p35189024
/
InfiniTensor
forked from jiuyuan/InfiniTensor
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

format

This commit is contained in:
wanghailu 2023-01-05 14:23:46 +08:00
parent 156a40806d
commit 2b8bca17e2
60 changed files with 517 additions and 425 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
include/core/data_type.h
View File

@ -7,7 +7,8 @@ class DataType {
static const DataType Float32;
static const DataType UInt32;
static const DataType Int32;
static constexpr size_t sizePerElement[]{sizeof(float), sizeof(uint32_t), sizeof(int32_t)};
static constexpr size_t sizePerElement[]{sizeof(float), sizeof(uint32_t),
sizeof(int32_t)};
static constexpr std::string_view names[]{"Float32", "UInt32", "Int32"};
private:

2
include/core/operator.h
View File

@ -250,7 +250,7 @@ class OperatorObj : public Object {
public:
OperatorObj(OpType opType, TensorVec inputs, TensorVec outputs);
OperatorObj(OpType opType);
void setInputs(TensorVec inputsTensor) {
void setInputs(TensorVec inputsTensor) {
inputs = inputsTensor;
for (auto &t : inputs)
IT_ASSERT(t != nullptr);

14
include/operators/activation_backward.h
View File

@ -4,7 +4,8 @@
namespace infini {
class ActivationBackwardObj : public OperatorObj {
public:
ActivationBackwardObj(OpType type, GraphObj *graph, Tensor y, Tensor diff_y, Tensor x, Tensor diff_x);
ActivationBackwardObj(OpType type, GraphObj *graph, Tensor y, Tensor diff_y,
Tensor x, Tensor diff_x);
optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
std::string toString() const override;
@ -16,11 +17,12 @@ class ActivationBackwardObj : public OperatorObj {
vector<int> getOpAttrVector() const override;
};
#define DEFINE_ACTIVATION_BACKWARD_OBJ(prefix, type) \
class prefix##Obj : public ActivationBackwardObj { \
public: \
prefix##Obj(GraphObj *graph, Tensor y, Tensor diff_y, Tensor x, Tensor diff_x) \
: ActivationBackwardObj(type, graph, y, diff_y, x, diff_x) {} \
#define DEFINE_ACTIVATION_BACKWARD_OBJ(prefix, type) \
class prefix##Obj : public ActivationBackwardObj { \
public: \
prefix##Obj(GraphObj *graph, Tensor y, Tensor diff_y, Tensor x, \
Tensor diff_x) \
: ActivationBackwardObj(type, graph, y, diff_y, x, diff_x) {} \
};
DEFINE_ACTIVATION_BACKWARD_OBJ(ReluBackward, OpType::ReluBackward)

15
include/operators/conv.h
View File

@ -91,14 +91,15 @@ class ConvBackwardFilterObj : public ConvBaseObj {
ActType act;
public:
ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Tensor diffY, Tensor diffW, int ph,
int pw, int sh = 1, int sw = 1, int dh = 1, int dw = 1,
Tensor bias = nullptr, ActType act = ActType::None);
ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Tensor diffY,
Tensor diffW, int ph, int pw, int sh = 1, int sw = 1,
int dh = 1, int dw = 1, Tensor bias = nullptr,
ActType act = ActType::None);
// Constructors for setting padding mode
ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Tensor diffY, Tensor diffW,
PaddingMode mode = PaddingMode::Same, int sh = 1, int sw = 1,
int dh = 1, int dw = 1, Tensor bias = nullptr,
ActType act = ActType::None);
ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Tensor diffY,
Tensor diffW, PaddingMode mode = PaddingMode::Same,
int sh = 1, int sw = 1, int dh = 1, int dw = 1,
Tensor bias = nullptr, ActType act = ActType::None);
optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
ActType getAct() const { return act; }

3
include/operators/element_wise.h
View File

@ -20,7 +20,8 @@ class ElementWiseObj : public OperatorObj {
class MSELossObj : public OperatorObj {
public:
enum Reduction { None = 0, Sum, Mean };
MSELossObj(GraphObj *graph, Tensor input0, Tensor input1, Reduction reduction, Tensor output);
MSELossObj(GraphObj *graph, Tensor input0, Tensor input1,
Reduction reduction, Tensor output);
optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
Reduction getReduction() const { return reductionMode; }

