desc() != z_user_mem.get_desc();
auto z_mkl_mem = zReorder ? dnnl::memory(op_prim_desc.dst_desc(), engine) : z_user_mem;
args[DNNL_ARG_DST] = z_mkl_mem;
// run calculations
dnnl::convolution_forward(op_prim_desc).execute(stream, args);
// reorder outputs if necessary
if (zReorder)
dnnl::reorder(z_mkl_mem, z_user_mem).execute(stream, z_mkl_mem, z_user_mem);
stream.wait();
// shape::printArray(z_mkl_mem.map_data<float>(),8);
}
//////////////////////////////////////////////////////////////////////
static void conv2dBpMKLDNN(const NDArray *input, const NDArray *weights, const NDArray *bias, const NDArray *gradO,
NDArray *gradI, NDArray *gradW, NDArray *gradB,
const int kH, const int kW, const int sH, const int sW, const int pH, const int pW, const int dH, const int dW,
const int paddingMode, const int isNCHW) {
// weights/gradW [kH, kW, iC, oC], we'll perform permutation since mkl support [oC, iC, kH, kW]
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
const int pWSame = (paddingMode == 2 && dW > 1) ? ((oW - 1) * sW + (kW - 1) * dW + 1 - iW) / 2 : pW; // dH == 1 for causal mode in conv1d
dnnl::memory::dims strides = { sH, sW };
dnnl::memory::dims padding = { pH, pW };
dnnl::memory::dims padding_r = { (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pWSame };
dnnl::memory::dims dilation = { dH-1, dW-1};
auto xzFrmat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
dnnl::memory::dims wDims = {oC, iC, kH, kW};
dnnl::memory::dims zDims = {bS, oC, oH, oW};
auto type = dnnl::memory::data_type::f32;
// memory descriptors for arrays
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, type, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(3); // permute [kH, kW, iC, oC] -> [oC, iC, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(2);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, type, xzFrmat);
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(2);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(3);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(2);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(3);
}
// gradW
dnnl::memory::desc gradW_mkl_md = dnnl::memory::desc(wDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradW_user_md = dnnl::memory::desc(wDims, type, wFormat);
gradW_user_md.data.format_kind = dnnl_blocked; // overrides format
gradW_user_md.data.format_desc.blocking.strides[0] = gradW->strideAt(3); // permute [kH, kW, iC, oC] -> [oC, iC, kH, kW]
gradW_user_md.data.format_desc.blocking.strides[1] = gradW->strideAt(2);
gradW_user_md.data.format_desc.blocking.strides[2] = gradW->strideAt(0);
gradW_user_md.data.format_desc.blocking.strides[3] = gradW->strideAt(1);
// gradB
dnnl::memory::desc gradB_mkl_md;
if(gradB != nullptr)
gradB_mkl_md = dnnl::memory::desc({oC}, type, dnnl::memory::format_tag::x);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// forward primitive description
dnnl::convolution_forward::desc op_ff_desc(dnnl::prop_kind::forward_inference, dnnl::algorithm::convolution_auto, x_mkl_md, w_mkl_md, gradB_mkl_md, gradO_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_forward::primitive_desc op_ff_prim_desc(op_ff_desc, engine);
// backward data primitive description
dnnl::convolution_backward_data::desc op_data_bp_desc(dnnl::algorithm::convolution_auto, gradI_mkl_md, w_mkl_md, gradO_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_backward_data::primitive_desc op_data_bp_prim_desc(op_data_bp_desc, engine, op_ff_prim_desc);
// backward weights primitive description
dnnl::convolution_backward_weights::desc op_weights_bp_desc(dnnl::algorithm::convolution_auto, x_mkl_md, gradW_mkl_md, gradB_mkl_md, gradO_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_backward_weights::primitive_desc op_weights_bp_prim_desc(op_weights_bp_desc, engine, op_ff_prim_desc);
// arguments (memory buffers) necessary for calculations
std::unordered_map<int, dnnl::memory> args;
dnnl::stream stream(engine);
// provide memory buffers and check whether reorder is required
// input
auto x_user_mem = dnnl::memory(x_user_md, engine, input->getBuffer());
const bool xReorder = op_weights_bp_prim_desc.src_desc() != x_user_mem.get_desc();
auto x_mkl_mem = xReorder ? dnnl::memory(op_weights_bp_prim_desc.src_desc(), engine) : x_user_mem;
if (xReorder)
dnnl::reorder(x_user_mem, x_mkl_mem).execute(stream, x_user_mem, x_mkl_mem);
args[DNNL_ARG_SRC] = x_mkl_mem;
// weights
auto w_user_mem = dnnl::memory(w_user_md, engine, weights->getBuffer());
const bool wReorder = op_data_bp_prim_desc.weights_desc() != w_user_mem.get_desc();
auto w_mkl_mem = wReorder ? dnnl::memory(op_data_bp_prim_desc.weights_desc(), engine) : w_user_mem;
if (wReorder)
dnnl::reorder(w_user_mem, w_mkl_mem).execute(stream, w_user_mem, w_mkl_mem);
args[DNNL_ARG_WEIGHTS] = w_mkl_mem;
// gradO
auto gradO_user_mem = dnnl::memory(gradO_user_md, engine, gradO->getBuffer());
const bool gradOReorderW = op_weights_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
const bool gradOReorderD = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_memW = gradOReorderW ? dnnl::memory(op_weights_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
auto gradO_mkl_memD = gradOReorderD ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorderW)
dnnl::reorder(gradO_user_mem, gradO_mkl_memW).execute(stream, gradO_user_mem, gradO_mkl_memW);
if (gradOReorderD)
dnnl::reorder(gradO_user_mem, gradO_mkl_memD).execute(stream, gradO_user_mem, gradO_mkl_memD);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memD;
// gradI
auto gradI_user_mem = dnnl::memory(gradI_user_md, engine, gradI->getBuffer());
const bool gradIReorder = op_data_bp_prim_desc.diff_src_desc() != gradI_user_mem.get_desc();
auto gradI_mkl_mem = gradIReorder ? dnnl::memory(op_data_bp_prim_desc.diff_src_desc(), engine) : gradI_user_mem;
args[DNNL_ARG_DIFF_SRC] = gradI_mkl_mem;
// gradW
auto gradW_user_mem = dnnl::memory(gradW_user_md, engine, gradW->getBuffer());
const bool gradWReorder = op_weights_bp_prim_desc.diff_weights_desc() != gradW_user_mem.get_desc();
auto gradW_mkl_mem = gradWReorder ? dnnl::memory(op_weights_bp_prim_desc.diff_weights_desc(), engine) : gradW_user_mem;
args[DNNL_ARG_DIFF_WEIGHTS] = gradW_mkl_mem;
// gradB
if(gradB != nullptr) {
auto gradB_mkl_mem = dnnl::memory(gradB_mkl_md, engine, gradB->getBuffer());
args[DNNL_ARG_DIFF_BIAS] = gradB_mkl_mem;
}
// run backward data calculations
dnnl::convolution_backward_data(op_data_bp_prim_desc).execute(stream, args);
if(gradOReorderW || gradOReorderD)
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memW;
// run backward weights calculations
dnnl::convolution_backward_weights(op_weights_bp_prim_desc).execute(stream, args);
// reorder gradI if necessary
if (gradIReorder)
dnnl::reorder(gradI_mkl_mem, gradI_user_mem).execute(stream, gradI_mkl_mem, gradI_user_mem);
if (gradWReorder)
dnnl::reorder(gradW_mkl_mem, gradW_user_mem).execute(stream, gradW_mkl_mem, gradW_user_mem);
stream.wait();
// shape::printArray(z_mkl_mem.map_data<float>(),8);
}
/*
//////////////////////////////////////////////////////////////////////
static void conv2dMKLDNN(nd4j::graph::Context &block, const NDArray *input, const NDArray *weights,
const NDArray *bias, NDArray *output, const int kH, const int kW, const int sH,
@ -46,37 +339,37 @@ static void conv2dMKLDNN(nd4j::graph::Context &block, const NDArray *input, cons
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW, paddingMode);
dnnl_memory_desc_t empty;
dnnl::memory::desc conv_src_md(empty), conv_weights_md(empty), conv_bias_md(empty), conv_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_weights_md(empty), user_bias_md(empty), user_dst_md(empty);
dnnl::memory::desc x_mkl_md(empty), w_mkl_md(empty), b_mkl_md(empty), z_mkl_md(empty);
dnnl::memory::desc x_user_md(empty), w_user_md(empty), b_user_md(empty), z_user_md(empty);
dnnl::memory::dims conv_strides, conv_padding, conv_padding_r, conv_dilation;
dnnl::memory::dims strides, padding, padding_r, dilation;
mkldnnUtils::getMKLDNNMemoryDescConv2d(kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW,
bS, iC, iH, iW, oC, oH, oW, input, nullptr, weights, nullptr,
bias, output,
&conv_src_md, nullptr, &conv_weights_md, nullptr,
&conv_bias_md, &conv_dst_md,
&user_src_md, nullptr, &user_weights_md, nullptr,
&user_bias_md, &user_dst_md,
conv_strides, conv_padding, conv_padding_r, conv_dilation);
&x_mkl_md, nullptr, &w_mkl_md, nullptr,
&b_mkl_md, &z_mkl_md,
&x_user_md, nullptr, &w_user_md, nullptr,
&b_user_md, &z_user_md,
strides, padding, padding_r, dilation);
auto conv_desc = bias != nullptr ? convolution_forward::desc(prop_kind::forward,
algorithm::convolution_auto, conv_src_md,
conv_weights_md, conv_bias_md,
conv_dst_md, conv_strides, conv_dilation, conv_padding,
conv_padding_r)
algorithm::convolution_auto, x_mkl_md,
w_mkl_md, b_mkl_md,
z_mkl_md, strides, dilation, padding,
padding_r)
: convolution_forward::desc(prop_kind::forward,
algorithm::convolution_auto, conv_src_md,
conv_weights_md,
conv_dst_md, conv_strides, conv_dilation, conv_padding,
conv_padding_r);
algorithm::convolution_auto, x_mkl_md,
w_mkl_md,
z_mkl_md, strides, dilation, padding,
padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
auto conv_prim_desc = convolution_forward::primitive_desc(conv_desc, engine);
auto user_src_memory = dnnl::memory(user_src_md, engine, const_cast<NDArray *>(input)->buffer());
auto user_weights_memory = dnnl::memory(user_weights_md, engine,
auto user_src_memory = dnnl::memory(x_user_md, engine, const_cast<NDArray *>(input)->buffer());
auto user_weights_memory = dnnl::memory(w_user_md, engine,
const_cast<NDArray *>(weights)->buffer());
auto user_dst_memory = dnnl::memory(user_dst_md, engine, output->buffer());
auto user_dst_memory = dnnl::memory(z_user_md, engine, output->buffer());
auto conv_src_memory = user_src_memory;
if (conv_prim_desc.src_desc() != user_src_memory.get_desc()) {
conv_src_memory = dnnl::memory(conv_prim_desc.src_desc(), engine);
@ -239,13 +532,16 @@ static void conv2dBpMKLDNN(nd4j::graph::Context &block,
}
}
*/
//////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(conv2d, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto weights = INPUT_VARIABLE(1); // [kH, kW, iC, oC] always
auto bias = block.width() > 2 ? INPUT_VARIABLE(2) : nullptr; // [oC]
auto output = OUTPUT_VARIABLE(0); // [bS, oH, oW, oC] (NHWC) or [bS, oC, oH, oW] (NCHW)
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto weights = INPUT_VARIABLE(1); // [kH, kW, iC, oC] always
auto bias = block.width() > 2 ? INPUT_VARIABLE(2) : nullptr; // [oC]
auto output = OUTPUT_VARIABLE(0); // [bS, oH, oW, oC] (NHWC) or [bS, oC, oH, oW] (NCHW)
int sH = INT_ARG(2); // strides height
int sW = INT_ARG(3); // strides width
@ -254,16 +550,28 @@ PLATFORM_IMPL(conv2d, ENGINE_CPU) {
int dH = INT_ARG(6); // dilations height
int dW = INT_ARG(7); // dilations width
int paddingMode = INT_ARG(8); // 0-VALID, 1-SAME
bool isNCHW = block.getIArguments()->size() > 9 ? !INT_ARG(9) : 1; // INT_ARG(9): 0-NCHW, 1-NHWC
bool isNCHW = block.getIArguments()->size() > 9 ? !INT_ARG(9) : 1; // INT_ARG(9): 0-NCHW, 1-NHWC
int kH = INT_ARG(0) > 0 ? INT_ARG(0) : static_cast<int>(weights->sizeAt(0)); // filter(kernel) height
int kW = INT_ARG(1) > 0 ? INT_ARG(1) : static_cast<int>(weights->sizeAt(1)); // filter(kernel) width
conv2dMKLDNN(block, input, weights, bias, output, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW);
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW, paddingMode);
std::vector<Nd4jLong> expectedWeightsShape = {kH, kW, iC, oC};
REQUIRE_TRUE(weights->isSameShape(expectedWeightsShape), 0, "CONV2D MKLDNN OP: wrong shape of weights array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedWeightsShape).c_str(), ShapeUtils::shapeAsString(weights).c_str());
if (bias)
REQUIRE_TRUE(bias->rankOf() <= 2 && oC == bias->lengthOf(), 0, "CONV2D MKLDNN OP: wrong shape of array with biases, expected rank, length: <=2, %i, but got %i, %i instead !", oC, bias->rankOf(), bias->lengthOf());
conv2dMKLDNN(input, weights, bias, output, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW);
return Status::OK();
}
PLATFORM_CHECK(conv2d, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto weights = INPUT_VARIABLE(1);
@ -276,10 +584,10 @@ PLATFORM_CHECK(conv2d, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(conv2d_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto weights = INPUT_VARIABLE(1); // [kH, kW, iC, oC] always
auto bias = block.width() > 3 ? INPUT_VARIABLE(2) : nullptr; // [oC]
auto gradO = block.width() > 3 ? INPUT_VARIABLE(3) : INPUT_VARIABLE(2); // [bS, oH, oW, oC] (NHWC) or [bS, oC, oH, oW] (NCHW), epsilon_next
auto bias = block.width() > 3 ? INPUT_VARIABLE(2) : nullptr; // [oC]
auto gradO = block.width() > 3 ? INPUT_VARIABLE(3) : INPUT_VARIABLE(2); // [bS, oH, oW, oC] (NHWC) or [bS, oC, oH, oW] (NCHW), epsilon_next
auto gradI = OUTPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW), epsilon
auto gradW = OUTPUT_VARIABLE(1); // [kH, kW, iC, oC] always
@ -293,19 +601,33 @@ PLATFORM_IMPL(conv2d_bp, ENGINE_CPU) {
int pW = INT_ARG(5); // paddings width
int dH = INT_ARG(6); // dilations height
int dW = INT_ARG(7); // dilations width
int paddingMode = INT_ARG(8); // 0-VALID, 1-SAME
int isNCHW = block.getIArguments()->size() > 9 ? !INT_ARG(9) : 1; // INT_ARG(9): 0-NCHW, 1-NHWC
int paddingMode = INT_ARG(8); // 0-VALID, 1-SAME
int isNCHW = block.getIArguments()->size() > 9 ? !INT_ARG(9) : 1; // INT_ARG(9): 0-NCHW, 1-NHWC
REQUIRE_TRUE(input->rankOf() == 4, 0,"CUSTOM CONV2D_BP OP: rank of input array must be equal to 4, but got %i instead !",input->rankOf());
REQUIRE_TRUE(weights->rankOf() == 4, 0,"CUSTOM CONV2D_BP OP: rank of weights array must be equal to 4, but got %i instead !",weights->rankOf());
REQUIRE_TRUE(gradO->rankOf() == 4, 0,"CUSTOM CONV2D_BP OP: rank of output's gradients (next epsilon) array must be equal to 4, but got %i instead !",gradO->rankOf());
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
conv2dBpMKLDNN(block, input, weights, bias, gradO, gradI, gradW, gradB, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW);
int trueoH, trueoW; // true output height, width
ConvolutionUtils::calcOutSizePool2D(trueoH, trueoW, kH, kW, sH, sW, pH, pW, dH, dW, iH, iW, paddingMode);
if(paddingMode) // SAME
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW, paddingMode);
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx({bS,oC,trueoH,trueoW, 0,indIOioC,indOoH,indOoH+1});
std::vector<Nd4jLong> expectedWeightsShape = {kH, kW, iC, oC};
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "CONV2D_BP MKLDNN OP: wrong shape of output gradients (next epsilon) array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(weights->isSameShape(expectedWeightsShape), 0, "CONV2D_BP MKLDNN OP: wrong shape of weights array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedWeightsShape).c_str(), ShapeUtils::shapeAsString(weights).c_str());
if(bias)
REQUIRE_TRUE(bias->rankOf() <= 2 && oC == bias->lengthOf(), 0, "CONV2D_BP MKLDNN OP: wrong shape of array with biases, expected rank, length: <=2, %i, but got %i, %i instead !", oC, bias->rankOf(), bias->lengthOf());
conv2dBpMKLDNN(input, weights, bias, gradO, gradI, gradW, gradB, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW);
return Status::OK();
}
PLATFORM_CHECK(conv2d_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NHWC) or [bS, iC, iH, iW] (NCHW)
auto weights = INPUT_VARIABLE(1); // [kH, kW, iC, oC] always
auto bias = block.width() > 3 ? INPUT_VARIABLE(2) : nullptr; // [oC]

