Skip to content

Conv2dMemFuse: Fix Conv2d parameters #16

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
pavan02 wants to merge 15 commits into
base: master
Choose a base branch
from
Open

Conv2dMemFuse: Fix Conv2d parameters #16

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
15 commits
Select commit Hold shift + click to select a range
6a354f2
Conv2dMemFuse: Fix Conv2d parameters
Feb 11, 2022
3eb7150
Conv2dProjTest: Add input, kernel, stride input from cli
pavan02 Feb 12, 2022
70ac087
Merge pull request #17 from asu-cactus/conv2dprojtest/add-dimension-cli
pavan02 Feb 12, 2022
1274eaf
Conv2dMemFuse: Fix Conv2d parameters
Feb 11, 2022
dd3ea9e
Merge branch 'Conv2dMemFuse/fix-parameters' of https://github.com/asu…
pavan02 Feb 19, 2022
fefc998
conv2d memfuse fix
pavan02 Feb 21, 2022
e3a46fc
double changes
Feb 22, 2022
72357e6
double changes
Feb 22, 2022
c68a41a
read input, kernel dimensions from cli arguments
pavan02 Apr 9, 2022
acd9fba
Merge branch 'Conv2dMemFuse/fix-parameters' of https://github.com/asu…
pavan02 Apr 9, 2022
823292c
large case allocation block
Apr 10, 2022
b5c315c
large image load
Apr 11, 2022
f612e63
Working for 2500*2500*3, 64,3,1,1
Apr 17, 2022
141bc2e
Remove the unnecessary sleep(1) statement from PangeaStorageServer's …
jiazou-bigdata May 2, 2022
15d63ff
Merge branch 'Conv2dMemFuse/fix-parameters' of https://github.com/asu…
pavan02 May 4, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
186 changes: 93 additions & 93 deletions src/FF/headers/FFInputLayerJoin.h
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,93 +1,93 @@
#ifndef FF_INPUTLAYER_JOIN_H
#define FF_INPUTLAYER_JOIN_H

#include "FFMatrixBlock.h"
#include "JoinComp.h"
#include "Lambda.h"
#include "LambdaCreationFunctions.h"

// LA libraries:
#include <eigen3/Eigen/Dense>

using namespace pdb;

class FFInputLayerJoin
: public JoinComp<FFMatrixBlock, FFMatrixBlock, FFMatrixBlock> {

public:
ENABLE_DEEP_COPY

FFInputLayerJoin() {}

Lambda<bool> getSelection(Handle<FFMatrixBlock> in1,
Handle<FFMatrixBlock> in2) override {
// return makeLambda(

// in1, in2, [](Handle<FFMatrixBlock> &in1, Handle<FFMatrixBlock> &in2)
// {
// return in1->getBlockColIndex() == in2->getBlockRowIndex();
// });
return makeLambdaFromMethod(in1, getBlockColIndex) ==
makeLambdaFromMethod(in2, getBlockRowIndex);
}

Lambda<Handle<FFMatrixBlock>>
getProjection(Handle<FFMatrixBlock> in1, Handle<FFMatrixBlock> in2) override {
return makeLambda(
in1, in2, [](Handle<FFMatrixBlock> &in1, Handle<FFMatrixBlock> &in2) {
if (FFMatrixBlock::librayCode == EIGEN_CODE) {
// get the sizes
uint32_t I = in1->getRowNums();
uint32_t J = in2->getColNums();
uint32_t K = in1->getColNums();

// make an output
pdb::Handle<FFMatrixBlock> resultFFMatrixBlock =
pdb::makeObject<FFMatrixBlock>(
in1->getBlockRowIndex(), in2->getBlockColIndex(), I, J,
in1->getTotalRowNums(), in2->getTotalColNums(), false);

// get the ptrs
double *outData = resultFFMatrixBlock->getValue().rawData->c_ptr();
double *in1Data = in1->getValue().rawData->c_ptr();
double *in2Data = in2->getValue().rawData->c_ptr();

// do the multiply
Eigen::Map<Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic,
Eigen::RowMajor>>
currentMatrix1(in1->getValue().rawData->c_ptr(),
in1->getRowNums(), in1->getColNums());
Eigen::Map<Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic,
Eigen::RowMajor>>
currentMatrix2(in2->getValue().rawData->c_ptr(),
in2->getRowNums(), in2->getColNums());