75
include/operators/unary.h
View File

@ -28,7 +28,7 @@ class ClipObj : public OperatorObj {
int numOutputs() const override { return 1; }
private:
float minValue,maxValue;
float minValue, maxValue;
vector<int> getWorkloadVector() const override;
vector<int> getOpAttrVector() const override;
};
@ -65,7 +65,8 @@ class L2LossObj : public OperatorObj {
class TransformObj : public OperatorObj {
public:
TransformObj(GraphObj *graph, Tensor input, Tensor output, float alpha, float beta);
TransformObj(GraphObj *graph, Tensor input, Tensor output, float alpha,
float beta);
optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
std::string toString() const override;
@ -82,18 +83,52 @@ class TransformObj : public OperatorObj {
class CastObj : public OperatorObj {
public:
enum CastType { Float2Half = 0, Float2HalfIEEE754, Float2Double, Float2Int64, Float2Int32, Float2Int16, Float2Int8, Float2Bool,
Half2Float, Half2Int32, Half2Int64, Half2Int16, Half2Int8, Half2Uint8, Half2Bool, Half2FloatInf,
Int322Float, Int322Half, Int322Int8, Int322Int16,
Int162Float, Int162Half, Int162Int32,
Int82Float, Int82Half, Int82Int16, Int82Int32,
Uint82Float, Uint82Half, Uint82Int32, Uint82Int64,
Bool2Float, Bool2Half, Bool2Int32,
Int322Int64, Int322Bool,
Int642Int32, Int642Uint32, Int642Float, Int642Half,
Uint642Uint32,
Uint322Int64, Uint322Uint64,
Double2Float};
enum CastType {
Float2Half = 0,
Float2HalfIEEE754,
Float2Double,
Float2Int64,
Float2Int32,
Float2Int16,
Float2Int8,
Float2Bool,
Half2Float,
Half2Int32,
Half2Int64,
Half2Int16,
Half2Int8,
Half2Uint8,
Half2Bool,
Half2FloatInf,
Int322Float,
Int322Half,
Int322Int8,
Int322Int16,
Int162Float,
Int162Half,
Int162Int32,
Int82Float,
Int82Half,
Int82Int16,
Int82Int32,
Uint82Float,
Uint82Half,
Uint82Int32,
Uint82Int64,
Bool2Float,
Bool2Half,
Bool2Int32,
Int322Int64,
Int322Bool,
Int642Int32,
Int642Uint32,
Int642Float,
Int642Half,
Uint642Uint32,
Uint322Int64,
Uint322Uint64,
Double2Float
};
CastObj(GraphObj *graph, Tensor input, Tensor output, CastType type);
optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
@ -110,7 +145,8 @@ class CastObj : public OperatorObj {
class CumsumObj : public OperatorObj {
public:
CumsumObj(GraphObj *graph, Tensor input, Tensor output, int axis, bool exclusive, bool reverse);
CumsumObj(GraphObj *graph, Tensor input, Tensor output, int axis,
bool exclusive, bool reverse);
optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
std::string toString() const override;
@ -129,16 +165,17 @@ class CumsumObj : public OperatorObj {
// class CumprodObj : public OperatorObj {
// public:
// CumprodObj(GraphObj *graph, Tensor input, Tensor output, int axis, bool exclusive, bool reverse);
// optional<vector<Shape>> inferShape(const TensorVec &inputs) const override;
//
// CumprodObj(GraphObj *graph, Tensor input, Tensor output, int axis, bool
// exclusive, bool reverse); optional<vector<Shape>> inferShape(const
// TensorVec &inputs) const override;
//
// std::string toString() const override;
// int getAxis() const { return axisValue; }
// float getExclusive() const { return exclusiveValue; }
// float getReverse() const { return reverseValue; }
// int numInputs() const override { return 1; }
// int numOutputs() const override { return 1; }
//
//
// private:
// int axisValue;
// bool exclusiveValue, reverseValue;

7
src/core/operator.cc
View File

@ -10,9 +10,7 @@ OperatorObj::OperatorObj(OpType opType, TensorVec inputs, TensorVec outputs)
IT_ASSERT(t != nullptr);
}
OperatorObj::OperatorObj(OpType opType)
: type(opType){
}
OperatorObj::OperatorObj(OpType opType) : type(opType) {}
bool OperatorObj::isLinearOp() const {
return enum_to_underlying(type) >= 100 && enum_to_underlying(type) < 200;
@ -91,7 +89,8 @@ bool OperatorObj::checkValid(GraphObj *graph, DataType type) {
if (shapes.size() != outputs.size())
return false;
if (graph) { // if graph != nullptr, outputs should be created
auto dataTypes = vector(numOutputs(), type);;
auto dataTypes = vector(numOutputs(), type);
;
for (size_t i = 0; i < outputs.size(); i++) {
IT_ASSERT(!outputs[i]);
outputs[i] = graph->addTensor(shapes[i], dataTypes[i]);