326
libnd4j/include/ops/declarable/platform/mkldnn/conv3d.cpp
View File

@ -33,6 +33,314 @@ namespace nd4j {
namespace ops {
namespace platforms {
//////////////////////////////////////////////////////////////////////
static void conv3dMKLDNN(const NDArray *input, const NDArray *weights,
const NDArray *bias, NDArray *output,
const int kD, const int kH, const int kW,
const int sD, const int sH, const int sW,
const int pD, const int pH, const int pW,
const int dD, const int dH, const int dW,
const int paddingMode, const int isNCDHW) {
// weights [kD, kH, kW, iC, oC], we'll perform permutation since mkl support [oC, iC, kD, kH, kW]
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
// const int pWSame = (paddingMode == 2 && dW > 1) ? ((oW - 1) * sW + (kW - 1) * dW + 1 - iW) / 2 : pW; // dH == 1 for causal mode in conv1d
dnnl::memory::dims strides = {sD, sH, sW};
dnnl::memory::dims padding = {pD, pH, pW};
// dnnl::memory::dims padding_r = { (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pWSame };
dnnl::memory::dims padding_r = {(oD - 1) * sD - iD + kD - pD, (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pW};
dnnl::memory::dims dilation = {dD-1, dH-1, dW-1};
auto xzFrmat = isNCDHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oidhw;
dnnl::memory::dims xDims = {bS, iC, iD, iH, iW};
dnnl::memory::dims wDims = {oC, iC, kD, kH, kW};
dnnl::memory::dims zDims = {bS, oC, oD, oH, oW};
auto type = dnnl::memory::data_type::f32;
// memory descriptors for arrays
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
x_user_md.data.format_desc.blocking.strides[4] = input->strideAt(4);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, type, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(4); // permute [kD, kH, kW, iC, oC] -> [oC, iC, kD, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(3);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
w_user_md.data.format_desc.blocking.strides[4] = weights->strideAt(2);
// bias
dnnl::memory::desc b_mkl_md;
if(bias != nullptr)
b_mkl_md = dnnl::memory::desc({oC}, type, dnnl::memory::format_tag::x);
// output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(zDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(zDims, type, xzFrmat);
if(output->ews() != 1 || output->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(1);
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(2);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(3);
z_user_md.data.format_desc.blocking.strides[4] = output->strideAt(4);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// operation primitive description
dnnl::convolution_forward::desc op_desc(dnnl::prop_kind::forward_inference, dnnl::algorithm::convolution_auto, x_mkl_md, w_mkl_md, b_mkl_md, z_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_forward::primitive_desc op_prim_desc(op_desc, engine);
// arguments (memory buffers) necessary for calculations
std::unordered_map<int, dnnl::memory> args;
dnnl::stream stream(engine);
// provide memory buffers and check whether reorder is required
// input
auto x_user_mem = dnnl::memory(x_user_md, engine, input->getBuffer());
const bool xReorder = op_prim_desc.src_desc() != x_user_mem.get_desc();
auto x_mkl_mem = xReorder ? dnnl::memory(op_prim_desc.src_desc(), engine) : x_user_mem;
if (xReorder)
dnnl::reorder(x_user_mem, x_mkl_mem).execute(stream, x_user_mem, x_mkl_mem);
args[DNNL_ARG_SRC] = x_mkl_mem;
// weights
auto w_user_mem = dnnl::memory(w_user_md, engine, weights->getBuffer());
const bool wReorder = op_prim_desc.weights_desc() != w_user_mem.get_desc();
auto w_mkl_mem = wReorder ? dnnl::memory(op_prim_desc.weights_desc(), engine) : w_user_mem;
if (wReorder)
dnnl::reorder(w_user_mem, w_mkl_mem).execute(stream, w_user_mem, w_mkl_mem);
args[DNNL_ARG_WEIGHTS] = w_mkl_mem;
// bias
if(bias != nullptr) {
auto b_mkl_mem = dnnl::memory(b_mkl_md, engine, bias->getBuffer());
args[DNNL_ARG_BIAS] = b_mkl_mem;
}
// output
auto z_user_mem = dnnl::memory(z_user_md, engine, output->getBuffer());
const bool zReorder = op_prim_desc.dst_desc() != z_user_mem.get_desc();
auto z_mkl_mem = zReorder ? dnnl::memory(op_prim_desc.dst_desc(), engine) : z_user_mem;
args[DNNL_ARG_DST] = z_mkl_mem;
// run calculations
dnnl::convolution_forward(op_prim_desc).execute(stream, args);
// reorder outputs if necessary
if (zReorder)
dnnl::reorder(z_mkl_mem, z_user_mem).execute(stream, z_mkl_mem, z_user_mem);
stream.wait();
}
//////////////////////////////////////////////////////////////////////
static void conv3dBpMKLDNN(const NDArray *input, const NDArray *weights, const NDArray *bias, const NDArray *gradO,
NDArray *gradI, NDArray *gradW, NDArray *gradB,
const int kD, const int kH, const int kW,
const int sD, const int sH, const int sW,
const int pD, const int pH, const int pW,
const int dD, const int dH, const int dW,
const int paddingMode, const int isNCDHW) {
// weights/gradW [kD, kH, kW, iC, oC], we'll perform permutation since mkl support [oC, iC, kD, kH, kW]
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
// const int pWSame = (paddingMode == 2 && dW > 1) ? ((oW - 1) * sW + (kW - 1) * dW + 1 - iW) / 2 : pW; // dH == 1 for causal mode in conv1d
dnnl::memory::dims strides = {sD, sH, sW};
dnnl::memory::dims padding = {pD, pH, pW};
// dnnl::memory::dims padding_r = { (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pWSame };
dnnl::memory::dims padding_r = {(oD - 1) * sD - iD + kD - pD, (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pW};
dnnl::memory::dims dilation = {dD-1, dH-1, dW-1};
auto xzFrmat = isNCDHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oidhw;
dnnl::memory::dims xDims = {bS, iC, iD, iH, iW};
dnnl::memory::dims wDims = {oC, iC, kD, kH, kW};
dnnl::memory::dims zDims = {bS, oC, oD, oH, oW};
auto type = dnnl::memory::data_type::f32;
// memory descriptors for arrays
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
x_user_md.data.format_desc.blocking.strides[4] = input->strideAt(4);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, type, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(4); // permute [kD, kH, kW, iC, oC] -> [oC, iC, kD, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(3);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
w_user_md.data.format_desc.blocking.strides[4] = weights->strideAt(2);
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, type, xzFrmat);
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(2);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(3);
gradO_user_md.data.format_desc.blocking.strides[4] = gradO->strideAt(4);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(2);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(3);
gradI_user_md.data.format_desc.blocking.strides[4] = gradI->strideAt(4);
}
// gradW
dnnl::memory::desc gradW_mkl_md = dnnl::memory::desc(wDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradW_user_md = dnnl::memory::desc(wDims, type, wFormat);
gradW_user_md.data.format_kind = dnnl_blocked; // overrides format
gradW_user_md.data.format_desc.blocking.strides[0] = gradW->strideAt(4); // permute [kD, kH, kW, iC, oC] -> [oC, iC, kD, kH, kW]
gradW_user_md.data.format_desc.blocking.strides[1] = gradW->strideAt(3);
gradW_user_md.data.format_desc.blocking.strides[2] = gradW->strideAt(0);
gradW_user_md.data.format_desc.blocking.strides[3] = gradW->strideAt(1);
gradW_user_md.data.format_desc.blocking.strides[4] = gradW->strideAt(2);
// gradB
dnnl::memory::desc gradB_mkl_md;
if(gradB != nullptr)
gradB_mkl_md = dnnl::memory::desc({oC}, type, dnnl::memory::format_tag::x);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// forward primitive description
dnnl::convolution_forward::desc op_ff_desc(dnnl::prop_kind::forward_inference, dnnl::algorithm::convolution_auto, x_mkl_md, w_mkl_md, gradB_mkl_md, gradO_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_forward::primitive_desc op_ff_prim_desc(op_ff_desc, engine);
// backward data primitive description
dnnl::convolution_backward_data::desc op_data_bp_desc(dnnl::algorithm::convolution_auto, gradI_mkl_md, w_mkl_md, gradO_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_backward_data::primitive_desc op_data_bp_prim_desc(op_data_bp_desc, engine, op_ff_prim_desc);
// backward weights primitive description
dnnl::convolution_backward_weights::desc op_weights_bp_desc(dnnl::algorithm::convolution_auto, x_mkl_md, gradW_mkl_md, gradB_mkl_md, gradO_mkl_md, strides, dilation, padding, padding_r);
dnnl::convolution_backward_weights::primitive_desc op_weights_bp_prim_desc(op_weights_bp_desc, engine, op_ff_prim_desc);
// arguments (memory buffers) necessary for calculations
std::unordered_map<int, dnnl::memory> args;
dnnl::stream stream(engine);
// provide memory buffers and check whether reorder is required
// input
auto x_user_mem = dnnl::memory(x_user_md, engine, input->getBuffer());
const bool xReorder = op_weights_bp_prim_desc.src_desc() != x_user_mem.get_desc();
auto x_mkl_mem = xReorder ? dnnl::memory(op_weights_bp_prim_desc.src_desc(), engine) : x_user_mem;
if (xReorder)
dnnl::reorder(x_user_mem, x_mkl_mem).execute(stream, x_user_mem, x_mkl_mem);
args[DNNL_ARG_SRC] = x_mkl_mem;
// weights
auto w_user_mem = dnnl::memory(w_user_md, engine, weights->getBuffer());
const bool wReorder = op_data_bp_prim_desc.weights_desc() != w_user_mem.get_desc();
auto w_mkl_mem = wReorder ? dnnl::memory(op_data_bp_prim_desc.weights_desc(), engine) : w_user_mem;
if (wReorder)
dnnl::reorder(w_user_mem, w_mkl_mem).execute(stream, w_user_mem, w_mkl_mem);
args[DNNL_ARG_WEIGHTS] = w_mkl_mem;
// gradO
auto gradO_user_mem = dnnl::memory(gradO_user_md, engine, gradO->getBuffer());
const bool gradOReorderW = op_weights_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
const bool gradOReorderD = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_memW = gradOReorderW ? dnnl::memory(op_weights_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
auto gradO_mkl_memD = gradOReorderD ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorderW)
dnnl::reorder(gradO_user_mem, gradO_mkl_memW).execute(stream, gradO_user_mem, gradO_mkl_memW);
if (gradOReorderD)
dnnl::reorder(gradO_user_mem, gradO_mkl_memD).execute(stream, gradO_user_mem, gradO_mkl_memD);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memD;
// gradI
auto gradI_user_mem = dnnl::memory(gradI_user_md, engine, gradI->getBuffer());
const bool gradIReorder = op_data_bp_prim_desc.diff_src_desc() != gradI_user_mem.get_desc();
auto gradI_mkl_mem = gradIReorder ? dnnl::memory(op_data_bp_prim_desc.diff_src_desc(), engine) : gradI_user_mem;
args[DNNL_ARG_DIFF_SRC] = gradI_mkl_mem;
// gradW
auto gradW_user_mem = dnnl::memory(gradW_user_md, engine, gradW->getBuffer());
const bool gradWReorder = op_weights_bp_prim_desc.diff_weights_desc() != gradW_user_mem.get_desc();
auto gradW_mkl_mem = gradWReorder ? dnnl::memory(op_weights_bp_prim_desc.diff_weights_desc(), engine) : gradW_user_mem;
args[DNNL_ARG_DIFF_WEIGHTS] = gradW_mkl_mem;
// gradB
if(gradB != nullptr) {
auto gradB_mkl_mem = dnnl::memory(gradB_mkl_md, engine, gradB->getBuffer());
args[DNNL_ARG_DIFF_BIAS] = gradB_mkl_mem;
}
// run backward data calculations
dnnl::convolution_backward_data(op_data_bp_prim_desc).execute(stream, args);
if(gradOReorderW || gradOReorderD)
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memW;
// run backward weights calculations
dnnl::convolution_backward_weights(op_weights_bp_prim_desc).execute(stream, args);
// reorder gradI if necessary
if (gradIReorder)
dnnl::reorder(gradI_mkl_mem, gradI_user_mem).execute(stream, gradI_mkl_mem, gradI_user_mem);
if (gradWReorder)
dnnl::reorder(gradW_mkl_mem, gradW_user_mem).execute(stream, gradW_mkl_mem, gradW_user_mem);
stream.wait();
// shape::printArray(z_mkl_mem.map_data<float>(),8);
}
/*
//////////////////////////////////////////////////////////////////////
static void conv3dMKLDNN(nd4j::graph::Context &block,
const NDArray *input, const NDArray *weights, const NDArray *bias,
@ -225,6 +533,7 @@ static void conv3dBpMKLDNN(nd4j::graph::Context &block,
reorder(convI_src_memory, userI_src_memory).execute(stream, convI_src_memory, userI_src_memory);
}
}
*/
//////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(conv3dnew, ENGINE_CPU) {
@ -256,15 +565,15 @@ PLATFORM_IMPL(conv3dnew, ENGINE_CPU) {
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
std::string expectedWeightsShape = ShapeUtils::shapeAsString({kD, kH, kW, iC, oC});
REQUIRE_TRUE(expectedWeightsShape == ShapeUtils::shapeAsString(weights), 0, "CUSTOM CONV3D MKLDNN OP: wrong shape of weights array, expected is %s, but got %s instead !", expectedWeightsShape.c_str(), ShapeUtils::shapeAsString(weights).c_str());
std::vector<Nd4jLong> expectedWeightsShape = {kD, kH, kW, iC, oC};
REQUIRE_TRUE(weights->isSameShape(expectedWeightsShape), 0, "CUSTOM CONV3D MKLDNN OP: wrong shape of weights array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedWeightsShape).c_str(), ShapeUtils::shapeAsString(weights).c_str());
if (bias)
REQUIRE_TRUE(bias->rankOf() <= 2 && oC == bias->lengthOf(), 0, "CUSTOM CONV3D MKLDNN OP: wrong shape of array with biases, expected rank, length: <=2, %i, but got %i, %i instead !", oC, bias->rankOf(), bias->lengthOf());
if (paddingMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
conv3dMKLDNN(block, input, weights, bias, output, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, paddingMode, isNCDHW);
conv3dMKLDNN(input, weights, bias, output, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, paddingMode, isNCDHW);
return Status::OK();
}
@ -280,6 +589,7 @@ PLATFORM_CHECK(conv3dnew, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(conv3dnew_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto weights = INPUT_VARIABLE(1); // [kD, kH, kW, iC, oC] always
auto bias = block.width() > 3 ? INPUT_VARIABLE(2) : nullptr; // [oC]
@ -318,14 +628,14 @@ PLATFORM_IMPL(conv3dnew_bp, ENGINE_CPU) {
int trueoD, trueoH, trueoW; // true output depth/height/width
ConvolutionUtils::calcOutSizePool3D(trueoD, trueoH, trueoW, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, iD, iH, iW, paddingMode);
std::string expectedGradOShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx( {bS, oC, trueoD, trueoH, trueoW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
std::string expectedWeightsShape = ShapeUtils::shapeAsString({kD, kH, kW, iC, oC});
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0, "CUSTOM CONV3D_BP OP: wrong shape of output gradients (next epsilon) array, expected is %s, but got %s instead !", expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedWeightsShape == ShapeUtils::shapeAsString(weights), 0, "CUSTOM CONV3D_BP OP: wrong shape of weights array, expected is %s, but got %s instead !", expectedWeightsShape.c_str(), ShapeUtils::shapeAsString(weights).c_str());
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx( {bS, oC, trueoD, trueoH, trueoW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2});
std::vector<Nd4jLong> expectedWeightsShape = {kD, kH, kW, iC, oC};
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "CUSTOM CONV3D_BP OP: wrong shape of output gradients (next epsilon) array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(weights->isSameShape(expectedWeightsShape), 0, "CUSTOM CONV3D_BP OP: wrong shape of weights array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedWeightsShape).c_str(), ShapeUtils::shapeAsString(weights).c_str());
if (bias)
REQUIRE_TRUE(bias->rankOf() <= 2 && oC == bias->lengthOf(), 0, "CUSTOM CONV3D_BP OP: wrong shape of array with biases, expected rank, length: <=2, %i, but got %i, %i instead !", oC, bias->rankOf(), bias->lengthOf());
conv3dBpMKLDNN(block, input, weights, bias, gradO, gradI, gradW, gradB, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, paddingMode, isNCDHW);
conv3dBpMKLDNN(input, weights, bias, gradO, gradI, gradW, gradB, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, paddingMode, isNCDHW);
return Status::OK();
}