Eigen::Map<Eigen::Matrix<double, Eigen::Dynamic, Eigen::Dynamic,
Eigen::RowMajor>>
productMatrix(resultFFMatrixBlock->getValue().rawData->c_ptr(),
I, J);

productMatrix = currentMatrix1 * currentMatrix2;

// process the bias if necessary
if (in2->getValue().bias != nullptr) {
auto bias = in2->getValue().bias->c_ptr();
for (uint32_t r = 0; r < I; r++) {
for (uint32_t c = 0; c < J; c++) {

// add the bias
outData[r * J + c] += bias[c];
}
}
}

return resultFFMatrixBlock;
} else {
std::cerr << "Wrong librayCode!" << std::endl;
exit(1);
}
});
}
};

#endif
#ifndef FF_INPUTLAYER_JOIN_H
#define FF_INPUTLAYER_JOIN_H
#include "FFMatrixBlock.h"
#include "JoinComp.h"
#include "Lambda.h"
#include "LambdaCreationFunctions.h"
// LA libraries:
#include <eigen3/Eigen/Dense>
using namespace pdb;
class FFInputLayerJoin
: public JoinComp<FFMatrixBlock, FFMatrixBlock, FFMatrixBlock> {
public:
ENABLE_DEEP_COPY
FFInputLayerJoin() {}
Lambda<bool> getSelection(Handle<FFMatrixBlock> in1,
Handle<FFMatrixBlock> in2) override {
// return makeLambda(
// in1, in2, [](Handle<FFMatrixBlock> &in1, Handle<FFMatrixBlock> &in2)
// {
// return in1->getBlockColIndex() == in2->getBlockRowIndex();
// });
return makeLambdaFromMethod(in1, getBlockColIndex) ==
makeLambdaFromMethod(in2, getBlockRowIndex);
}
Lambda<Handle<FFMatrixBlock>>
getProjection(Handle<FFMatrixBlock> in1, Handle<FFMatrixBlock> in2) override {
return makeLambda(
in1, in2, [](Handle<FFMatrixBlock> &in1, Handle<FFMatrixBlock> &in2) {
if (FFMatrixBlock::librayCode == EIGEN_CODE) {
// get the sizes
uint32_t I = in1->getRowNums();
uint32_t J = in2->getColNums();
uint32_t K = in1->getColNums();
// make an output
pdb::Handle<FFMatrixBlock> resultFFMatrixBlock =
pdb::makeObject<FFMatrixBlock>(
in1->getBlockRowIndex(), in2->getBlockColIndex(), I, J,
in1->getTotalRowNums(), in2->getTotalColNums(), false);
// get the ptrs
float *outData = resultFFMatrixBlock->getValue().rawData->c_ptr();
float *in1Data = in1->getValue().rawData->c_ptr();
float *in2Data = in2->getValue().rawData->c_ptr();
// do the multiply
Eigen::Map<Eigen::Matrix<float, Eigen::Dynamic, Eigen::Dynamic,
Eigen::RowMajor>>
currentMatrix1(in1->getValue().rawData->c_ptr(),
in1->getRowNums(), in1->getColNums());
Eigen::Map<Eigen::Matrix<float, Eigen::Dynamic, Eigen::Dynamic,
Eigen::RowMajor>>
currentMatrix2(in2->getValue().rawData->c_ptr(),
in2->getRowNums(), in2->getColNums());
Eigen::Map<Eigen::Matrix<float, Eigen::Dynamic, Eigen::Dynamic,
Eigen::RowMajor>>
productMatrix(resultFFMatrixBlock->getValue().rawData->c_ptr(),
I, J);
productMatrix = currentMatrix1 * currentMatrix2;
// process the bias if necessary
if (in2->getValue().bias != nullptr) {
auto bias = in2->getValue().bias->c_ptr();
for (uint32_t r = 0; r < I; r++) {
for (uint32_t c = 0; c < J; c++) {
// add the bias
outData[r * J + c] += bias[c];
}
}
}
return resultFFMatrixBlock;
} else {
std::cerr << "Wrong librayCode!" << std::endl;
exit(1);
}
});
}
};
#endif
Loading