22
src/kernels/bang/activation_backward.cc
View File

@ -1,6 +1,6 @@
#include "operators/activation_backward.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/activation_backward.h"
namespace infini {
class ActivationBackwardCnnl : public BangKernelWithoutConfig {
@ -47,11 +47,9 @@ class ActivationBackwardCnnl : public BangKernelWithoutConfig {
opDesc, getOpType(), CNNL_NOT_PROPAGATE_NAN, getCoef()));
auto [alpha, beta] = getAlphBeta();
cnnlStatus_t stat =
cnnlActivationBackward(context->cnnlHandle(), opDesc, &alpha, yDesc, yData,
diffYDesc, diffYData,
xDesc, xData,
&beta, diffXDesc, diffXData);
cnnlStatus_t stat = cnnlActivationBackward(
context->cnnlHandle(), opDesc, &alpha, yDesc, yData, diffYDesc,
diffYData, xDesc, xData, &beta, diffXDesc, diffXData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -86,11 +84,11 @@ class TanhBackwardCnnl : public ActivationBackwardCnnl {
float getCoef() const override { return 0.0; }
};
REGISTER_KERNEL(Device::BANG, OpType::ReluBackward, DataType::Float32, ReluBackwardCnnl,
"ReluBackward_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::SigmoidBackward, DataType::Float32, SigmoidBackwardCnnl,
"SigmoidBackward_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::TanhBackward, DataType::Float32, TanhBackwardCnnl,
"TanhBackward_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::ReluBackward, DataType::Float32,
ReluBackwardCnnl, "ReluBackward_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::SigmoidBackward, DataType::Float32,
SigmoidBackwardCnnl, "SigmoidBackward_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::TanhBackward, DataType::Float32,
TanhBackwardCnnl, "TanhBackward_cnnl_BANG_Float32");
}; // namespace infini

22
src/kernels/bang/addn.cc
View File

@ -1,6 +1,6 @@
#include "operators/element_wise.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/element_wise.h"
namespace infini {
class AddNCnnl : public BangKernelWithoutConfig {
@ -10,8 +10,8 @@ class AddNCnnl : public BangKernelWithoutConfig {
auto context = dynamic_cast<const BangRuntimeObj *>(_context);
int num = op->numInputs();
void *argv[num];
for(int i = 0; i < num; ++i) {
argv[i] = op->getInputs(i)->getRawDataPtr<void *>();
for (int i = 0; i < num; ++i) {
argv[i] = op->getInputs(i)->getRawDataPtr<void *>();
}
void *const cData = (op->getOutput()->getRawDataPtr<void *>());
@ -24,20 +24,22 @@ class AddNCnnl : public BangKernelWithoutConfig {
checkCnnlError(cnnlCreateTensorDescriptor(&desc));
checkCnnlError(cnnlSetTensorDescriptor(desc, CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlTensorDescriptor_t descArray[num];
for(int i = 0; i < num; ++i) {
checkCnnlError(cnnlCreateTensorDescriptor(&descArray[i]));
checkCnnlError(cnnlSetTensorDescriptor(descArray[i], CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlTensorDescriptor_t descArray[num];
for (int i = 0; i < num; ++i) {
checkCnnlError(cnnlCreateTensorDescriptor(&descArray[i]));
checkCnnlError(
cnnlSetTensorDescriptor(descArray[i], CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
}
cnnlStatus_t stat = cnnlAddN(context->cnnlHandle(), descArray, argv, num, desc, cData);
cnnlStatus_t stat =
cnnlAddN(context->cnnlHandle(), descArray, argv, num, desc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
// Destories in BANG does not require sync. But cnnl does not state
// whether sync is required before destories.
for(int i = 0; i < num; ++i) {
for (int i = 0; i < num; ++i) {
checkCnnlError(cnnlDestroyTensorDescriptor(descArray[i]));
}
checkCnnlError(cnnlDestroyTensorDescriptor(desc));