167
libnd4j/include/ops/declarable/platform/mkldnn/deconv2d.cpp
View File

@ -34,17 +34,13 @@ namespace platforms {
//////////////////////////////////////////////////////////////////////////
static void deconv2dMKLDNN(const NDArray* input, const NDArray* weights, const NDArray* bias, NDArray* output,
const int kH, const int kW, const int sH, const int sW, const int pH, const int pW, const int dH, const int dW,
const int paddingMode) {
const int paddingMode, const bool isNCHW) {
// input [bS, iC, iH, iW] nchw, mkl doesn't support format nhwc
// weights [oC, iC, kH, kW] always, mkl doesn't support weights format [kH, kW, oC, iC]
// bias [oC], may be nullptr
// output [bS, oC, oH, oW] nchw, mkl doesn't support format nhwc
// weights [oC, iC, kH, kW] always, mkl doesn't support [kH, kW, oC, iC], so we'll perform permutation
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(true, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH);
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH);
dnnl::memory::dims strides = { sH, sW };
dnnl::memory::dims padding = { pH, pW };
@ -80,8 +76,7 @@ static void deconv2dMKLDNN(const NDArray* input, const NDArray* weights, const N
else
zType = dnnl::memory::data_type::s32;
dnnl::memory::format_tag xFormat = dnnl::memory::format_tag::nchw; // isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag xFormat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
@ -93,20 +88,22 @@ static void deconv2dMKLDNN(const NDArray* input, const NDArray* weights, const N
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, xType, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xFormat);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->stridesOf()[0];
x_user_md.data.format_desc.blocking.strides[1] = input->stridesOf()[1];
x_user_md.data.format_desc.blocking.strides[2] = input->stridesOf()[2];
x_user_md.data.format_desc.blocking.strides[3] = input->stridesOf()[3];
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, wType, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->stridesOf()[0];
w_user_md.data.format_desc.blocking.strides[1] = weights->stridesOf()[1];
w_user_md.data.format_desc.blocking.strides[2] = weights->stridesOf()[2];
w_user_md.data.format_desc.blocking.strides[3] = weights->stridesOf()[3];
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(2); // [kH, kW, oC, iC] -> [oC, iC, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(3);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
// bias
dnnl::memory::desc b_mkl_md;
@ -116,11 +113,13 @@ static void deconv2dMKLDNN(const NDArray* input, const NDArray* weights, const N
// output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(zDims, zType, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(zDims, zType, xFormat);
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->stridesOf()[0];
z_user_md.data.format_desc.blocking.strides[1] = output->stridesOf()[1];
z_user_md.data.format_desc.blocking.strides[2] = output->stridesOf()[2];
z_user_md.data.format_desc.blocking.strides[3] = output->stridesOf()[3];
if(output->ews() != 1 || output->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(1);
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(2);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(3);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
@ -179,21 +178,19 @@ static void deconv2dMKLDNN(const NDArray* input, const NDArray* weights, const N
//////////////////////////////////////////////////////////////////////////
static void deconv2dBpMKLDNN(const NDArray* input, const NDArray* weights, const NDArray* gradO, NDArray* gradI, NDArray* gradW, NDArray* gradB,
const int kH, const int kW, const int sH, const int sW, const int pH, const int pW, const int dH, const int dW,
const int paddingMode) {
const int paddingMode, const bool isNCHW) {
// input and gradI [bS, iC, iH, iW], mkl doesn't support ndhwc format
// weights and gradW [oC, iC, kH, kW] always, mkl doesn't support weights format [kH, kW, oC, iC]
// gradB [oC], may be nullptr
// gradO [bS, oC, oH, oW]
// weights and gradW [oC, iC, kH, kW] always, mkl doesn't support [kH, kW, oC, iC], so we'll perform permutation
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(true, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH);
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH);
dnnl::memory::dims strides = { sH, sW };
dnnl::memory::dims padding = { pH, pW };
dnnl::memory::dims padding_r = { (iH - 1) * sH - oH + kH - pH, (iW - 1) * sW - oW + kW - pW };
dnnl::memory::dims dilation = { dH-1, dW-1 };
// input type
dnnl::memory::data_type xType = input->dataType() == DataType::FLOAT32 ? dnnl::memory::data_type::f32 : dnnl::memory::data_type::bf16;
// weights type
@ -207,7 +204,7 @@ static void deconv2dBpMKLDNN(const NDArray* input, const NDArray* weights, const
// gradB type
dnnl::memory::data_type gradBType = gradB != nullptr ? (gradB->dataType() == DataType::FLOAT32 ? dnnl::memory::data_type::f32 : dnnl::memory::data_type::bf16) : dnnl::memory::data_type::f32;
dnnl::memory::format_tag xFormat = dnnl::memory::format_tag::nchw; // isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag xFormat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
@ -219,54 +216,59 @@ static void deconv2dBpMKLDNN(const NDArray* input, const NDArray* weights, const
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, xType, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xFormat);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->stridesOf()[0];
x_user_md.data.format_desc.blocking.strides[1] = input->stridesOf()[1];
x_user_md.data.format_desc.blocking.strides[2] = input->stridesOf()[2];
x_user_md.data.format_desc.blocking.strides[3] = input->stridesOf()[3];
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, wType, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->stridesOf()[0];
w_user_md.data.format_desc.blocking.strides[1] = weights->stridesOf()[1];
w_user_md.data.format_desc.blocking.strides[2] = weights->stridesOf()[2];
w_user_md.data.format_desc.blocking.strides[3] = weights->stridesOf()[3];
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(2); // [kH, kW, oC, iC] -> [oC, iC, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(3);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, gradOType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, gradOType, xFormat);
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->stridesOf()[0];
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->stridesOf()[1];
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->stridesOf()[2];
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->stridesOf()[3];
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(2);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(3);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, gradIType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, gradIType, xFormat);
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->stridesOf()[0];
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->stridesOf()[1];
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->stridesOf()[2];
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->stridesOf()[3];
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(2);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(3);
}
// gradW
dnnl::memory::desc gradW_mkl_md = dnnl::memory::desc(wDims, gradWType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradW_user_md = dnnl::memory::desc(wDims, gradWType, wFormat);
gradW_user_md.data.format_kind = dnnl_blocked; // overrides format
gradW_user_md.data.format_desc.blocking.strides[0] = gradW->stridesOf()[0];
gradW_user_md.data.format_desc.blocking.strides[1] = gradW->stridesOf()[1];
gradW_user_md.data.format_desc.blocking.strides[2] = gradW->stridesOf()[2];
gradW_user_md.data.format_desc.blocking.strides[3] = gradW->stridesOf()[3];
gradW_user_md.data.format_desc.blocking.strides[0] = gradW->strideAt(2); // [kH, kW, oC, iC] -> [oC, iC, kH, kW]
gradW_user_md.data.format_desc.blocking.strides[1] = gradW->strideAt(3);
gradW_user_md.data.format_desc.blocking.strides[2] = gradW->strideAt(0);
gradW_user_md.data.format_desc.blocking.strides[3] = gradW->strideAt(1);
// gradB
dnnl::memory::desc gradB_mkl_md;
if(gradB != nullptr)
gradB_mkl_md = dnnl::memory::desc({oC}, gradBType, dnnl::memory::format_tag::x);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// forward primitive description
@ -306,11 +308,15 @@ static void deconv2dBpMKLDNN(const NDArray* input, const NDArray* weights, const
// gradO
auto gradO_user_mem = dnnl::memory(gradO_user_md, engine, gradO->getBuffer());
const bool gradOReorder = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_mem = gradOReorder ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorder)
dnnl::reorder(gradO_user_mem, gradO_mkl_mem).execute(stream, gradO_user_mem, gradO_mkl_mem);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_mem;
const bool gradOReorderW = op_weights_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
const bool gradOReorderD = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_memW = gradOReorderW ? dnnl::memory(op_weights_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
auto gradO_mkl_memD = gradOReorderD ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorderW)
dnnl::reorder(gradO_user_mem, gradO_mkl_memW).execute(stream, gradO_user_mem, gradO_mkl_memW);
if (gradOReorderD)
dnnl::reorder(gradO_user_mem, gradO_mkl_memD).execute(stream, gradO_user_mem, gradO_mkl_memD);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memD;
// gradI
auto gradI_user_mem = dnnl::memory(gradI_user_md, engine, gradI->getBuffer());
@ -333,6 +339,9 @@ static void deconv2dBpMKLDNN(const NDArray* input, const NDArray* weights, const
// run backward data calculations
dnnl::deconvolution_backward_data(op_data_bp_prim_desc).execute(stream, args);
if(gradOReorderW || gradOReorderD)
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memW;
// run backward weights calculations
dnnl::deconvolution_backward_weights(op_weights_bp_prim_desc).execute(stream, args);
@ -385,23 +394,7 @@ PLATFORM_IMPL(deconv2d, ENGINE_CPU) {
ConvolutionUtils::calcPadding2D(pH, pW, iH, iW, oH, oW, kH, kW, sH, sW, dH, dW);
}
// mkl supports only [oC, iC, kH, kW] format for weights
weights = new NDArray(weights->permute({2,3,0,1})); // [kH, kW, oC, iC] -> [oC, iC, kH, kW]
// mkl supports only NCHW
if(!isNCHW) {
input = new NDArray(input->permute({0,3,1,2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
output = new NDArray(output->permute({0,3,1,2})); // [bS, oH, oW, oC] -> [bS, oC, oH, oW]
}
deconv2dMKLDNN(input, weights, bias, output, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode);
delete weights;
if(!isNCHW) {
delete input;
delete output;
}
deconv2dMKLDNN(input, weights, bias, output, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW);
return Status::OK();
}
@ -477,27 +470,7 @@ PLATFORM_IMPL(deconv2d_bp, ENGINE_CPU) {
ConvolutionUtils::calcPadding2D(pH, pW, iH, iW, oH, oW, kH, kW, sH, sW, dH, dW);
}
// mkl supports only [oC, iC, kH, kW] for weights
weights = new NDArray(weights->permute({2,3,0,1})); // [kH, kW, oC, iC] -> [oC, iC, kH, kW]
gradW = new NDArray(gradW->permute({2,3,0,1})); // [kH, kW, oC, iC] -> [oC, iC, kH, kW]
// mkl supports NCHW format only
if(!isNCHW) {
input = new NDArray(input->permute({0,3,1,2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradI = new NDArray(gradI->permute({0,3,1,2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradO = new NDArray(gradO->permute({0,3,1,2})); // [bS, oH, oW, oC] -> [bS, oC, oH, oW]
}
deconv2dBpMKLDNN(input, weights, gradO, gradI, gradW, gradB, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode);
delete weights;
delete gradW;
if(!isNCHW) {
delete input;
delete gradI;
delete gradO;
}
deconv2dBpMKLDNN(input, weights, gradO, gradI, gradW, gradB, kH, kW, sH, sW, pH, pW, dH, dW, paddingMode, isNCHW);
return Status::OK();
}

74
libnd4j/include/ops/declarable/platform/mkldnn/deconv2d_tf.cpp
View File

@ -33,7 +33,8 @@ namespace platforms {
//////////////////////////////////////////////////////////////////////////
static void deconv2TFdBackPropMKLDNN(const NDArray* weights, const NDArray* gradO, NDArray* gradI,
const int bS, const int iC, const int iH, const int iW, const int oC, const int oH, const int oW,
const int kH, const int kW, const int sH, const int sW, const int pH, const int pW, const int dH, const int dW) {
const int kH, const int kW, const int sH, const int sW, const int pH, const int pW, const int dH, const int dW,
const bool isNCHW) {
// gradI [bS, iH, iW, iC], mkl doesn't support ndhwc format
// weights [oC, iC, kH, kW] always, mkl doesn't support weights format [kH, kW, iC, oC]
@ -51,7 +52,7 @@ static void deconv2TFdBackPropMKLDNN(const NDArray* weights, const NDArray* grad
// gradI type
dnnl::memory::data_type gradIType = gradI->dataType() == DataType::FLOAT32 ? dnnl::memory::data_type::f32 : dnnl::memory::data_type::bf16;
dnnl::memory::format_tag xFormat = dnnl::memory::format_tag::nchw; // isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag xFormat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
@ -67,29 +68,32 @@ static void deconv2TFdBackPropMKLDNN(const NDArray* weights, const NDArray* grad
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, wType, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->stridesOf()[0];
w_user_md.data.format_desc.blocking.strides[1] = weights->stridesOf()[1];
w_user_md.data.format_desc.blocking.strides[2] = weights->stridesOf()[2];
w_user_md.data.format_desc.blocking.strides[3] = weights->stridesOf()[3];
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(3); // permute [kH, kW, iC, oC] -> [oC, iC, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(2);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, gradOType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, gradOType, xFormat);
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->stridesOf()[0];
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->stridesOf()[1];
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->stridesOf()[2];
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->stridesOf()[3];
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(2);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(3);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, gradIType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, gradIType, xFormat);
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->stridesOf()[0];
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->stridesOf()[1];
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->stridesOf()[2];
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->stridesOf()[3];
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(2);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(3);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
@ -166,9 +170,9 @@ PLATFORM_IMPL(deconv2d_tf, ENGINE_CPU) {
const int rank = gradO->rankOf();
REQUIRE_TRUE(weights->rankOf() == rank, 0, "CUSTOM DECONV2D_TF OP: rank of weights array must be equal to 4, but got %i instead !", weights->rankOf());
REQUIRE_TRUE(gradIShape->rankOf() == 1, 0, "CUSTOM DECONV2D_TF OP: rank of array with output shape must be equal to 1, but got %i instead !", gradIShape->rankOf());
REQUIRE_TRUE(gradIShape->lengthOf() == rank, 0, "CUSTOM DECONV2D_TF OP: length of array with output shape must be equal to 4, but got %i instead !", gradIShape->lengthOf());
REQUIRE_TRUE(weights->rankOf() == rank, 0, "CUSTOM DECONV2D_TF MKLDNN OP: rank of weights array must be equal to 4, but got %i instead !", weights->rankOf());
REQUIRE_TRUE(gradIShape->rankOf() == 1, 0, "CUSTOM DECONV2D_TF MKLDNN OP: rank of array with output shape must be equal to 1, but got %i instead !", gradIShape->rankOf());
REQUIRE_TRUE(gradIShape->lengthOf() == rank, 0, "CUSTOM DECONV2D_TF MKLDNN OP: length of array with output shape must be equal to 4, but got %i instead !", gradIShape->lengthOf());
int indIOioC, indIiH, indWoC(3), indOoH;
if(!isNCHW) {
@ -193,29 +197,29 @@ PLATFORM_IMPL(deconv2d_tf, ENGINE_CPU) {
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx({bS,oC,trueoH,trueoW, 0,indIOioC,indOoH,indOoH+1});
std::vector<Nd4jLong> expectedWeightsShape = {kH, kW, iC, oC};
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "CUSTOM DECONV2D_TF OP: wrong shape of input array, basing on array with output shape expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(weights->isSameShape(expectedWeightsShape), 0, "CUSTOM DECONV2D_TF OP: wrong shape of weights array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedWeightsShape).c_str(), ShapeUtils::shapeAsString(weights).c_str());
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "CUSTOM DECONV2D_TF MKLDNN OP: wrong shape of input array, basing on array with output shape expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(weights->isSameShape(expectedWeightsShape), 0, "CUSTOM DECONV2D_TF MKLDNN OP: wrong shape of weights array, expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedWeightsShape).c_str(), ShapeUtils::shapeAsString(weights).c_str());
if(isSameMode) // SAME
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW);
// mkl supports only [oC, iC, kH, kW] for weights
weights = new NDArray(weights->permute({3,2,0,1})); // [kH, kW, iC, oC] -> [oC, iC, kH, kW]
// // mkl supports only [oC, iC, kH, kW] for weights
// weights = new NDArray(weights->permute({3,2,0,1})); // [kH, kW, iC, oC] -> [oC, iC, kH, kW]
// mkl supports NCHW format only
if(!isNCHW) {
gradI = new NDArray(gradI->permute({0,3,1,2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradO = new NDArray(gradO->permute({0,3,1,2})); // [bS, oH, oW, oC] -> [bS, oC, oH, oW]
}
// // mkl supports NCHW format only
// if(!isNCHW) {
// gradI = new NDArray(gradI->permute({0,3,1,2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
// gradO = new NDArray(gradO->permute({0,3,1,2})); // [bS, oH, oW, oC] -> [bS, oC, oH, oW]
// }
deconv2TFdBackPropMKLDNN(weights, gradO, gradI, bS, iC, iH, iW, oC, oH, oW, kH, kW, sH, sW, pH, pW, dH, dW);
deconv2TFdBackPropMKLDNN(weights, gradO, gradI, bS, iC, iH, iW, oC, oH, oW, kH, kW, sH, sW, pH, pW, dH, dW, isNCHW);
delete weights;
// delete weights;
if(!isNCHW) {
delete gradI;
delete gradO;
}
// if(!isNCHW) {
// delete gradI;
// delete gradO;
// }
// ConvolutionUtils::conv2dBP(block, &input, weights, nullptr, gradO, gradI, nullptr, nullptr, kH,kW,sH,sW,pH,pW,dH,dW,isSameMode,isNCHW);