156
src/kernels/bang/cast.cc
View File

@ -23,83 +23,87 @@ class CastCnnl : public BangKernelWithoutConfig {
checkCnnlError(cnnlCreateTensorDescriptor(&cDesc));
cnnlCastDataType_t NlCastType;
CastObj::CastType type = op->getType();
switch(type){
case CastObj::Float2Half:
checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, dim_array));
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_HALF, 4, dim_array));
NlCastType = CNNL_CAST_FLOAT_TO_HALF;
break;
case CastObj::Float2HalfIEEE754:
case CastObj::Float2Double:
case CastObj::Float2Int64:
case CastObj::Float2Int32:
checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, dim_array));
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_INT32, 4, dim_array));
NlCastType = CNNL_CAST_FLOAT_TO_INT32;
case CastObj::Float2Int16:
case CastObj::Float2Int8:
case CastObj::Float2Bool:
//Todo
break;
case CastObj::Half2Float:
case CastObj::Half2Int32:
case CastObj::Half2Int64:
case CastObj::Half2Int16:
case CastObj::Half2Int8:
case CastObj::Half2Uint8:
case CastObj::Half2Bool:
case CastObj::Half2FloatInf:
//todo
break;
case CastObj::Int322Float:
case CastObj::Int322Half:
case CastObj::Int322Int8:
case CastObj::Int322Int16:
//todo
break;
case CastObj::Int162Float:
case CastObj::Int162Half:
case CastObj::Int162Int32:
//todo
break;
case CastObj::Int82Float:
case CastObj::Int82Half:
case CastObj::Int82Int16:
case CastObj::Int82Int32:
//todo
break;
case CastObj::Uint82Float:
case CastObj::Uint82Half:
case CastObj::Uint82Int32:
case CastObj::Uint82Int64:
//todo
break;
case CastObj::Bool2Float:
case CastObj::Bool2Half:
case CastObj::Bool2Int32:
//todo
break;
case CastObj::Int322Int64:
case CastObj::Int322Bool:
//todo
break;
case CastObj::Int642Int32:
case CastObj::Int642Uint32:
case CastObj::Int642Float:
case CastObj::Int642Half:
//todo
break;
case CastObj::Uint642Uint32:
case CastObj::Uint322Int64:
case CastObj::Uint322Uint64:
//todo
break;
case CastObj::Double2Float:
//todo
break;
switch (type) {
case CastObj::Float2Half:
checkCnnlError(cnnlSetTensorDescriptor(
aDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, dim_array));
checkCnnlError(cnnlSetTensorDescriptor(
cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_HALF, 4, dim_array));
NlCastType = CNNL_CAST_FLOAT_TO_HALF;
break;
case CastObj::Float2HalfIEEE754:
case CastObj::Float2Double:
case CastObj::Float2Int64:
case CastObj::Float2Int32:
checkCnnlError(cnnlSetTensorDescriptor(
aDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, dim_array));
checkCnnlError(cnnlSetTensorDescriptor(
cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_INT32, 4, dim_array));
NlCastType = CNNL_CAST_FLOAT_TO_INT32;
case CastObj::Float2Int16:
case CastObj::Float2Int8:
case CastObj::Float2Bool:
// Todo
break;
case CastObj::Half2Float:
case CastObj::Half2Int32:
case CastObj::Half2Int64:
case CastObj::Half2Int16:
case CastObj::Half2Int8:
case CastObj::Half2Uint8:
case CastObj::Half2Bool:
case CastObj::Half2FloatInf:
// todo
break;
case CastObj::Int322Float:
case CastObj::Int322Half:
case CastObj::Int322Int8:
case CastObj::Int322Int16:
// todo
break;
case CastObj::Int162Float:
case CastObj::Int162Half:
case CastObj::Int162Int32:
// todo
break;
case CastObj::Int82Float:
case CastObj::Int82Half:
case CastObj::Int82Int16:
case CastObj::Int82Int32:
// todo
break;
case CastObj::Uint82Float:
case CastObj::Uint82Half:
case CastObj::Uint82Int32:
case CastObj::Uint82Int64:
// todo
break;
case CastObj::Bool2Float:
case CastObj::Bool2Half:
case CastObj::Bool2Int32:
// todo
break;
case CastObj::Int322Int64:
case CastObj::Int322Bool:
// todo
break;
case CastObj::Int642Int32:
case CastObj::Int642Uint32:
case CastObj::Int642Float:
case CastObj::Int642Half:
// todo
break;
case CastObj::Uint642Uint32:
case CastObj::Uint322Int64:
case CastObj::Uint322Uint64:
// todo
break;
case CastObj::Double2Float:
// todo
break;
}
cnnlStatus_t stat =
cnnlCastDataType(context->cnnlHandle(), aDesc, aData, NlCastType, cDesc, cData);
cnnlStatus_t stat = cnnlCastDataType(context->cnnlHandle(), aDesc,
aData, NlCastType, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

71
src/kernels/bang/convbpfilter.cc
View File

@ -1,6 +1,6 @@
#include "operators/conv.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/conv.h"
namespace infini {
class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
@ -27,7 +27,8 @@ class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
void *const bData = (op->getInputs(1)->getRawDataPtr<void *>());
void *const cData = (op->getOutput()->getRawDataPtr<void *>());
cnnlTensorDescriptor_t aDesc, bDesc, cDesc, aDescTrans, bDescTrans, cDescTrans;
cnnlTensorDescriptor_t aDesc, bDesc, cDesc, aDescTrans, bDescTrans,
cDescTrans;
auto dimInputs0 = op->getInputs(0)->getDims();
auto dimInputs1 = op->getInputs(1)->getDims();
auto dimOutput = op->getOutput()->getDims();
@ -47,17 +48,16 @@ class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
dimOutput[3]};
int inputs0ArrayTrans[4] = {dimInputs0[0], dimInputs0[2], dimInputs0[3],
dimInputs0[1]};
dimInputs0[1]};
int inputs1ArrayTrans[4] = {dimInputs1[0], dimInputs1[2], dimInputs1[3],
dimInputs1[1]};
dimInputs1[1]};
int outputArrayTrans[4] = {dimOutput[0], dimOutput[2], dimOutput[3],
dimOutput[1]};
dimOutput[1]};
int transMode[4] = {0, 2, 3, 1};
cnnlTransposeDescriptor_t transDesc;
checkCnnlError(cnnlCreateTransposeDescriptor(&transDesc));
checkCnnlError(cnnlSetTransposeDescriptor(
transDesc, 4, transMode));
checkCnnlError(cnnlSetTransposeDescriptor(transDesc, 4, transMode));
// get inputs
checkCnnlError(cnnlCreateTensorDescriptor(&aDesc));
@ -65,13 +65,17 @@ class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
aDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, inputs0Array));
checkCnnlError(cnnlCreateTensorDescriptor(&aDescTrans));
checkCnnlError(cnnlSetTensorDescriptor(
aDescTrans, CNNL_LAYOUT_NHWC, CNNL_DTYPE_FLOAT, 4, inputs0ArrayTrans));
checkCnnlError(cnnlSetTensorDescriptor(aDescTrans, CNNL_LAYOUT_NHWC,
CNNL_DTYPE_FLOAT, 4,
inputs0ArrayTrans));
size_t wsTrans1Size = dimInputs0[0] * dimInputs0[1] * dimInputs0[2] * dimInputs0[3] * sizeof(float);
size_t wsTrans1Size = dimInputs0[0] * dimInputs0[1] * dimInputs0[2] *
dimInputs0[3] * sizeof(float);
BangPtr wsTrans1Data = context->getWorkspace(wsTrans1Size);
cnnlStatus_t stat = cnnlTranspose(context->cnnlHandle(), transDesc, aDesc, aData, aDescTrans, wsTrans1Data);
cnnlStatus_t stat =
cnnlTranspose(context->cnnlHandle(), transDesc, aDesc, aData,
aDescTrans, wsTrans1Data);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -80,13 +84,16 @@ class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
bDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, inputs1Array));
checkCnnlError(cnnlCreateTensorDescriptor(&bDescTrans));
checkCnnlError(cnnlSetTensorDescriptor(
bDescTrans, CNNL_LAYOUT_NHWC, CNNL_DTYPE_FLOAT, 4, inputs1ArrayTrans));
checkCnnlError(cnnlSetTensorDescriptor(bDescTrans, CNNL_LAYOUT_NHWC,
CNNL_DTYPE_FLOAT, 4,
inputs1ArrayTrans));
size_t wsTrans2Size = dimInputs1[0] * dimInputs1[1] * dimInputs1[2] * dimInputs1[3] * sizeof(float);
size_t wsTrans2Size = dimInputs1[0] * dimInputs1[1] * dimInputs1[2] *
dimInputs1[3] * sizeof(float);
BangPtr wsTrans2Data = context->getWorkspace(wsTrans2Size);
stat = cnnlTranspose(context->cnnlHandle(), transDesc, bDesc, bData, bDescTrans, wsTrans2Data);
stat = cnnlTranspose(context->cnnlHandle(), transDesc, bDesc, bData,
bDescTrans, wsTrans2Data);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -96,36 +103,40 @@ class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, outputArray));
checkCnnlError(cnnlCreateTensorDescriptor(&cDescTrans));
checkCnnlError(cnnlSetTensorDescriptor(
cDescTrans, CNNL_LAYOUT_NHWC, CNNL_DTYPE_FLOAT, 4, outputArrayTrans));
checkCnnlError(cnnlSetTensorDescriptor(cDescTrans, CNNL_LAYOUT_NHWC,
CNNL_DTYPE_FLOAT, 4,
outputArrayTrans));
size_t wsTrans3Size = dimOutput[0] * dimOutput[1] * dimOutput[2] * dimOutput[3] * sizeof(float);
size_t wsTrans3Size = dimOutput[0] * dimOutput[1] * dimOutput[2] *
dimOutput[3] * sizeof(float);
BangPtr wsTrans3Data = context->getWorkspace(wsTrans3Size);
cnnlConvolutionBwdFilterAlgo_t algo;
cnnlGetConvolutionBackwardFilterAlgorithm(context->cnnlHandle(), convDesc,
aDescTrans, bDescTrans, cDescTrans,
CNNL_CONVOLUTION_BWD_FILTER_FASTEST, &algo);
cnnlGetConvolutionBackwardFilterAlgorithm(
context->cnnlHandle(), convDesc, aDescTrans, bDescTrans, cDescTrans,
CNNL_CONVOLUTION_BWD_FILTER_FASTEST, &algo);
size_t wsSize;
cnnlGetConvolutionBackwardFilterWorkspaceSize(context->cnnlHandle(), aDescTrans,
bDescTrans, cDescTrans, convDesc,
algo, &wsSize);
cnnlGetConvolutionBackwardFilterWorkspaceSize(
context->cnnlHandle(), aDescTrans, bDescTrans, cDescTrans, convDesc,
algo, &wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
stat = cnnlConvolutionBackwardFilter(
context->cnnlHandle(), NULL, aDescTrans, wsTrans1Data, bDescTrans,
wsTrans2Data, convDesc, algo, wsData, wsSize, NULL, cDescTrans, wsTrans3Data);
wsTrans2Data, convDesc, algo, wsData, wsSize, NULL, cDescTrans,
wsTrans3Data);
if (stat != CNNL_STATUS_SUCCESS)
return;
int transMode2[4] = {0, 3, 1, 2};
cnnlTransposeDescriptor_t transOutputDesc;
checkCnnlError(cnnlCreateTransposeDescriptor(&transOutputDesc));
checkCnnlError(cnnlSetTransposeDescriptor(
transOutputDesc, 4, transMode2));
checkCnnlError(
cnnlSetTransposeDescriptor(transOutputDesc, 4, transMode2));
stat = cnnlTranspose(context->cnnlHandle(), transOutputDesc, cDescTrans, wsTrans3Data, cDesc, cData);
stat = cnnlTranspose(context->cnnlHandle(), transOutputDesc, cDescTrans,
wsTrans3Data, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -143,6 +154,6 @@ class ConvBackwardFilterCnnl : public BangKernelWithoutConfig {
}
};
REGISTER_KERNEL(Device::BANG, OpType::ConvBackwardFilter, DataType::Float32, ConvBackwardFilterCnnl,
"ConvBackwardFilter_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::ConvBackwardFilter, DataType::Float32,
ConvBackwardFilterCnnl, "ConvBackwardFilter_cnnl_BANG_Float32");
}; // namespace infini