183
libnd4j/include/ops/declarable/platform/mkldnn/deconv3d.cpp
View File

@ -34,17 +34,14 @@ namespace platforms {
//////////////////////////////////////////////////////////////////////////
static void deconv3dMKLDNN(const NDArray* input, const NDArray* weights, const NDArray* bias, NDArray* output,
const int kD, const int kH, const int kW, const int sD, const int sH, const int sW,
const int pD, const int pH, const int pW, const int dD, const int dH, const int dW) {
const int pD, const int pH, const int pW, const int dD, const int dH, const int dW,
const bool isNCDHW) {
// input [bS, iD, iH, iW, iC] ncdhw, mkl doesn't support format ndhwc
// weights [oC, iC, kD, kH, kW] always, mkl doesn't support weights format [kD, kH, kW, oC, iC]
// bias [oC], may be nullptr
// output [bS, oD, oH, oW, oC] ncdhw, mkl doesn't support format ndhwc
// weights [oC, iC, kD, kH, kW] always, mkl doesn't support [kD, kH, kW, oC, iC], so we'll perform permutation
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(true, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWoC, indWiC, indWkD);
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWoC, indWiC, indWkD);
dnnl::memory::dims strides = { sD, sH, sW };
dnnl::memory::dims padding = { pD, pH, pW };
@ -80,8 +77,7 @@ static void deconv3dMKLDNN(const NDArray* input, const NDArray* weights, const N
else
zType = dnnl::memory::data_type::s32;
dnnl::memory::format_tag xFormat = dnnl::memory::format_tag::ncdhw;
dnnl::memory::format_tag xFormat = isNCDHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oidhw;
dnnl::memory::dims xDims = {bS, iC, iD, iH, iW};
@ -93,22 +89,24 @@ static void deconv3dMKLDNN(const NDArray* input, const NDArray* weights, const N
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, xType, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xFormat);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->stridesOf()[0];
x_user_md.data.format_desc.blocking.strides[1] = input->stridesOf()[1];
x_user_md.data.format_desc.blocking.strides[2] = input->stridesOf()[2];
x_user_md.data.format_desc.blocking.strides[3] = input->stridesOf()[3];
x_user_md.data.format_desc.blocking.strides[4] = input->stridesOf()[4];
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
x_user_md.data.format_desc.blocking.strides[4] = input->strideAt(4);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, wType, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->stridesOf()[0];
w_user_md.data.format_desc.blocking.strides[1] = weights->stridesOf()[1];
w_user_md.data.format_desc.blocking.strides[2] = weights->stridesOf()[2];
w_user_md.data.format_desc.blocking.strides[3] = weights->stridesOf()[3];
w_user_md.data.format_desc.blocking.strides[4] = weights->stridesOf()[4];
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(3); // [kD, kH, kW, oC, iC] -> [oC, iC, kD, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(4);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
w_user_md.data.format_desc.blocking.strides[4] = weights->strideAt(2);
// bias
dnnl::memory::desc b_mkl_md;
@ -118,12 +116,14 @@ static void deconv3dMKLDNN(const NDArray* input, const NDArray* weights, const N
// output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(zDims, zType, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(zDims, zType, xFormat);
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->stridesOf()[0];
z_user_md.data.format_desc.blocking.strides[1] = output->stridesOf()[1];
z_user_md.data.format_desc.blocking.strides[2] = output->stridesOf()[2];
z_user_md.data.format_desc.blocking.strides[3] = output->stridesOf()[3];
z_user_md.data.format_desc.blocking.strides[4] = output->stridesOf()[4];
if(output->ews() !=1 || output->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(1);
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(2);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(3);
z_user_md.data.format_desc.blocking.strides[4] = output->strideAt(4);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
@ -184,16 +184,14 @@ static void deconv3dBackPropMKLDNN(const NDArray* input, const NDArray* weights,
const int kD, const int kH, const int kW,
const int sD, const int sH, const int sW,
const int pD, const int pH, const int pW,
const int dD, const int dH, const int dW) {
const int dD, const int dH, const int dW,
const bool isNCDHW) {
// input and gradI [bS, iD, iH, iW, iC], mkl doesn't support ndhwc format
// weights and gradW [oC, iC, kD, kH, kW] always, mkl doesn't support weights format [kD, kH, kW, oC, iC]
// gradB [oC], may be nullptr
// gradO [bS, oD, oH, oW, oC]
// weights and gradW [oC, iC, kD, kH, kW] always, mkl doesn't support [kD, kH, kW, oC, iC], so we'll perform permutation
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(true, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWoC, indWiC, indWkD);
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWoC, indWiC, indWkD);
dnnl::memory::dims strides = { sD, sH, sW };
dnnl::memory::dims padding = { pD, pH, pW };
@ -213,7 +211,7 @@ static void deconv3dBackPropMKLDNN(const NDArray* input, const NDArray* weights,
// gradB type
dnnl::memory::data_type gradBType = gradB != nullptr ? (gradB->dataType() == DataType::FLOAT32 ? dnnl::memory::data_type::f32 : dnnl::memory::data_type::bf16) : dnnl::memory::data_type::f32;
dnnl::memory::format_tag xFormat = dnnl::memory::format_tag::ncdhw; // isNCDHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
dnnl::memory::format_tag xFormat = isNCDHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::oidhw;
dnnl::memory::dims xDims = {bS, iC, iD, iH, iW};
@ -225,52 +223,58 @@ static void deconv3dBackPropMKLDNN(const NDArray* input, const NDArray* weights,
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, xType, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xFormat);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->stridesOf()[0];
x_user_md.data.format_desc.blocking.strides[1] = input->stridesOf()[1];
x_user_md.data.format_desc.blocking.strides[2] = input->stridesOf()[2];
x_user_md.data.format_desc.blocking.strides[3] = input->stridesOf()[3];
x_user_md.data.format_desc.blocking.strides[4] = input->stridesOf()[4];
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
x_user_md.data.format_desc.blocking.strides[4] = input->strideAt(4);
}
// weights
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
dnnl::memory::desc w_user_md = dnnl::memory::desc(wDims, wType, wFormat);
w_user_md.data.format_kind = dnnl_blocked; // overrides format
w_user_md.data.format_desc.blocking.strides[0] = weights->stridesOf()[0];
w_user_md.data.format_desc.blocking.strides[1] = weights->stridesOf()[1];
w_user_md.data.format_desc.blocking.strides[2] = weights->stridesOf()[2];
w_user_md.data.format_desc.blocking.strides[3] = weights->stridesOf()[3];
w_user_md.data.format_desc.blocking.strides[4] = weights->stridesOf()[4];
w_user_md.data.format_desc.blocking.strides[0] = weights->strideAt(3); // [kD, kH, kW, oC, iC] -> [oC, iC, kD, kH, kW]
w_user_md.data.format_desc.blocking.strides[1] = weights->strideAt(4);
w_user_md.data.format_desc.blocking.strides[2] = weights->strideAt(0);
w_user_md.data.format_desc.blocking.strides[3] = weights->strideAt(1);
w_user_md.data.format_desc.blocking.strides[4] = weights->strideAt(2);
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, gradOType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, gradOType, xFormat);
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->stridesOf()[0];
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->stridesOf()[1];
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->stridesOf()[2];
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->stridesOf()[3];
gradO_user_md.data.format_desc.blocking.strides[4] = gradO->stridesOf()[4];
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(2);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(3);
gradO_user_md.data.format_desc.blocking.strides[4] = gradO->strideAt(4);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, gradIType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, gradIType, xFormat);
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->stridesOf()[0];
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->stridesOf()[1];
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->stridesOf()[2];
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->stridesOf()[3];
gradI_user_md.data.format_desc.blocking.strides[4] = gradI->stridesOf()[4];
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(2);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(3);
gradI_user_md.data.format_desc.blocking.strides[4] = gradI->strideAt(4);
}
// gradW
dnnl::memory::desc gradW_mkl_md = dnnl::memory::desc(wDims, gradWType, wFormat);
dnnl::memory::desc gradW_user_md = dnnl::memory::desc(wDims, gradWType, wFormat);
gradW_user_md.data.format_kind = dnnl_blocked; // overrides format
gradW_user_md.data.format_desc.blocking.strides[0] = gradW->stridesOf()[0];
gradW_user_md.data.format_desc.blocking.strides[1] = gradW->stridesOf()[1];
gradW_user_md.data.format_desc.blocking.strides[2] = gradW->stridesOf()[2];
gradW_user_md.data.format_desc.blocking.strides[3] = gradW->stridesOf()[3];
gradW_user_md.data.format_desc.blocking.strides[4] = gradW->stridesOf()[4];
gradW_user_md.data.format_desc.blocking.strides[0] = gradW->strideAt(3); // [kD, kH, kW, oC, iC] -> [oC, iC, kD, kH, kW]
gradW_user_md.data.format_desc.blocking.strides[1] = gradW->strideAt(4);
gradW_user_md.data.format_desc.blocking.strides[2] = gradW->strideAt(0);
gradW_user_md.data.format_desc.blocking.strides[3] = gradW->strideAt(1);
gradW_user_md.data.format_desc.blocking.strides[4] = gradW->strideAt(2);
// gradB
dnnl::memory::desc gradB_mkl_md;
@ -317,11 +321,15 @@ static void deconv3dBackPropMKLDNN(const NDArray* input, const NDArray* weights,
// gradO
auto gradO_user_mem = dnnl::memory(gradO_user_md, engine, gradO->getBuffer());
const bool gradOReorder = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_mem = gradOReorder ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorder)
dnnl::reorder(gradO_user_mem, gradO_mkl_mem).execute(stream, gradO_user_mem, gradO_mkl_mem);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_mem;
const bool gradOReorderW = op_weights_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
const bool gradOReorderD = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_memW = gradOReorderW ? dnnl::memory(op_weights_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
auto gradO_mkl_memD = gradOReorderD ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorderW)
dnnl::reorder(gradO_user_mem, gradO_mkl_memW).execute(stream, gradO_user_mem, gradO_mkl_memW);
if (gradOReorderD)
dnnl::reorder(gradO_user_mem, gradO_mkl_memD).execute(stream, gradO_user_mem, gradO_mkl_memD);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memD;
// gradI
auto gradI_user_mem = dnnl::memory(gradI_user_md, engine, gradI->getBuffer());
@ -344,6 +352,9 @@ static void deconv3dBackPropMKLDNN(const NDArray* input, const NDArray* weights,
// run backward data calculations
dnnl::deconvolution_backward_data(op_data_bp_prim_desc).execute(stream, args);
if(gradOReorderW || gradOReorderD)
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memW;
// run backward weights calculations
dnnl::deconvolution_backward_weights(op_weights_bp_prim_desc).execute(stream, args);
@ -400,23 +411,7 @@ PLATFORM_IMPL(deconv3d, ENGINE_CPU) {
ConvolutionUtils::calcPadding3D(pD, pH, pW, iD, iH, iW, oD, oH, oW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
}
// mkl supports only [oC, iC, kD, kH, kW] format for weights
weights = new NDArray(weights->permute({3,4,0,1,2})); // [kD, kH, kW, oC, iC] -> [oC, iC, kD, kH, kW]
// mkl supports only NCDHW
if(!isNCDHW) {
input = new NDArray(input->permute({0,4,1,2,3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
output = new NDArray(output->permute({0,4,1,2,3})); // [bS, oD, oH, oW, oC] -> [bS, oC, oD, oH, oW]
}
deconv3dMKLDNN(input, weights, bias, output, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW);
delete weights;
if(!isNCDHW) {
delete input;
delete output;
}
deconv3dMKLDNN(input, weights, bias, output, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, isNCDHW);
return Status::OK();
}
@ -495,27 +490,7 @@ PLATFORM_IMPL(deconv3d_bp, ENGINE_CPU) {
if(isSameMode) // Note: we're intentionally swapping iH and oH, to calculated the padding for a"normal" conv (not deconv) forward pass
ConvolutionUtils::calcPadding3D(pD, pH, pW, iD, iH, iW, oD, oH, oW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
// mkl supports only [oC, iC, kD, kH, kW] for weights
weights = new NDArray(weights->permute({3,4,0,1,2})); // [kD, kH, kW, oC, iC] -> [oC, iC, kD, kH, kW]
gradW = new NDArray(gradW->permute({3,4,0,1,2})); // [kD, kH, kW, oC, iC] -> [oC, iC, kD, kH, kW]
// mkl supports NCDHW format only
if(!isNCDHW) {
input = new NDArray(input->permute({0,4,1,2,3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
gradI = new NDArray(gradI->permute({0,4,1,2,3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
gradO = new NDArray(gradO->permute({0,4,1,2,3})); // [bS, oD, oH, oW, oC] -> [bS, oC, oD, oH, oW]
}
deconv3dBackPropMKLDNN(input, weights, gradO, gradI, gradW, gradB, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW);
delete weights;
delete gradW;
if(!isNCDHW) {
delete input;
delete gradI;
delete gradO;
}
deconv3dBackPropMKLDNN(input, weights, gradO, gradI, gradW, gradB, kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, isNCDHW);
return Status::OK();
}

89
libnd4j/include/ops/declarable/platform/mkldnn/depthwiseConv2d.cpp
View File

@ -86,7 +86,7 @@ static void depthwiseConv2dMKLDNN(const NDArray* input, const NDArray* weights,
else
zType = dnnl::memory::data_type::s32;
dnnl::memory::format_tag xzFrmat = dnnl::memory::format_tag::nchw;
dnnl::memory::format_tag xzFrmat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::goihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
@ -98,11 +98,13 @@ static void depthwiseConv2dMKLDNN(const NDArray* input, const NDArray* weights,
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, xType, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xzFrmat);
x_user_md.data.format_kind = dnnl_blocked; // overrides format NHWC -> NCHW
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(isNCHW ? 1 : 3);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(isNCHW ? 2 : 1);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(isNCHW ? 3 : 2);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1); // do permutation NHWC -> NCHW
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
}
// weights, make permute [kH, kW, iC, mC] -> [iC, mC, 1, kH, kW];
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
@ -122,11 +124,13 @@ static void depthwiseConv2dMKLDNN(const NDArray* input, const NDArray* weights,
// output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(zDims, zType, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(zDims, zType, xzFrmat);
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(isNCHW ? 1 : 3);
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(isNCHW ? 2 : 1);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(isNCHW ? 3 : 2);
if(output->ews() != 1 || output->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(1); // do permutation NHWC -> NCHW
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(2);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(3);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
@ -219,7 +223,7 @@ static void depthwiseConv2dNackPropMKLDNN(const NDArray* input, const NDArray* w
// gradB type
dnnl::memory::data_type gradBType = gradB != nullptr ? (gradB->dataType() == DataType::FLOAT32 ? dnnl::memory::data_type::f32 : dnnl::memory::data_type::bf16) : dnnl::memory::data_type::f32;
dnnl::memory::format_tag xFormat = dnnl::memory::format_tag::nchw; // isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag xzFrmat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
dnnl::memory::format_tag wFormat = dnnl::memory::format_tag::goihw;
dnnl::memory::dims xDims = {bS, iC, iH, iW};
@ -230,12 +234,14 @@ static void depthwiseConv2dNackPropMKLDNN(const NDArray* input, const NDArray* w
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, xType, dnnl::memory::format_tag::any);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xFormat);
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(isNCHW ? 1 : 3);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(isNCHW ? 2 : 1);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(isNCHW ? 3 : 2);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, xType, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(2);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(3);
}
// weights, make permute [kH, kW, iC, mC] -> [iC, mC, 1, kH, kW];
dnnl::memory::desc w_mkl_md = dnnl::memory::desc(wDims, wType, dnnl::memory::format_tag::any);
@ -249,21 +255,25 @@ static void depthwiseConv2dNackPropMKLDNN(const NDArray* input, const NDArray* w
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, gradOType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, gradOType, xFormat);
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(isNCHW ? 1 : 3);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(isNCHW ? 2 : 1);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(isNCHW ? 3 : 2);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, gradOType, xzFrmat);
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(2);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(3);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, gradIType, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, gradIType, xFormat);
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(isNCHW ? 1 : 3);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(isNCHW ? 2 : 1);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(isNCHW ? 3 : 2);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, gradIType, xzFrmat);
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(2);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(3);
}
// gradW, make permute [kH, kW, iC, mC] -> [iC, mC, 1, kH, kW];
dnnl::memory::desc gradW_mkl_md = dnnl::memory::desc(wDims, gradWType, dnnl::memory::format_tag::any);
@ -319,11 +329,15 @@ static void depthwiseConv2dNackPropMKLDNN(const NDArray* input, const NDArray* w
// gradO
auto gradO_user_mem = dnnl::memory(gradO_user_md, engine, gradO->getBuffer());
const bool gradOReorder = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_mem = gradOReorder ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorder)
dnnl::reorder(gradO_user_mem, gradO_mkl_mem).execute(stream, gradO_user_mem, gradO_mkl_mem);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_mem;
const bool gradOReorderW = op_weights_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
const bool gradOReorderD = op_data_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_memW = gradOReorderW ? dnnl::memory(op_weights_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
auto gradO_mkl_memD = gradOReorderD ? dnnl::memory(op_data_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorderW)
dnnl::reorder(gradO_user_mem, gradO_mkl_memW).execute(stream, gradO_user_mem, gradO_mkl_memW);
if (gradOReorderD)
dnnl::reorder(gradO_user_mem, gradO_mkl_memD).execute(stream, gradO_user_mem, gradO_mkl_memD);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memD;
// gradI
auto gradI_user_mem = dnnl::memory(gradI_user_md, engine, gradI->getBuffer());
@ -346,6 +360,9 @@ static void depthwiseConv2dNackPropMKLDNN(const NDArray* input, const NDArray* w
// run backward data calculations
dnnl::convolution_backward_data(op_data_bp_prim_desc).execute(stream, args);
if(gradOReorderW || gradOReorderD)
args[DNNL_ARG_DIFF_DST] = gradO_mkl_memW;
// run backward weights calculations
dnnl::convolution_backward_weights(op_weights_bp_prim_desc).execute(stream, args);
@ -401,6 +418,7 @@ PLATFORM_IMPL(depthwise_conv2d, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(depthwise_conv2d, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
auto weights = INPUT_VARIABLE(1);
auto bias = block.width() > 2 ? INPUT_VARIABLE(2) : nullptr;
@ -473,6 +491,7 @@ PLATFORM_IMPL(depthwise_conv2d_bp, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////
PLATFORM_CHECK(depthwise_conv2d_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iH, iW, iC] (NDHWC) or [bS, iC, iH, iW] (NCDHW)
auto weights = INPUT_VARIABLE(1); // [kH, kW, iC, mC] always
auto bias = block.width() > 3 ? INPUT_VARIABLE(2) : nullptr; // [oC] = [iC*mC]