3
src/kernels/bang/cumsum.cc
View File

@ -32,7 +32,8 @@ class CumsumCnnl : public BangKernelWithoutConfig {
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlStatus_t stat =
cnnlCumsum(context->cnnlHandle(), aDesc, aData, axis, exclusive, reverse, CNNL_NOT_PROPAGATE_NAN, cDesc, cData);
cnnlCumsum(context->cnnlHandle(), aDesc, aData, axis, exclusive,
reverse, CNNL_NOT_PROPAGATE_NAN, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

15
src/kernels/bang/det.cc
View File

@ -1,6 +1,6 @@
#include "operators/det.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/det.h"
namespace infini {
class DetCnnl : public BangKernelWithoutConfig {
@ -13,7 +13,7 @@ class DetCnnl : public BangKernelWithoutConfig {
void *const cData = (op->getOutput()->getRawDataPtr<void *>());
DetObj::Mode mode = op->getMode();
cnnlDetMode_t nlMode;
if(mode == DetObj::LogDet) {
if (mode == DetObj::LogDet) {
nlMode = CNNL_DET_MODE_LOGDET;
} else {
nlMode = CNNL_DET_MODE_DET;
@ -28,15 +28,16 @@ class DetCnnl : public BangKernelWithoutConfig {
int dimout_array[2] = {dimout[0], dimout[1]};
// get inputs
checkCnnlError(cnnlCreateTensorDescriptor(&aDesc));
checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_ARRAY,
CNNL_DTYPE_FLOAT, 4, dimin_array));
checkCnnlError(cnnlSetTensorDescriptor(
aDesc, CNNL_LAYOUT_ARRAY, CNNL_DTYPE_FLOAT, 4, dimin_array));
// get outputs
checkCnnlError(cnnlCreateTensorDescriptor(&cDesc));
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_ARRAY,
CNNL_DTYPE_FLOAT, 2, dimout_array));
checkCnnlError(cnnlSetTensorDescriptor(
cDesc, CNNL_LAYOUT_ARRAY, CNNL_DTYPE_FLOAT, 2, dimout_array));
cnnlStatus_t stat = cnnlDet(context->cnnlHandle(), nlMode, aDesc, aData, cDesc, cData);
cnnlStatus_t stat =
cnnlDet(context->cnnlHandle(), nlMode, aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