223
libnd4j/include/ops/declarable/platform/mkldnn/maxpooling2d.cpp
View File

@ -17,6 +17,7 @@
//
// @author saudet
// @author raver119@gmail.com
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <ops/declarable/PlatformHelper.h>
@ -33,105 +34,38 @@ namespace nd4j {
namespace ops {
namespace platforms {
//////////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(maxpool2d, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0);
REQUIRE_TRUE(input->rankOf() == 4, 0, "Input should have rank of 4, but got %i instead",
input->rankOf());
// 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
auto argI = *(block.getIArguments());
auto output = OUTPUT_VARIABLE(0);
const auto kH = INT_ARG(0);
const auto kW = INT_ARG(1);
const auto sH = INT_ARG(2);
const auto sW = INT_ARG(3);
int pH = INT_ARG(4);
int pW = INT_ARG(5);
const auto dH = INT_ARG(6);
const auto dW = INT_ARG(7);
const auto isSameMode = static_cast<bool>(INT_ARG(8));
REQUIRE_TRUE(input->rankOf() == 4, 0, "MAXPOOL2D MKLDNN OP: input array should have rank of 4, but got %i instead", input->rankOf());
REQUIRE_TRUE(dH != 0 && dW != 0, 0, "AVGPOOL2D op: dilation must not be zero, but got instead {%i, %i}",
dH, dW);
// 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
const int kH = INT_ARG(0);
const int kW = INT_ARG(1);
const int sH = INT_ARG(2);
const int sW = INT_ARG(3);
int pH = INT_ARG(4);
int pW = INT_ARG(5);
const int dH = INT_ARG(6);
const int dW = INT_ARG(7);
const int paddingMode = INT_ARG(8);
// const int extraParam0 = INT_ARG(9);
const int isNCHW = block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 1-NHWC, 0-NCHW
int oH = 0;
int oW = 0;
REQUIRE_TRUE(dH != 0 && dW != 0, 0, "MAXPOOL2D MKLDNN op: dilation must not be zero, but got instead {%i, %i}", dH, dW);
int isNCHW = block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
const int iH = static_cast<int>(isNCHW ? input->sizeAt(2) : input->sizeAt(1));
const int iW = static_cast<int>(isNCHW ? input->sizeAt(3) : input->sizeAt(2));
if (!isNCHW) {
input = new NDArray(
input->permute({0, 3, 1, 2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
output = new NDArray(
output->permute({0, 3, 1, 2})); // [bS, oH, oW, iC] -> [bS, iC, oH, oW]
}
ConvolutionUtils::calcOutSizePool2D(oH, oW, kH, kW, sH, sW, pH, pW, dH, dW, iH, iW, isSameMode);
if (isSameMode)
if (paddingMode)
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW);
const int bS = input->sizeAt(0);
const int iC = input->sizeAt(1);
const int oC = output->sizeAt(1);
auto poolingMode = PoolingType::MAX_POOL;
int extraParam0 = 1;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool2d(kH, kW, sH, sW, pH, pW, dH, dW, poolingMode, extraParam0,
true,
bS, iC, iH, iW, oC, oH, oW, input, nullptr, output,
algorithm,
&pool_src_md, nullptr, &pool_dst_md, &user_src_md, nullptr,
&user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto pool_desc = pooling_forward::desc(prop_kind::forward_inference, algorithm, pool_src_md,
pool_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto user_src_memory = dnnl::memory(user_src_md, engine, input->buffer());
auto user_dst_memory = dnnl::memory(user_dst_md, engine, output->buffer());
auto pool_src_memory = user_src_memory;
dnnl::stream stream(engine);
if (pool_prim_desc.src_desc() != user_src_memory.get_desc()) {
pool_src_memory = dnnl::memory(pool_prim_desc.src_desc(), engine);
reorder(user_src_memory, pool_src_memory).execute(stream, user_src_memory, pool_src_memory);
}
auto pool_dst_memory = user_dst_memory;
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
pool_dst_memory = dnnl::memory(pool_prim_desc.dst_desc(), engine);
}
pooling_forward(pool_prim_desc).execute(stream, {{DNNL_ARG_SRC, pool_src_memory},
{DNNL_ARG_DST, pool_dst_memory}});
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
reorder(pool_dst_memory, user_dst_memory).execute(stream, pool_dst_memory, user_dst_memory);
}
stream.wait();
if (!isNCHW) {
delete input;
delete output;
}
mkldnnUtils::poolingMKLDNN(input, output, 0,kH,kW, 0,sH,sW, 0,pH,pW, isNCHW, algorithm::pooling_max);
return Status::OK();
}
@ -159,117 +93,24 @@ PLATFORM_IMPL(maxpool2d_bp, ENGINE_CPU) {
int pW = INT_ARG(5); // paddings width
int dH = INT_ARG(6); // dilations height
int dW = INT_ARG(7); // dilations width
int isSameMode = INT_ARG(8); // 0-VALID, 1-SAME
int extraParam0 = INT_ARG(9);
int isNCHW =
block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
int paddingMode = INT_ARG(8); // 0-VALID, 1-SAME
// int extraParam0 = INT_ARG(9);
int isNCHW = block.getIArguments()->size() > 10 ? !INT_ARG(10) : 1; // INT_ARG(10): 0-NCHW, 1-NHWC
REQUIRE_TRUE(input->rankOf() == 4, 0,
"AVGPOOL2D_BP op: input should have rank of 4, but got %i instead", input->rankOf());
REQUIRE_TRUE(dH != 0 && dW != 0, 0,
"AVGPOOL2D_BP op: dilation must not be zero, but got instead {%i, %i}", dH, dW);
REQUIRE_TRUE(input->rankOf() == 4, 0, "MAXPOOL2D_BP MKLDNN op: input should have rank of 4, but got %i instead", input->rankOf());
REQUIRE_TRUE(dH != 0 && dW != 0, 0, "MAXPOOL2D_BP MKLDNN op: dilation must not be zero, but got instead {%i, %i}", dH, dW);
int bS, iC, iH, iW, oC, oH, oW; // batch size, input channels, input height/width, output channels, output height/width;
int indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC,
indIiH, indWiC, indWoC, indWkH, indOoH);
ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
std::string expectedGradOShape = ShapeUtils::shapeAsString(
ShapeUtils::composeShapeUsingDimsAndIdx({bS, iC, oH, oW, 0, indIOioC, indIiH, indIiH + 1}));
std::string expectedGradIShape = ShapeUtils::shapeAsString(
ShapeUtils::composeShapeUsingDimsAndIdx({bS, iC, iH, iW, 0, indIOioC, indIiH, indIiH + 1}));
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0,
"AVGPOOL2D_BP op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !",
expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedGradIShape == ShapeUtils::shapeAsString(gradI), 0,
"AVGPOOL2D_BP op: wrong shape of input's gradients array (epsilon), expected is %s, but got %s instead !",
expectedGradIShape.c_str(), ShapeUtils::shapeAsString(gradI).c_str());
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx({bS, iC, oH, oW, 0, indIOioC, indIiH, indIiH + 1});
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "MAXPOOL2D_BP MKLDNN op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
if (!isNCHW) {
input = new NDArray(input->permute(
{0, 3, 1, 2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradI = new NDArray(gradI->permute(
{0, 3, 1, 2})); // [bS, iH, iW, iC] -> [bS, iC, iH, iW]
gradO = new NDArray(gradO->permute(
{0, 3, 1, 2})); // [bS, oH, oW, iC] -> [bS, iC, oH, oW]
}
if (isSameMode) // SAME
if (paddingMode) // SAME
ConvolutionUtils::calcPadding2D(pH, pW, oH, oW, iH, iW, kH, kW, sH, sW, dH, dW);
auto poolingMode = PoolingType::MAX_POOL;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_diff_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_diff_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool2d(kH, kW, sH, sW, pH, pW, dH, dW, poolingMode, extraParam0,
true,
bS, iC, iH, iW, oC, oH, oW, input, gradI, gradO, algorithm,
&pool_src_md, &pool_diff_src_md, &pool_dst_md, &user_src_md,
&user_diff_src_md, &user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
// input is sometimes null, so we can't rely on pool_src_md being valid
auto pool_desc = pooling_forward::desc(prop_kind::forward, algorithm,
input->buffer() != nullptr ? pool_src_md : pool_diff_src_md,
pool_dst_md, pool_strides, pool_kernel, pool_padding,
pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto poolB_desc = pooling_backward::desc(algorithm, pool_diff_src_md, pool_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto poolB_prim_desc = pooling_backward::primitive_desc(poolB_desc, engine, pool_prim_desc);
auto userB_src_memory = dnnl::memory(user_src_md, engine, gradI->buffer());
auto userB_dst_memory = dnnl::memory(user_dst_md, engine, gradO->buffer());
auto poolB_src_memory = userB_src_memory;
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
poolB_src_memory = dnnl::memory(poolB_prim_desc.diff_src_desc(), engine);
}
auto poolB_dst_memory = userB_dst_memory;
if (poolB_prim_desc.diff_dst_desc() != userB_dst_memory.get_desc()) {
poolB_dst_memory = dnnl::memory(poolB_prim_desc.diff_dst_desc(), engine);
reorder(userB_dst_memory, poolB_dst_memory).execute(stream, userB_dst_memory, poolB_dst_memory);
}
auto user_src_memory = dnnl::memory(user_src_md, engine, input->buffer());
auto pool_src_memory = user_src_memory;
if (pool_prim_desc.src_desc() != user_src_memory.get_desc()) {
pool_src_memory = dnnl::memory(pool_prim_desc.src_desc(), engine);
reorder(user_src_memory, pool_src_memory).execute(stream, user_src_memory, pool_src_memory);
}
auto pool_dst_memory = dnnl::memory(pool_prim_desc.dst_desc(), engine);
auto pool_workspace_memory = dnnl::memory(pool_prim_desc.workspace_desc(), engine);
pooling_forward(pool_prim_desc).execute(stream, {{DNNL_ARG_SRC, pool_src_memory},
{DNNL_ARG_DST, pool_dst_memory},
{DNNL_ARG_WORKSPACE, pool_workspace_memory}});
// probably wrong, fix that
pooling_backward(poolB_prim_desc).execute(stream, {{DNNL_ARG_DIFF_DST, poolB_dst_memory},
{DNNL_ARG_WORKSPACE, pool_workspace_memory},
{DNNL_ARG_DIFF_SRC, poolB_src_memory}});
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
reorder(poolB_src_memory, userB_src_memory).execute(stream, poolB_src_memory, userB_src_memory);
}
stream.wait();
if (!isNCHW) {
delete input;
delete gradI;
delete gradO;
}
mkldnnUtils::poolingBpMKLDNN(input, gradO, gradI, 0,kH,kW, 0,sH,sW, 0,pH,pW, isNCHW, algorithm::pooling_max);
return Status::OK();
}