123
src/kernels/bang/element_wise.cc
View File

@ -98,13 +98,13 @@ class DivCnnl : public BangKernelWithoutConfig {
size_t wsSize;
cnnlGetDivWorkspaceSize(context->cnnlHandle(), aDesc, bDesc, cDesc,
&wsSize);
&wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlDiv_v2(context->cnnlHandle(),
CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, bDesc, bData, wsData, wsSize, cDesc, cData);
cnnlStatus_t stat = cnnlDiv_v2(
context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc,
aData, bDesc, bData, wsData, wsSize, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -152,9 +152,9 @@ class DivNoNanCnnl : public BangKernelWithoutConfig {
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlDivNoNan_v2(context->cnnlHandle(),
CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, bDesc, bData, wsData, wsSize, cDesc, cData);
cnnlStatus_t stat = cnnlDivNoNan_v2(
context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc,
aData, bDesc, bData, wsData, wsSize, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -201,8 +201,9 @@ class MaximumCnnl : public BangKernelWithoutConfig {
cnnlGetMaximumWorkspaceSize(context->cnnlHandle(), cDesc, &wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlMaximum(context->cnnlHandle(), aDesc, aData, bDesc, bData,
cDesc, cData, wsData, wsSize);
cnnlStatus_t stat =
cnnlMaximum(context->cnnlHandle(), aDesc, aData, bDesc, bData,
cDesc, cData, wsData, wsSize);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -249,8 +250,9 @@ class MinimumCnnl : public BangKernelWithoutConfig {
cnnlGetMinimumWorkspaceSize(context->cnnlHandle(), cDesc, &wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlMinimum(context->cnnlHandle(), aDesc, aData, bDesc, bData,
cDesc, cData, wsData, wsSize);
cnnlStatus_t stat =
cnnlMinimum(context->cnnlHandle(), aDesc, aData, bDesc, bData,
cDesc, cData, wsData, wsSize);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -278,7 +280,7 @@ class MSELossCnnl : public BangKernelWithoutConfig {
IT_TODO_HALT();
int dim_array[4] = {dim[0], dim[1], dim[2], dim[3]};
int dim_out[4] ={1,1,1,1};
int dim_out[4] = {1, 1, 1, 1};
// get inputs
checkCnnlError(cnnlCreateTensorDescriptor(&aDesc));
checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_NCHW,
@ -290,23 +292,23 @@ class MSELossCnnl : public BangKernelWithoutConfig {
// get outputs
checkCnnlError(cnnlCreateTensorDescriptor(&cDesc));
if ( reduction == MSELossObj::None ) {
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
if (reduction == MSELossObj::None) {
checkCnnlError(cnnlSetTensorDescriptor(
cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, dim_array));
} else {
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_out));
checkCnnlError(cnnlSetTensorDescriptor(
cDesc, CNNL_LAYOUT_NCHW, CNNL_DTYPE_FLOAT, 4, dim_out));
}
cnnlStatus_t stat;
if( reduction == MSELossObj::None ) {
stat = cnnlMSELoss(context->cnnlHandle(), CNNL_MSE_LOSS_NONE, aDesc, aData, bDesc, bData,
cDesc, cData);
if (reduction == MSELossObj::None) {
stat = cnnlMSELoss(context->cnnlHandle(), CNNL_MSE_LOSS_NONE, aDesc,
aData, bDesc, bData, cDesc, cData);
} else if (reduction == MSELossObj::Sum) {
stat = cnnlMSELoss(context->cnnlHandle(), CNNL_MSE_LOSS_SUM, aDesc, aData, bDesc, bData,
cDesc, cData);
stat = cnnlMSELoss(context->cnnlHandle(), CNNL_MSE_LOSS_SUM, aDesc,
aData, bDesc, bData, cDesc, cData);
} else {
stat = cnnlMSELoss(context->cnnlHandle(), CNNL_MSE_LOSS_MEAN, aDesc, aData, bDesc, bData,
cDesc, cData);
stat = cnnlMSELoss(context->cnnlHandle(), CNNL_MSE_LOSS_MEAN, aDesc,
aData, bDesc, bData, cDesc, cData);
}
if (stat != CNNL_STATUS_SUCCESS)
@ -352,11 +354,13 @@ class PowerCnnl : public BangKernelWithoutConfig {
// get op descriptor
size_t wsSize;
cnnlGetPowWorkspaceSize(context->cnnlHandle(), aDesc, bDesc, cDesc, &wsSize);
cnnlGetPowWorkspaceSize(context->cnnlHandle(), aDesc, bDesc, cDesc,
&wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlPow(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, bDesc, bData, wsData, wsSize, cDesc, cData);
cnnlStatus_t stat =
cnnlPow(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, bDesc, bData, wsData, wsSize, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -404,9 +408,9 @@ class FloorDivCnnl : public BangKernelWithoutConfig {
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlFloorDiv_v2(context->cnnlHandle(),
CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, bDesc, bData, cDesc, cData, wsData, wsSize);
cnnlStatus_t stat = cnnlFloorDiv_v2(
context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc,
aData, bDesc, bData, cDesc, cData, wsData, wsSize);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -449,14 +453,14 @@ class FloorDivTruncCnnl : public BangKernelWithoutConfig {
CNNL_DTYPE_FLOAT, 4, dim_array));
size_t wsSize;
cnnlGetFloorDivTruncWorkspaceSize(context->cnnlHandle(), aDesc, bDesc, cDesc,
&wsSize);
cnnlGetFloorDivTruncWorkspaceSize(context->cnnlHandle(), aDesc, bDesc,
cDesc, &wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlFloorDivTrunc(context->cnnlHandle(),
CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, bDesc, bData, cDesc, cData, wsData, wsSize);
cnnlStatus_t stat = cnnlFloorDivTrunc(
context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc,
aData, bDesc, bData, cDesc, cData, wsData, wsSize);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -504,8 +508,9 @@ class FloorModCnnl : public BangKernelWithoutConfig {
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlFloorMod(context->cnnlHandle(),
aDesc, aData, bDesc, bData, cDesc, cData, wsData, wsSize);
cnnlStatus_t stat =
cnnlFloorMod(context->cnnlHandle(), aDesc, aData, bDesc, bData,
cDesc, cData, wsData, wsSize);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -522,42 +527,47 @@ class FloorModCnnl : public BangKernelWithoutConfig {
// const RuntimeObj *_context) const override {
// auto op = as<ElementWiseObj>(_op);
// auto context = dynamic_cast<const BangRuntimeObj *>(_context);
//
//
// void *const aData = (op->getInputs(0)->getRawDataPtr<void *>());
// void *const bData = (op->getInputs(1)->getRawDataPtr<void *>());
// void *const cData = (op->getOutput()->getRawDataPtr<void *>());
//
//
// cnnlTensorDescriptor_t aDesc, bDesc, cDesc;
// auto dim = op->getInputs(0)->getDims();
// if (dim.size() != 4)
// IT_TODO_HALT();
//
//
// int dim_array[4] = {dim[0], dim[1], dim[2], dim[3]};
// // get inputs
// checkCnnlError(cnnlCreateTensorDescriptor(&aDesc));
// checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_NCHW,
// CNNL_DTYPE_FLOAT, 4, dim_array));
//
// CNNL_DTYPE_FLOAT, 4,
// dim_array));
//
// checkCnnlError(cnnlCreateTensorDescriptor(&bDesc));
// checkCnnlError(cnnlSetTensorDescriptor(bDesc, CNNL_LAYOUT_NCHW,
// CNNL_DTYPE_FLOAT, 4, dim_array));
//
// CNNL_DTYPE_FLOAT, 4,
// dim_array));
//
// // get outputs
// checkCnnlError(cnnlCreateTensorDescriptor(&cDesc));
// checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_NCHW,
// CNNL_DTYPE_FLOAT, 4, dim_array));
//
// CNNL_DTYPE_FLOAT, 4,
// dim_array));
//
// size_t wsSize;
// cnnlGetFloorModTruncWorkspaceSize(context->cnnlHandle(), aDesc, bDesc, cDesc,
// cnnlGetFloorModTruncWorkspaceSize(context->cnnlHandle(), aDesc,
// bDesc, cDesc,
// &wsSize);
//
//
// BangPtr wsData = context->getWorkspace(wsSize);
//
//
// cnnlStatus_t stat = cnnlFloorModTrunc(context->cnnlHandle(),
// aDesc, aData, bDesc, bData, cDesc, cData, wsData, wsSize);
// aDesc, aData, bDesc, bData, cDesc,
// cData, wsData, wsSize);
// if (stat != CNNL_STATUS_SUCCESS)
// return;
//
//
// // Destories in BANG does not require sync. But cnnl does not state
// // whether sync is required before destories.
// checkCnnlError(cnnlDestroyTensorDescriptor(aDesc));
@ -595,8 +605,8 @@ REGISTER_KERNEL(Device::BANG, OpType::Sub, DataType::Float32, SubCnnl,
REGISTER_KERNEL(Device::BANG, OpType::Mul, DataType::Float32, MulCnnl,
"Mul_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::DivDemo, DataType::Float32, ElementWiseBang,
"DivDemo_Bang_Float32");
REGISTER_KERNEL(Device::BANG, OpType::DivDemo, DataType::Float32,
ElementWiseBang, "DivDemo_Bang_Float32");
REGISTER_KERNEL(Device::BANG, OpType::Div, DataType::Float32, DivCnnl,
"Div_cnnl_Float32");
REGISTER_KERNEL(Device::BANG, OpType::DivNoNan, DataType::Float32, DivNoNanCnnl,
@ -611,11 +621,12 @@ REGISTER_KERNEL(Device::BANG, OpType::Power, DataType::Float32, PowerCnnl,
"Power_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::FloorDiv, DataType::Float32, FloorDivCnnl,
"FloorDiv_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::FloorDivTrunc, DataType::Float32, FloorDivTruncCnnl,
"FloorDivTrunc_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::FloorDivTrunc, DataType::Float32,
FloorDivTruncCnnl, "FloorDivTrunc_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::FloorMod, DataType::Float32, FloorModCnnl,
"FloorMod_cnnl_BANG_Float32");
// REGISTER_KERNEL(Device::BANG, OpType::FloorModTrunc, DataType::Float32, FloorModTruncCnnl,
// REGISTER_KERNEL(Device::BANG, OpType::FloorModTrunc, DataType::Float32,
// FloorModTruncCnnl,
// "FloorModTrunc_cnnl_BANG_Float32");
// REGISTER_KERNEL(Device::BANG, OpType::Pow, DataType::Float32,
// ElementWiseBang,