223
libnd4j/include/ops/declarable/platform/mkldnn/maxpooling3d.cpp
View File

@ -16,6 +16,7 @@
//
// @author raver119@gmail.com
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <ops/declarable/PlatformHelper.h>
@ -34,10 +35,9 @@ namespace platforms {
//////////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(maxpool3dnew, ENGINE_CPU) {
auto input = INPUT_VARIABLE(
0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto output = OUTPUT_VARIABLE(
0); // [bS, oD, oH, oW, iC] (NDHWC) or [bS, iC, oD, oH, oW] (NCDHW)
auto input = INPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto output = OUTPUT_VARIABLE(0); // [bS, oD, oH, oW, iC] (NDHWC) or [bS, iC, oD, oH, oW] (NCDHW)
int kD = INT_ARG(0); // filter(kernel) depth
int kH = INT_ARG(1); // filter(kernel) height
@ -51,95 +51,24 @@ PLATFORM_IMPL(maxpool3dnew, ENGINE_CPU) {
int dD = INT_ARG(9); // dilations depth
int dH = INT_ARG(10); // dilations height
int dW = INT_ARG(11); // dilations width
int isSameMode = INT_ARG(12); // 1-SAME, 0-VALID
int paddingMode = INT_ARG(12); // 1-SAME, 0-VALID
// int extraParam0 = INT_ARG(13); // unnecessary for max case, required only for avg and pnorm cases
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
REQUIRE_TRUE(input->rankOf() == 5, 0,
"MAXPOOL3DNEW OP: rank of input array must be equal to 5, but got %i instead !",
input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0,
"MAXPOOL3DNEW op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
REQUIRE_TRUE(input->rankOf() == 5, 0, "MAXPOOL3DNEW MKLDNN OP: rank of input array must be equal to 5, but got %i instead !", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0, "MAXPOOL3DNEW MKLDNN op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW,
indIOioC, indIOioD, indWiC, indWoC, indWkD);
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
std::string expectedOutputShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx(
{bS, iC, oD, oH, oW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
REQUIRE_TRUE(expectedOutputShape == ShapeUtils::shapeAsString(output), 0,
"MAXPOOL3D op: wrong shape of output array, expected is %s, but got %s instead !",
expectedOutputShape.c_str(), ShapeUtils::shapeAsString(output).c_str());
// REQUIRE_TRUE(iD >= kD && iH >= kH && iW >= kW, 0, "MAXPOOL3D OP: the input depth/height/width must be greater or equal to kernel(filter) depth/height/width, but got [%i, %i, %i] and [%i, %i, %i] correspondingly !", iD,iH,iW, kD,kH,kW);
// REQUIRE_TRUE(kD/2 >= pD && kH/2 >= pH && kW/2 >= pW, 0, "MAXPOOL3D OP: pad depth/height/width must not be greater than half of kernel depth/height/width, but got [%i, %i, %i] and [%i, %i, %i] correspondingly !", pD,pH,pW, kD,kH,kW);
if(paddingMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
if (!isNCDHW) {
input = new NDArray(
input->permute({0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
output = new NDArray(
output->permute({0, 4, 1, 2, 3})); // [bS, oD, oH, oW, iC] -> [bS, iC, oD, oH, oW]
}
if (isSameMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH,
dW);
auto poolingMode = PoolingType::MAX_POOL;
auto extraParam0 = 1;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool3d(kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, poolingMode,
extraParam0, true,
bS, iC, iD, iH, iW, oC, oD, oH, oW, input, nullptr, output,
algorithm,
&pool_src_md, nullptr, &pool_dst_md, &user_src_md, nullptr,
&user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto pool_desc = pooling_forward::desc(prop_kind::forward_inference, algorithm, pool_src_md,
pool_dst_md, pool_strides, pool_kernel, pool_padding,
pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto user_src_memory = dnnl::memory(user_src_md, engine, input->buffer());
auto user_dst_memory = dnnl::memory(user_dst_md, engine, output->buffer());
auto pool_src_memory = user_src_memory;
if (pool_prim_desc.src_desc() != user_src_memory.get_desc()) {
pool_src_memory = dnnl::memory(pool_prim_desc.src_desc(), engine);
reorder(user_src_memory, pool_src_memory).execute(stream, user_src_memory, pool_src_memory);
}
auto pool_dst_memory = user_dst_memory;
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
pool_dst_memory = dnnl::memory(pool_prim_desc.dst_desc(), engine);
}
pooling_forward(pool_prim_desc).execute(stream, {{DNNL_ARG_SRC, pool_src_memory},
{DNNL_ARG_DST, pool_dst_memory}});
if (pool_prim_desc.dst_desc() != user_dst_memory.get_desc()) {
reorder(pool_dst_memory, user_dst_memory).execute(stream, pool_dst_memory, user_dst_memory);
}
stream.wait();
if (!isNCDHW) {
delete input;
delete output;
}
mkldnnUtils::poolingMKLDNN(input, output, kD,kH,kW, sD,sH,sW, pD,pH,pW, isNCDHW, algorithm::pooling_max);
return Status::OK();
}
//////////////////////////////////////////////////////////////////////////
@ -152,6 +81,7 @@ PLATFORM_CHECK(maxpool3dnew, ENGINE_CPU) {
//////////////////////////////////////////////////////////////////////////
PLATFORM_IMPL(maxpool3dnew_bp, ENGINE_CPU) {
auto input = INPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW)
auto gradO = INPUT_VARIABLE(1); // [bS, oD, oH, oW, oC] (NDHWC) or [bS, oC, oD, oH, oW] (NCDHW), epsilon_next
auto gradI = OUTPUT_VARIABLE(0); // [bS, iD, iH, iW, iC] (NDHWC) or [bS, iC, iD, iH, iW] (NCDHW), epsilon
@ -162,127 +92,30 @@ PLATFORM_IMPL(maxpool3dnew_bp, ENGINE_CPU) {
const int sD = INT_ARG(3); // strides depth
const int sH = INT_ARG(4); // strides height
const int sW = INT_ARG(5); // strides width
int pD = INT_ARG(6); // paddings depth
int pH = INT_ARG(7); // paddings height
int pW = INT_ARG(8); // paddings width
int pD = INT_ARG(6); // paddings depth
int pH = INT_ARG(7); // paddings height
int pW = INT_ARG(8); // paddings width
const int dD = INT_ARG(9); // dilations depth
const int dH = INT_ARG(10); // dilations height
const int dW = INT_ARG(11); // dilations width
const int isSameMode = INT_ARG(12); // 1-SAME, 0-VALID
const int paddngMode = INT_ARG(12); // 1-SAME, 0-VALID
// int extraParam0 = INT_ARG(13); // unnecessary for max case, required only for avg and pnorm cases
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
int isNCDHW = block.getIArguments()->size() > 14 ? !INT_ARG(14) : 1; // 1-NDHWC, 0-NCDHW
REQUIRE_TRUE(input->rankOf() == 5, 0,
"MAXPOOL3D_BP op: input should have rank of 5, but got %i instead", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0,
"MAXPOOL3DNEW op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
REQUIRE_TRUE(input->rankOf() == 5, 0, "MAXPOOL3DNEW_BP MKLDNN op: input should have rank of 5, but got %i instead", input->rankOf());
REQUIRE_TRUE(dD != 0 && dH != 0 && dW != 0, 0, "MAXPOOL3DNEW_BP MKLDNN op: dilation must not be zero, but got instead {%i, %i, %i}", dD, dH, dW);
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int bS, iC, iD, iH, iW, oC, oD, oH, oW; // batch size, input channels, input depth/height/width, output channels, output depth/height/width;
int indIOioC, indIOioD, indWoC, indWiC, indWkD; // corresponding indexes
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW,
indIOioC, indIOioD, indWiC, indWoC, indWkD);
ConvolutionUtils::getSizesAndIndexesConv3d(isNCDHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIOioD, indWiC, indWoC, indWkD);
std::string expectedGradOShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx(
{bS, iC, oD, oH, oW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
std::string expectedGradIShape = ShapeUtils::shapeAsString(ShapeUtils::composeShapeUsingDimsAndIdx(
{bS, iC, iD, iH, iW, 0, indIOioC, indIOioD, indIOioD + 1, indIOioD + 2}));
REQUIRE_TRUE(expectedGradOShape == ShapeUtils::shapeAsString(gradO), 0,
"MAXPOOL3D_BP op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !",
expectedGradOShape.c_str(), ShapeUtils::shapeAsString(gradO).c_str());
REQUIRE_TRUE(expectedGradIShape == ShapeUtils::shapeAsString(gradI), 0,
"MAXPOOL3D_BP op: wrong shape of input's gradients array (epsilon), expected is %s, but got %s instead !",
expectedGradIShape.c_str(), ShapeUtils::shapeAsString(gradI).c_str());
std::vector<Nd4jLong> expectedGradOShape = ShapeUtils::composeShapeUsingDimsAndIdx({bS,iC,oD,oH,oW, 0,indIOioC,indIOioD,indIOioD+1,indIOioD+2});
REQUIRE_TRUE(gradO->isSameShape(expectedGradOShape), 0, "MAXPOOL3DNEW_BP MKLDNN op: wrong shape of output's gradients array (next epsilon), expected is %s, but got %s instead !", ShapeUtils::shapeAsString(expectedGradOShape).c_str(), ShapeUtils::shapeAsString(gradO).c_str());
if (!isNCDHW) {
input = new NDArray(input->permute(
{0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
gradI = new NDArray(gradI->permute(
{0, 4, 1, 2, 3})); // [bS, iD, iH, iW, iC] -> [bS, iC, iD, iH, iW]
gradO = new NDArray(gradO->permute(
{0, 4, 1, 2, 3})); // [bS, oD, oH, oW, iC] -> [bS, iC, oD, oH, oW]
}
if(paddngMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH, dW);
if (isSameMode) // SAME
ConvolutionUtils::calcPadding3D(pD, pH, pW, oD, oH, oW, iD, iH, iW, kD, kH, kW, sD, sH, sW, dD, dH,
dW);
auto poolingMode = PoolingType::MAX_POOL;
auto extraParam0 = 1;
dnnl_memory_desc_t empty;
dnnl::memory::desc pool_src_md(empty), pool_diff_src_md(empty), pool_dst_md(empty);
dnnl::memory::desc user_src_md(empty), user_diff_src_md(empty), user_dst_md(empty);
dnnl::memory::dims pool_strides, pool_kernel, pool_padding, pool_padding_r;
dnnl::algorithm algorithm;
mkldnnUtils::getMKLDNNMemoryDescPool3d(kD, kH, kW, sD, sH, sW, pD, pH, pW, dD, dH, dW, poolingMode,
extraParam0, true,
bS, iC, iD, iH, iW, oC, oD, oH, oW, input, gradI, gradO,
algorithm,
&pool_src_md, &pool_diff_src_md, &pool_dst_md, &user_src_md,
&user_diff_src_md, &user_dst_md,
pool_strides, pool_kernel, pool_padding, pool_padding_r);
// input is sometimes null, so we can't rely on pool_src_md being valid
if (input->buffer() == nullptr) {
pool_src_md = pool_diff_src_md;
user_src_md = user_diff_src_md;
}
auto pool_desc = pooling_forward::desc(prop_kind::forward, algorithm, pool_src_md, pool_dst_md, pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
auto pool_prim_desc = pooling_forward::primitive_desc(pool_desc, engine);
auto poolB_desc = pooling_backward::desc(algorithm, pool_diff_src_md, pool_dst_md, pool_strides, pool_kernel, pool_padding, pool_padding_r);
auto poolB_prim_desc = pooling_backward::primitive_desc(poolB_desc, engine, pool_prim_desc);
auto userB_src_memory = dnnl::memory(user_diff_src_md, engine, gradI->buffer());
auto userB_dst_memory = dnnl::memory(user_dst_md, engine, gradO->buffer());
auto poolB_src_memory = userB_src_memory;
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
poolB_src_memory = dnnl::memory(poolB_prim_desc.diff_src_desc(), engine);
}
auto poolB_dst_memory = userB_dst_memory;
if (poolB_prim_desc.diff_dst_desc() != userB_dst_memory.get_desc()) {
poolB_dst_memory = dnnl::memory(poolB_prim_desc.diff_dst_desc(), engine);
reorder(userB_dst_memory, poolB_dst_memory).execute(stream, userB_dst_memory, poolB_dst_memory);
}
auto user_src_memory = dnnl::memory(user_src_md, engine, input->buffer());
auto pool_src_memory = user_src_memory;
if (pool_prim_desc.src_desc() != user_src_memory.get_desc()) {
pool_src_memory = dnnl::memory(pool_prim_desc.src_desc(), engine);
reorder(user_src_memory, pool_src_memory).execute(stream, user_src_memory, pool_src_memory);
}
auto pool_dst_memory = dnnl::memory(pool_prim_desc.dst_desc(), engine);
auto pool_workspace_memory = dnnl::memory(pool_prim_desc.workspace_desc(), engine);
pooling_forward(pool_prim_desc).execute(stream, {{DNNL_ARG_SRC, pool_src_memory},
{DNNL_ARG_DST, pool_dst_memory},
{DNNL_ARG_WORKSPACE, pool_workspace_memory}});
pooling_backward(poolB_prim_desc).execute(stream, {{DNNL_ARG_DIFF_DST, poolB_dst_memory},
{DNNL_ARG_WORKSPACE, pool_workspace_memory},
{DNNL_ARG_DIFF_SRC, poolB_src_memory}});
if (poolB_prim_desc.diff_src_desc() != userB_src_memory.get_desc()) {
reorder(poolB_src_memory, userB_src_memory).execute(stream, poolB_src_memory, userB_src_memory);
}
stream.wait();
if (!isNCDHW) {
delete input;
delete gradI;
delete gradO;
}
mkldnnUtils::poolingBpMKLDNN(input, gradO, gradI, kD,kH,kW, sD,sH,sW, pD,pH,pW, isNCDHW, algorithm::pooling_max);
return Status::OK();
}

411
libnd4j/include/ops/declarable/platform/mkldnn/mkldnnUtils.cpp
View File

@ -16,9 +16,11 @@
//
// @author saudet
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#include <dnnl_types.h>
#include <ops/declarable/helpers/convolutions.h>
#include "mkldnnUtils.h"
using namespace dnnl;
@ -26,6 +28,314 @@ using namespace dnnl;
namespace nd4j {
namespace mkldnnUtils {
//////////////////////////////////////////////////////////////////////
void poolingMKLDNN(const NDArray *input, NDArray *output,
const int kD, const int kH, const int kW,
const int sD, const int sH, const int sW,
const int pD, const int pH, const int pW,
const int isNCHW, const dnnl::algorithm mode) {
// unfortunately mkl dnn doesn't support any format (dnnl::memory::format_tag::any) for input
const int rank = input->rankOf();
int bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH;
dnnl::memory::dims strides, kernel, padding, padding_r, xDims, zDims;
dnnl::memory::format_tag xzFrmat;
const auto type = dnnl::memory::data_type::f32;
if(rank == 4) { // 2d
ops::ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *output, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
strides = { sH, sW };
kernel = { kH, kW };
padding = { pH, pW };
padding_r = { (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pW };
xDims = {bS, iC, iH, iW};
zDims = {bS, oC, oH, oW};
xzFrmat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
}
else { // 3d
ops::ConvolutionUtils::getSizesAndIndexesConv3d(isNCHW, *input, *output, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH);
strides = { sD, sH, sW };
kernel = { kD, kH, kW };
padding = { pD, pH, pW };
padding_r = { (oD - 1) * sD - iD + kD - pD, (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pW };
xDims = {bS, iC, iD, iH, iW};
zDims = {bS, oC, oD, oH, oW};
xzFrmat = isNCHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
}
// memory descriptors for arrays
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, type, xzFrmat);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(isNCHW ? 1 :-1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(isNCHW ? 2 : 1);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(isNCHW ? 3 : 2);
if(rank == 5)
x_user_md.data.format_desc.blocking.strides[4] = input->strideAt(isNCHW ? 4 : 3);
}
// output
dnnl::memory::desc z_mkl_md = dnnl::memory::desc(zDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc z_user_md = dnnl::memory::desc(zDims, type, xzFrmat);
if(output->ews() != 1 || output->ordering() != 'c') {
z_user_md.data.format_kind = dnnl_blocked; // overrides format
z_user_md.data.format_desc.blocking.strides[0] = output->strideAt(0);
z_user_md.data.format_desc.blocking.strides[1] = output->strideAt(isNCHW ? 1 :-1);
z_user_md.data.format_desc.blocking.strides[2] = output->strideAt(isNCHW ? 2 : 1);
z_user_md.data.format_desc.blocking.strides[3] = output->strideAt(isNCHW ? 3 : 2);
if(rank == 5)
z_user_md.data.format_desc.blocking.strides[4] = output->strideAt(isNCHW ? 4 : 3);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
// operation primitive description
dnnl::pooling_forward::desc op_desc(dnnl::prop_kind::forward_inference, mode, x_mkl_md, z_mkl_md, strides, kernel, padding, padding_r);
dnnl::pooling_forward::primitive_desc op_prim_desc(op_desc, engine);
// arguments (memory buffers) necessary for calculations
std::unordered_map<int, dnnl::memory> args;
dnnl::stream stream(engine);
// provide memory buffers and check whether reorder is required
// input
auto x_user_mem = dnnl::memory(x_user_md, engine, input->getBuffer());
const bool xReorder = op_prim_desc.src_desc() != x_user_mem.get_desc();
auto x_mkl_mem = xReorder ? dnnl::memory(op_prim_desc.src_desc(), engine) : x_user_mem;
if (xReorder)
dnnl::reorder(x_user_mem, x_mkl_mem).execute(stream, x_user_mem, x_mkl_mem);
args[DNNL_ARG_SRC] = x_mkl_mem;
// output
auto z_user_mem = dnnl::memory(z_user_md, engine, output->getBuffer());
const bool zReorder = op_prim_desc.dst_desc() != z_user_mem.get_desc();
auto z_mkl_mem = zReorder ? dnnl::memory(op_prim_desc.dst_desc(), engine) : z_user_mem;
args[DNNL_ARG_DST] = z_mkl_mem;
// run calculations
dnnl::pooling_forward(op_prim_desc).execute(stream, args);
// reorder outputs if necessary
if (zReorder)
dnnl::reorder(z_mkl_mem, z_user_mem).execute(stream, z_mkl_mem, z_user_mem);
stream.wait();
}
//////////////////////////////////////////////////////////////////////
void poolingBpMKLDNN(const NDArray *input, const NDArray *gradO, NDArray *gradI,
const int kD, const int kH, const int kW,
const int sD, const int sH, const int sW,
const int pD, const int pH, const int pW,
const int isNCHW, const dnnl::algorithm mode) {
// unfortunately mkl dnn doesn't support any format (dnnl::memory::format_tag::any) for input
const int rank = input->rankOf();
int bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIiH, indWoC, indWiC, indWkH, indOoH;
dnnl::memory::dims strides, kernel, padding, padding_r, xDims, zDims;
dnnl::memory::format_tag xzFrmat;
const auto type = dnnl::memory::data_type::f32;
if(rank == 4) { // 2d
ops::ConvolutionUtils::getSizesAndIndexesConv2d(isNCHW, *input, *gradO, bS, iC, iH, iW, oC, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH, indOoH);
strides = { sH, sW };
kernel = { kH, kW };
padding = { pH, pW };
padding_r = { (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pW };
xDims = {bS, iC, iH, iW};
zDims = {bS, oC, oH, oW};
xzFrmat = isNCHW ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
}
else { // 3d
ops::ConvolutionUtils::getSizesAndIndexesConv3d(isNCHW, *input, *gradO, bS, iC, iD, iH, iW, oC, oD, oH, oW, indIOioC, indIiH, indWiC, indWoC, indWkH);
strides = { sD, sH, sW };
kernel = { kD, kH, kW };
padding = { pD, pH, pW };
padding_r = { (oD - 1) * sD - iD + kD - pD, (oH - 1) * sH - iH + kH - pH, (oW - 1) * sW - iW + kW - pW };
xDims = {bS, iC, iD, iH, iW};
zDims = {bS, oC, oD, oH, oW};
xzFrmat = isNCHW ? dnnl::memory::format_tag::ncdhw : dnnl::memory::format_tag::ndhwc;
}
// memory descriptors for arrays
// input
dnnl::memory::desc x_mkl_md = dnnl::memory::desc(xDims, type, xzFrmat);
dnnl::memory::desc x_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(input->ews() != 1 || input->ordering() != 'c') {
x_user_md.data.format_kind = dnnl_blocked; // overrides format
x_user_md.data.format_desc.blocking.strides[0] = input->strideAt(0);
x_user_md.data.format_desc.blocking.strides[1] = input->strideAt(isNCHW ? 1 :-1);
x_user_md.data.format_desc.blocking.strides[2] = input->strideAt(isNCHW ? 2 : 1);
x_user_md.data.format_desc.blocking.strides[3] = input->strideAt(isNCHW ? 3 : 2);
if(rank == 5)
x_user_md.data.format_desc.blocking.strides[4] = input->strideAt(isNCHW ? 4 : 3);
}
// gradO
dnnl::memory::desc gradO_mkl_md = dnnl::memory::desc(zDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradO_user_md = dnnl::memory::desc(zDims, type, xzFrmat);
if(gradO->ews() != 1 || gradO->ordering() != 'c') {
gradO_user_md.data.format_kind = dnnl_blocked; // overrides format
gradO_user_md.data.format_desc.blocking.strides[0] = gradO->strideAt(0);
gradO_user_md.data.format_desc.blocking.strides[1] = gradO->strideAt(isNCHW ? 1 :-1);
gradO_user_md.data.format_desc.blocking.strides[2] = gradO->strideAt(isNCHW ? 2 : 1);
gradO_user_md.data.format_desc.blocking.strides[3] = gradO->strideAt(isNCHW ? 3 : 2);
if(rank == 5)
gradO_user_md.data.format_desc.blocking.strides[4] = gradO->strideAt(isNCHW ? 4 : 3);
}
// gradI
dnnl::memory::desc gradI_mkl_md = dnnl::memory::desc(xDims, type, dnnl::memory::format_tag::any);
dnnl::memory::desc gradI_user_md = dnnl::memory::desc(xDims, type, xzFrmat);
if(gradI->ews() != 1 || gradI->ordering() != 'c') {
gradI_user_md.data.format_kind = dnnl_blocked; // overrides format
gradI_user_md.data.format_desc.blocking.strides[0] = gradI->strideAt(0);
gradI_user_md.data.format_desc.blocking.strides[1] = gradI->strideAt(isNCHW ? 1 :-1);
gradI_user_md.data.format_desc.blocking.strides[2] = gradI->strideAt(isNCHW ? 2 : 1);
gradI_user_md.data.format_desc.blocking.strides[3] = gradI->strideAt(isNCHW ? 3 : 2);
if(rank == 5)
gradI_user_md.data.format_desc.blocking.strides[4] = gradI->strideAt(isNCHW ? 4 : 3);
}
auto engine = mkldnnUtils::getEngine(LaunchContext::defaultContext()->engine());
dnnl::stream stream(engine);
// forward primitive description
dnnl::pooling_forward::desc op_ff_desc(dnnl::prop_kind::forward, mode, x_mkl_md, gradO_mkl_md, strides, kernel, padding, padding_r);
dnnl::pooling_forward::primitive_desc op_ff_prim_desc(op_ff_desc, engine);
// backward primitive description
dnnl::pooling_backward::desc op_bp_desc(mode, gradI_mkl_md, gradO_mkl_md, strides, kernel, padding, padding_r);
dnnl::pooling_backward::primitive_desc op_bp_prim_desc(op_bp_desc, engine, op_ff_prim_desc);
// arguments (memory buffers) necessary for calculations
std::unordered_map<int, dnnl::memory> args;
// gradO
auto gradO_user_mem = dnnl::memory(gradO_user_md, engine, gradO->getBuffer());
const bool gradOReorder = op_bp_prim_desc.diff_dst_desc() != gradO_user_mem.get_desc();
auto gradO_mkl_mem = gradOReorder ? dnnl::memory(op_bp_prim_desc.diff_dst_desc(), engine) : gradO_user_mem;
if (gradOReorder)
dnnl::reorder(gradO_user_mem, gradO_mkl_mem).execute(stream, gradO_user_mem, gradO_mkl_mem);
args[DNNL_ARG_DIFF_DST] = gradO_mkl_mem;
// gradI
auto gradI_user_mem = dnnl::memory(gradI_user_md, engine, gradI->getBuffer());
const bool gradIReorder = op_bp_prim_desc.diff_src_desc() != gradI_user_mem.get_desc();
auto gradI_mkl_mem = gradIReorder ? dnnl::memory(op_bp_prim_desc.diff_src_desc(), engine) : gradI_user_mem;
args[DNNL_ARG_DIFF_SRC] = gradI_mkl_mem;
if(mode == algorithm::pooling_max) {
// input
auto x_user_mem = dnnl::memory(x_user_md, engine, input->getBuffer());
const bool xReorder = op_ff_prim_desc.src_desc() != x_user_mem.get_desc();
auto x_mkl_mem = xReorder ? dnnl::memory(op_ff_prim_desc.src_desc(), engine) : x_user_mem;
if (xReorder)
dnnl::reorder(x_user_mem, x_mkl_mem).execute(stream, x_user_mem, x_mkl_mem);
args[DNNL_ARG_SRC] = x_mkl_mem;
// z
auto z_mkl_mem = dnnl::memory(op_ff_prim_desc.dst_desc(), engine);
args[DNNL_ARG_DST] = z_mkl_mem;
// auxiliary memory allocation
auto workspace = dnnl::memory(op_ff_prim_desc.workspace_desc(), engine);
args[DNNL_ARG_WORKSPACE] = workspace;
// run forward calculations
dnnl::pooling_forward(op_ff_prim_desc).execute(stream, args);
}
// run backward calculations
dnnl::pooling_backward(op_bp_prim_desc).execute(stream, args);
// reorder gradI if necessary
if (gradIReorder)
dnnl::reorder(gradI_mkl_mem, gradI_user_mem).execute(stream, gradI_mkl_mem, gradI_user_mem);
stream.wait();
}
//////////////////////////////////////////////////////////////////////////
void getMKLDNNMemoryDescLrn(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
dnnl::memory::desc* lrn_src_md, dnnl::memory::desc* lrn_diff_src_md, dnnl::memory::desc* lrn_dst_md,
dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis) {
const Nd4jLong* shape = src->getShapeInfo();
long rank = shape[0];
long dim1 = axis; // MKL-DNN supports only 1 axis, which has to be the "channel" one
long dim2 = axis >= 2 ? 1 : 2;
long dim3 = axis >= 3 ? 2 : 3;
dnnl::memory::dims lrn_src_tz = { (int)shape[1], (int)shape[dim1 + 1], rank > 2 ? (int)shape[dim2 + 1] : 1, rank > 3 ? (int)shape[dim3 + 1] : 1};
auto type = dnnl::memory::data_type::f32;
auto format = axis == 1 ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
auto supposed_to_be_any_format = format; // doesn't work with "any"
if (src != nullptr && src->getBuffer() != nullptr && lrn_src_md != nullptr) {
*lrn_src_md = dnnl::memory::desc({ lrn_src_tz }, type, supposed_to_be_any_format);
*user_src_md = dnnl::memory::desc({ lrn_src_tz }, type, format);
user_src_md->data.format_kind = dnnl_blocked;
user_src_md->data.format_desc.blocking.strides[0] = src->stridesOf()[0];
user_src_md->data.format_desc.blocking.strides[1] = src->stridesOf()[dim1];
user_src_md->data.format_desc.blocking.strides[2] = rank > 2 ? src->stridesOf()[dim2] : 1;
user_src_md->data.format_desc.blocking.strides[3] = rank > 3 ? src->stridesOf()[dim3] : 1;
}
if (diff_src != nullptr && diff_src->getBuffer() != nullptr && lrn_diff_src_md != nullptr) {
*lrn_diff_src_md = dnnl::memory::desc({ lrn_src_tz }, type, supposed_to_be_any_format);
*user_diff_src_md = dnnl::memory::desc({ lrn_src_tz }, type, format);
user_diff_src_md->data.format_kind = dnnl_blocked;
user_diff_src_md->data.format_desc.blocking.strides[0] = diff_src->stridesOf()[0];
user_diff_src_md->data.format_desc.blocking.strides[1] = diff_src->stridesOf()[dim1];
user_diff_src_md->data.format_desc.blocking.strides[2] = rank > 2 ? diff_src->stridesOf()[dim2] : 1;
user_diff_src_md->data.format_desc.blocking.strides[3] = rank > 3 ? diff_src->stridesOf()[dim3] : 1;
}
if (dst != nullptr && dst->getBuffer() != nullptr && lrn_dst_md != nullptr) {
*lrn_dst_md = dnnl::memory::desc({ lrn_src_tz }, type, supposed_to_be_any_format);
*user_dst_md = dnnl::memory::desc({ lrn_src_tz }, type, format);
user_dst_md->data.format_kind = dnnl_blocked;
user_dst_md->data.format_desc.blocking.strides[0] = dst->stridesOf()[0];
user_dst_md->data.format_desc.blocking.strides[1] = dst->stridesOf()[dim1];
user_dst_md->data.format_desc.blocking.strides[2] = rank > 2 ? dst->stridesOf()[dim2] : 1;
user_dst_md->data.format_desc.blocking.strides[3] = rank > 3 ? dst->stridesOf()[dim3] : 1;
}
}
//////////////////////////////////////////////////////////////////////////
dnnl::engine& getEngine(void *ptr) {
auto eng = reinterpret_cast<dnnl::engine*>(ptr);
return *eng;
}
/*
//////////////////////////////////////////////////////////////////////////
void getMKLDNNMemoryDescPool2d(
int kH, int kW, int sH, int sW, int pH, int pW, int dH, int dW, int poolingMode, int extraParam0, bool isNCHW,
@ -307,104 +617,51 @@ void getMKLDNNMemoryDescConv3d(
}
};
// void getMKLDNNMemoryDescBatchNorm(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
// dnnl::memory::desc* batchnorm_src_md, dnnl::memory::desc* batchnorm_diff_src_md, dnnl::memory::desc* batchnorm_dst_md,
// dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis) {
// const Nd4jLong* shape = src->getShapeInfo();
// Nd4jLong rank = shape[0];
// Nd4jLong dim1 = axis; // MKL-DNN supports only 1 axis, which has to be the "channel" one
// Nd4jLong dim2 = axis >= 2 ? 1 : 2;
// Nd4jLong dim3 = axis >= 3 ? 2 : 3;
// dnnl::memory::dims batchnorm_src_tz = { (int)shape[1], (int)shape[dim1 + 1], rank > 2 ? (int)shape[dim2 + 1] : 1, rank > 3 ? (int)shape[dim3 + 1] : 1};
// auto type = dnnl::memory::data_type::f32;
// auto format = dnnl::memory::format_tag::nchw;
// auto supposed_to_be_any_format = dnnl::memory::format_tag::nChw8c; // doesn't work with "any"
// if (src != nullptr && src->getBuffer() != nullptr && batchnorm_src_md != nullptr) {
// *batchnorm_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, supposed_to_be_any_format);
// *user_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, format);
// user_src_md->data.format_kind = dnnl_blocked; // overrides format
// user_src_md->data.format_desc.blocking.strides[0] = src->stridesOf()[0];
// user_src_md->data.format_desc.blocking.strides[1] = src->stridesOf()[dim1];
// user_src_md->data.format_desc.blocking.strides[2] = rank > 2 ? src->stridesOf()[dim2] : 1;
// user_src_md->data.format_desc.blocking.strides[3] = rank > 3 ? src->stridesOf()[dim3] : 1;
// }
// if (diff_src != nullptr && diff_src->getBuffer() != nullptr && batchnorm_diff_src_md != nullptr) {
// *batchnorm_diff_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, supposed_to_be_any_format);
// *user_diff_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, format);
// user_diff_src_md->data.format_kind = dnnl_blocked; // overrides format
// user_diff_src_md->data.format_desc.blocking.strides[0] = diff_src->stridesOf()[0];
// user_diff_src_md->data.format_desc.blocking.strides[1] = diff_src->stridesOf()[dim1];
// user_diff_src_md->data.format_desc.blocking.strides[2] = rank > 2 ? diff_src->stridesOf()[dim2] : 1;
// user_diff_src_md->data.format_desc.blocking.strides[3] = rank > 3 ? diff_src->stridesOf()[dim3] : 1;
// }
// if (dst != nullptr && dst->getBuffer() != nullptr && batchnorm_dst_md != nullptr) {
// *batchnorm_dst_md = dnnl::memory::desc({ batchnorm_src_tz }, type, supposed_to_be_any_format);
// *user_dst_md = dnnl::memory::desc({ batchnorm_src_tz }, type, format);
// user_dst_md->data.format_kind = dnnl_blocked; // overrides format
// user_dst_md->data.format_desc.blocking.strides[0] = dst->stridesOf()[0];
// user_dst_md->data.format_desc.blocking.strides[1] = dst->stridesOf()[dim1];
// user_dst_md->data.format_desc.blocking.strides[2] = rank > 2 ? dst->stridesOf()[dim2] : 1;
// user_dst_md->data.format_desc.blocking.strides[3] = rank > 3 ? dst->stridesOf()[dim3] : 1;
// }
// };
//////////////////////////////////////////////////////////////////////////
void getMKLDNNMemoryDescLrn(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
dnnl::memory::desc* lrn_src_md, dnnl::memory::desc* lrn_diff_src_md, dnnl::memory::desc* lrn_dst_md,
dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis) {
void getMKLDNNMemoryDescBatchNorm(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
dnnl::memory::desc* batchnorm_src_md, dnnl::memory::desc* batchnorm_diff_src_md, dnnl::memory::desc* batchnorm_dst_md,
dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis) {
const Nd4jLong* shape = src->getShapeInfo();
long rank = shape[0];
long dim1 = axis; // MKL-DNN supports only 1 axis, which has to be the "channel" one
long dim2 = axis >= 2 ? 1 : 2;
long dim3 = axis >= 3 ? 2 : 3;
dnnl::memory::dims lrn_src_tz = { (int)shape[1], (int)shape[dim1 + 1], rank > 2 ? (int)shape[dim2 + 1] : 1, rank > 3 ? (int)shape[dim3 + 1] : 1};
Nd4jLong rank = shape[0];
Nd4jLong dim1 = axis; // MKL-DNN supports only 1 axis, which has to be the "channel" one
Nd4jLong dim2 = axis >= 2 ? 1 : 2;
Nd4jLong dim3 = axis >= 3 ? 2 : 3;
dnnl::memory::dims batchnorm_src_tz = { (int)shape[1], (int)shape[dim1 + 1], rank > 2 ? (int)shape[dim2 + 1] : 1, rank > 3 ? (int)shape[dim3 + 1] : 1};
auto type = dnnl::memory::data_type::f32;
auto format = axis == 1 ? dnnl::memory::format_tag::nchw : dnnl::memory::format_tag::nhwc;
auto supposed_to_be_any_format = format; // doesn't work with "any"
auto format = dnnl::memory::format_tag::nchw;
auto supposed_to_be_any_format = dnnl::memory::format_tag::nChw8c; // doesn't work with "any"
if (src != nullptr && src->getBuffer() != nullptr && lrn_src_md != nullptr) {
*lrn_src_md = dnnl::memory::desc({ lrn_src_tz }, type, supposed_to_be_any_format);
*user_src_md = dnnl::memory::desc({ lrn_src_tz }, type, format);
user_src_md->data.format_kind = dnnl_blocked;
if (src != nullptr && src->getBuffer() != nullptr && batchnorm_src_md != nullptr) {
*batchnorm_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, supposed_to_be_any_format);
*user_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, format);
user_src_md->data.format_kind = dnnl_blocked; // overrides format
user_src_md->data.format_desc.blocking.strides[0] = src->stridesOf()[0];
user_src_md->data.format_desc.blocking.strides[1] = src->stridesOf()[dim1];
user_src_md->data.format_desc.blocking.strides[2] = rank > 2 ? src->stridesOf()[dim2] : 1;
user_src_md->data.format_desc.blocking.strides[3] = rank > 3 ? src->stridesOf()[dim3] : 1;
}
if (diff_src != nullptr && diff_src->getBuffer() != nullptr && lrn_diff_src_md != nullptr) {
*lrn_diff_src_md = dnnl::memory::desc({ lrn_src_tz }, type, supposed_to_be_any_format);
*user_diff_src_md = dnnl::memory::desc({ lrn_src_tz }, type, format);
user_diff_src_md->data.format_kind = dnnl_blocked;
if (diff_src != nullptr && diff_src->getBuffer() != nullptr && batchnorm_diff_src_md != nullptr) {
*batchnorm_diff_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, supposed_to_be_any_format);
*user_diff_src_md = dnnl::memory::desc({ batchnorm_src_tz }, type, format);
user_diff_src_md->data.format_kind = dnnl_blocked; // overrides format
user_diff_src_md->data.format_desc.blocking.strides[0] = diff_src->stridesOf()[0];
user_diff_src_md->data.format_desc.blocking.strides[1] = diff_src->stridesOf()[dim1];
user_diff_src_md->data.format_desc.blocking.strides[2] = rank > 2 ? diff_src->stridesOf()[dim2] : 1;
user_diff_src_md->data.format_desc.blocking.strides[3] = rank > 3 ? diff_src->stridesOf()[dim3] : 1;
}
if (dst != nullptr && dst->getBuffer() != nullptr && lrn_dst_md != nullptr) {
*lrn_dst_md = dnnl::memory::desc({ lrn_src_tz }, type, supposed_to_be_any_format);
*user_dst_md = dnnl::memory::desc({ lrn_src_tz }, type, format);
user_dst_md->data.format_kind = dnnl_blocked;
if (dst != nullptr && dst->getBuffer() != nullptr && batchnorm_dst_md != nullptr) {
*batchnorm_dst_md = dnnl::memory::desc({ batchnorm_src_tz }, type, supposed_to_be_any_format);
*user_dst_md = dnnl::memory::desc({ batchnorm_src_tz }, type, format);
user_dst_md->data.format_kind = dnnl_blocked; // overrides format
user_dst_md->data.format_desc.blocking.strides[0] = dst->stridesOf()[0];
user_dst_md->data.format_desc.blocking.strides[1] = dst->stridesOf()[dim1];
user_dst_md->data.format_desc.blocking.strides[2] = rank > 2 ? dst->stridesOf()[dim2] : 1;
user_dst_md->data.format_desc.blocking.strides[3] = rank > 3 ? dst->stridesOf()[dim3] : 1;
}
}
//////////////////////////////////////////////////////////////////////////
dnnl::engine& getEngine(void *ptr) {
auto eng = reinterpret_cast<dnnl::engine*>(ptr);
return *eng;
}
};
*/
}
}

22
libnd4j/include/ops/declarable/platform/mkldnn/mkldnnUtils.h
View File

@ -16,6 +16,7 @@
//
// @author saudet
// @author Yurii Shyrma (iuriish@yahoo.com)
//
#ifndef DEV_TESTS_MKLDNNUTILS_H
@ -81,17 +82,27 @@ namespace nd4j{
DECLARE_PLATFORM(deconv3d_bp, ENGINE_CPU);
DECLARE_PLATFORM(depthwise_conv2d, ENGINE_CPU);
DECLARE_PLATFORM(depthwise_conv2d_bp, ENGINE_CPU);
}
}
namespace mkldnnUtils {
void poolingMKLDNN(const NDArray *input, NDArray *output, const int kD, const int kH, const int kW, const int sD, const int sH, const int sW, const int pD, const int pH, const int pW, const int isNCHW, const dnnl::algorithm mode);
void poolingBpMKLDNN(const NDArray *input, const NDArray *gradO, NDArray *gradI, const int kD, const int kH, const int kW, const int sD, const int sH, const int sW, const int pD, const int pH, const int pW, const int isNCHW, const dnnl::algorithm mode);
void getMKLDNNMemoryDescLrn(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
dnnl::memory::desc* lrn_src_md, dnnl::memory::desc* lrn_diff_src_md, dnnl::memory::desc* lrn_dst_md,
dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis);
dnnl::engine& getEngine(void *ptr);
/**
* Utility methods for MKLDNN
*/
void getMKLDNNMemoryDescConv2d(
/* void getMKLDNNMemoryDescConv2d(
int kH, int kW, int sH, int sW, int pH, int pW, int dH, int dW, const int paddingMode, bool isNCHW,
int bS, int iC, int iH, int iW, int oC, int oH, int oW, const NDArray* src, const NDArray* diff_src,
const NDArray* weights, const NDArray* diff_weights, const NDArray* bias, const NDArray* dst,
@ -130,12 +141,7 @@ namespace nd4j{
void getMKLDNNMemoryDescBatchNorm(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
dnnl::memory::desc* batchnorm_src_md, dnnl::memory::desc* batchnorm_diff_src_md, dnnl::memory::desc* batchnorm_dst_md,
dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis);
void getMKLDNNMemoryDescLrn(const NDArray* src, const NDArray* diff_src, const NDArray* dst,
dnnl::memory::desc* lrn_src_md, dnnl::memory::desc* lrn_diff_src_md, dnnl::memory::desc* lrn_dst_md,
dnnl::memory::desc* user_src_md, dnnl::memory::desc* user_diff_src_md, dnnl::memory::desc* user_dst_md, int axis);
dnnl::engine& getEngine(void *ptr);
*/
}
}

115
libnd4j/tests_cpu/layers_tests/DeclarableOpsTests1.cpp
View File

@ -2031,121 +2031,6 @@ TEST_F(DeclarableOpsTests1, Sum1) {
}
*/
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests1, Avgpool2d_test1) {
auto x = NDArrayFactory::create_<float>('c', {bS,iD,iH,iW});
auto exp = NDArrayFactory::create<float>('c',{bS,iD,oH,oW});
// auto z('c',{bS,iD,oH,oW});
auto variableSpace = new VariableSpace();
variableSpace->putVariable(-1, x);
// variableSpace->putVariable(1, &z);
auto block = new Context(1, variableSpace, false);
block->fillInputs({-1});
std::vector<int>* argI = block->getIArguments();
*argI = {kH,kW, sH,sW, pH,pW, dW,dH, 0, 0, 0}; // 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
nd4j::ops::avgpool2d pooling;
Nd4jStatus status = pooling.execute(block);
ASSERT_EQ(ND4J_STATUS_OK, status);
auto result = variableSpace->getVariable(block->getNodeId())->getNDArray();
ASSERT_TRUE(exp.isSameShape(result));
delete variableSpace;
delete block;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests1, Avgpool2d_test2) {
const int bS = 2;
const int iD = 1;
const int iH = 28;
const int iW = 28;
const int kH = 5;
const int kW = 5;
const int sH = 1;
const int sW = 1;
const int pH = 0;
const int pW = 0;
const int dH = 1;
const int dW = 1;
const int oH = (iH - kH - (kH-1)*(dH-1) + 2*pH)/sH + 1; // output height
const int oW = (iW - kW - (kW-1)*(dW-1) + 2*pW)/sW + 1; // output width
auto x = NDArrayFactory::create_<float>('c', {bS,iD,iH,iW});
auto exp = NDArrayFactory::create<float>('c',{bS,iD,oH,oW});
// auto z('c',{bS,iD,oH,oW});
auto variableSpace = new VariableSpace();
variableSpace->putVariable(-1, x);
// variableSpace->putVariable(1, &z);
auto block = new Context(1, variableSpace, false);
block->fillInputs({-1});
std::vector<int>* argI = block->getIArguments();
*argI = {kH,kW, sH,sW, pH,pW, dW,dH, 0, 0, 0}; // 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
nd4j::ops::avgpool2d pooling;
Nd4jStatus status = pooling.execute(block);
ASSERT_EQ(ND4J_STATUS_OK, status);
auto result = variableSpace->getVariable(block->getNodeId())->getNDArray();
// result->printShapeInfo();
ASSERT_TRUE(exp.isSameShape(result));
delete variableSpace;
delete block;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests1, Avgpool2d_test3) {
const int bS = 2;
const int iD = 1;
const int iH = 28;
const int iW = 28;
const int kH = 5;
const int kW = 5;
const int sH = 1;
const int sW = 1;
const int pH = 0;
const int pW = 0;
const int dH = 1;
const int dW = 1;
const int oH = (int) nd4j::math::nd4j_ceil<float, int>(iH * 1.f / sH);
const int oW = (int) nd4j::math::nd4j_ceil<float, int>(iW * 1.f / sW);
auto x = NDArrayFactory::create_<float>('c', {bS,iD,iH,iW});
auto exp = NDArrayFactory::create<float>('c',{bS,iD,oH,oW});
// auto z('c',{bS,iD,oH,oW});
auto variableSpace = new VariableSpace();
variableSpace->putVariable(-1, x);
// variableSpace->putVariable(1, &z);
auto block = new Context(1, variableSpace, false);
block->fillInputs({-1});
std::vector<int>* argI = block->getIArguments();
*argI = {kH,kW, sH,sW, pH,pW, dW,dH, 1, 0, 0}; // 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
nd4j::ops::avgpool2d pooling;
Nd4jStatus status = pooling.execute(block);
ASSERT_EQ(ND4J_STATUS_OK, status);
auto result = variableSpace->getVariable(block->getNodeId())->getNDArray();
// result->printShapeInfo();
ASSERT_TRUE(exp.isSameShape(result));
delete variableSpace;
delete block;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests1, Pnormpool2d1) {

155
libnd4j/tests_cpu/layers_tests/DeclarableOpsTests4.cpp
View File

@ -360,7 +360,6 @@ TEST_F(DeclarableOpsTests4, avgpool2d_12) {
917.5, 918.5, 919.5, 925. , 926. , 927. , 934. , 935. , 936. , 941.5, 942.5, 943.5, 992.5, 993.5, 994.5,1000. , 1001. , 1002. ,1009. , 1010. , 1011. ,1016.5, 1017.5, 1018.5,
1082.5, 1083.5, 1084.5,1090. , 1091. , 1092. ,1099. , 1100. , 1101. ,1106.5, 1107.5, 1108.5,1157.5, 1158.5, 1159.5,1165. , 1166. , 1167. ,1174. , 1175. , 1176. ,1181.5, 1182.5, 1183.5});
input.linspace(1.);
input.syncToDevice();
nd4j::ops::avgpool2d op;
auto results = op.evaluate({&input}, {kH,kW, sH,sW, pH,pW, dH,dW, paddingMode, 0, dataFormat});
@ -377,6 +376,160 @@ TEST_F(DeclarableOpsTests4, avgpool2d_12) {
delete results;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests4, avgpool2d_13) {
const int bS = 2; // batch size
const int iD = 1; // input depth (number of picture channels, for example rgb=3)
const int iH = 28; // picture height in pixels
const int iW = 28; // picture width in pixels
const int kH = 5; // kernel height in pixels
const int kW = 5; // kernel width in pixels
const int sH = 1; // stride step in horizontal direction
const int sW = 1; // stride step in vertical direction
const int pH = 0; // padding height
const int pW = 0; // padding width
const int dH = 2; // dilation height
const int dW = 2; // dilation width
const int oH = (iH - kH - (kH-1)*(dH-1) + 2*pH)/sH + 1; // output height
const int oW = (iW - kW - (kW-1)*(dW-1) + 2*pW)/sW + 1; // output width
auto x = NDArrayFactory::create_<float>('c', {bS,iD,iH,iW});
auto exp = NDArrayFactory::create<float>('c',{bS,iD,oH,oW});
// auto z('c',{bS,iD,oH,oW});
auto variableSpace = new VariableSpace();
variableSpace->putVariable(-1, x);
// variableSpace->putVariable(1, &z);
auto block = new Context(1, variableSpace, false);
block->fillInputs({-1});
std::vector<int>* argI = block->getIArguments();
*argI = {kH,kW, sH,sW, pH,pW, dW,dH, 0, 0, 0}; // 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
nd4j::ops::avgpool2d pooling;
Nd4jStatus status = pooling.execute(block);
ASSERT_EQ(ND4J_STATUS_OK, status);
auto result = variableSpace->getVariable(block->getNodeId())->getNDArray();
ASSERT_TRUE(exp.isSameShape(result));
delete variableSpace;
delete block;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests4, avgpool2d_14) {
const int bS = 2;
const int iD = 1;
const int iH = 28;
const int iW = 28;
const int kH = 5;
const int kW = 5;
const int sH = 1;
const int sW = 1;
const int pH = 0;
const int pW = 0;
const int dH = 1;
const int dW = 1;
const int oH = (iH - kH - (kH-1)*(dH-1) + 2*pH)/sH + 1; // output height
const int oW = (iW - kW - (kW-1)*(dW-1) + 2*pW)/sW + 1; // output width
auto x = NDArrayFactory::create_<float>('c', {bS,iD,iH,iW});
auto exp = NDArrayFactory::create<float>('c',{bS,iD,oH,oW});
// auto z('c',{bS,iD,oH,oW});
auto variableSpace = new VariableSpace();
variableSpace->putVariable(-1, x);
// variableSpace->putVariable(1, &z);
auto block = new Context(1, variableSpace, false);
block->fillInputs({-1});
std::vector<int>* argI = block->getIArguments();
*argI = {kH,kW, sH,sW, pH,pW, dW,dH, 0, 0, 0}; // 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
nd4j::ops::avgpool2d pooling;
Nd4jStatus status = pooling.execute(block);
ASSERT_EQ(ND4J_STATUS_OK, status);
auto result = variableSpace->getVariable(block->getNodeId())->getNDArray();
// result->printShapeInfo();
ASSERT_TRUE(exp.isSameShape(result));
delete variableSpace;
delete block;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests4, Avgpool2d_test15) {
const int bS = 2;
const int iD = 1;
const int iH = 28;
const int iW = 28;
const int kH = 5;
const int kW = 5;
const int sH = 1;
const int sW = 1;
const int pH = 0;
const int pW = 0;
const int dH = 1;
const int dW = 1;
const int oH = (int) nd4j::math::nd4j_ceil<float, int>(iH * 1.f / sH);
const int oW = (int) nd4j::math::nd4j_ceil<float, int>(iW * 1.f / sW);
auto x = NDArrayFactory::create_<float>('c', {bS,iD,iH,iW});
auto exp = NDArrayFactory::create<float>('c',{bS,iD,oH,oW});
// auto z('c',{bS,iD,oH,oW});
auto variableSpace = new VariableSpace();
variableSpace->putVariable(-1, x);
// variableSpace->putVariable(1, &z);
auto block = new Context(1, variableSpace, false);
block->fillInputs({-1});
std::vector<int>* argI = block->getIArguments();
*argI = {kH,kW, sH,sW, pH,pW, dW,dH, 1, 0, 0}; // 0,1 - kernel Height/Width; 2,3 - stride Height/Width; 4,5 - pad Height/Width; 6,7 - dilation Height/Width; 8 - same mode;
nd4j::ops::avgpool2d pooling;
Nd4jStatus status = pooling.execute(block);
ASSERT_EQ(ND4J_STATUS_OK, status);
auto result = variableSpace->getVariable(block->getNodeId())->getNDArray();
// result->printShapeInfo();
ASSERT_TRUE(exp.isSameShape(result));
delete variableSpace;
delete block;
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests4, avgpool2d_16) {
int bS=2, iH=4,iW=4, iC=2, kH=2,kW=2, sH=2,sW=2, pH=0,pW=0, dH=1,dW=1;
int oH=2,oW=2;
int paddingMode = 1; // 1-SAME, 0-VALID
int dataFormat = 1; // 1-NHWC, 0-NDHW
NDArray input('c', {bS, iH, iW, iC}, nd4j::DataType::FLOAT32);
NDArray output('f', {bS, oH, oW, iC}, nd4j::DataType::FLOAT32);
NDArray expected('c', {bS, oH, oW, iC}, {6.f, 7.f, 10.f, 11.f, 22.f, 23.f, 26.f, 27.f, 38.f, 39.f, 42.f, 43.f, 54.f, 55.f, 58.f, 59.f}, nd4j::DataType::FLOAT32);
input.linspace(1.);
nd4j::ops::avgpool2d op;
auto status = op.execute({&input}, {&output}, {}, {kH,kW, sH,sW, pH,pW, dH,dW, paddingMode, 0, dataFormat}, {});
ASSERT_EQ(Status::OK(), status);
// output.printBuffer();
//expected.printIndexedBuffer("expected");
ASSERT_TRUE(expected.equalsTo(output));
}
//////////////////////////////////////////////////////////////////////
TEST_F(DeclarableOpsTests4, biasadd_1) {
auto x = NDArrayFactory::create<double>('c', {2, 3, 3, 2});

52
libnd4j/tests_cpu/layers_tests/PlaygroundTests.cpp
View File

@ -422,50 +422,38 @@ TEST_F(PlaygroundTests, my) {
delete variableSpace;
}
#include<ops/declarable/helpers/batchnorm.h>
TEST_F(PlaygroundTests, my) {
const int N = 10000;
const Nd4jLong dim0(128), dim1(128), dim2(128);
int N = 100;
int bS=16, iH=128,iW=128, iC=32,oC=64, kH=4,kW=4, sH=1,sW=1, pH=0,pW=0, dH=1,dW=1;
int oH=128,oW=128;
NDArray input('c', {dim0,dim1,dim2}, nd4j::DataType::DOUBLE);
NDArray mean('c', {dim1}, nd4j::DataType::DOUBLE);
NDArray variance('c', {dim1}, nd4j::DataType::DOUBLE);
NDArray gamma('c', {dim1}, nd4j::DataType::DOUBLE);
NDArray beta ('c', {dim1}, nd4j::DataType::DOUBLE);
int paddingMode = 1; // 1-SAME, 0-VALID;
int dataFormat = 1; // 1-NHWC, 0-NCHW
NDArray output('c', {dim0,dim1,dim2}, nd4j::DataType::DOUBLE);
// NDArray input('c', {bS, iC, iH, iW}, nd4j::DataType::FLOAT32);
// NDArray output('c', {bS, oC, oH, oW}, nd4j::DataType::FLOAT32);
NDArray input('c', {bS, iH, iW, iC}, nd4j::DataType::FLOAT32);
NDArray output('c', {bS, oH, oW, oC}, nd4j::DataType::FLOAT32);
// NDArray weights('c', {kH, kW, iC, oC}, nd4j::DataType::FLOAT32); // permute [kH, kW, iC, oC] -> [oC, iC, kH, kW]
NDArray weights('c', {oC, iC, kH, kW}, nd4j::DataType::FLOAT32);
NDArray bias('c', {oC}, nd4j::DataType::FLOAT32);
input.linspace(-100, 0.1);
mean.linspace(-50, 0.15);
variance.linspace(-5, 0.2);
gamma = 1.5;
beta = -2.5;
input = 5.;
weights = 3.;
bias = 1.;
// warm up
ops::helpers::batchnorm(&input, &mean, &variance, &gamma, &beta, &output, {1}, 1e-5);
nd4j::ops::conv2d op;
auto err = op.execute({&input, &weights, &bias}, {&output}, {kH,kW, sH,sW, pH,pW, dH,dW, paddingMode, dataFormat});
auto timeStart = std::chrono::system_clock::now();
for (int i = 0; i < N; ++i)
ops::helpers::batchnorm(&input, &mean, &variance, &gamma, &beta, &output, {1}, 1e-5);
err = op.execute({&input, &weights, &bias}, {&output}, {kH,kW, sH,sW, pH,pW, dH,dW, paddingMode, dataFormat});
auto timeEnd = std::chrono::system_clock::now();
auto time = std::chrono::duration_cast<std::chrono::microseconds> ((timeEnd - timeStart)/N).count();
printf("time: %li \n", time);
auto time = std::chrono::duration_cast<std::chrono::microseconds> ((timeEnd - timeStart) / N).count();
printf("time: %i \n", time);
}
*/