3
src/kernels/bang/erf.cc
View File

@ -29,7 +29,8 @@ class ErfCnnl : public BangKernelWithoutConfig {
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlStatus_t stat =
cnnlErf_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc, aData, cDesc, cData);
cnnlErf_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

3
src/kernels/bang/exp.cc
View File

@ -29,7 +29,8 @@ class ExpCnnl : public BangKernelWithoutConfig {
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlStatus_t stat =
cnnlExp_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc, aData, cDesc, cData);
cnnlExp_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

10
src/kernels/bang/log.cc
View File

@ -31,7 +31,7 @@ class LogCnnl : public BangKernelWithoutConfig {
cnnlStatus_t stat =
cnnlLog_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
getOpType(), aDesc, aData, cDesc, cData);
getOpType(), aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -44,14 +44,14 @@ class LogCnnl : public BangKernelWithoutConfig {
class LogECnnl : public LogCnnl {
cnnlLogBase_t getOpType() const override { return CNNL_LOG_E; }
};
};
class Log2Cnnl : public LogCnnl {
cnnlLogBase_t getOpType() const override { return CNNL_LOG_2; }
};
};
class Log10Cnnl : public LogCnnl {
cnnlLogBase_t getOpType() const override { return CNNL_LOG_10; }
};
};
REGISTER_KERNEL(Device::BANG, OpType::Log_e, DataType::Float32, LogECnnl,
"Loge_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::Log_2, DataType::Float32, Log2Cnnl,

2
src/kernels/bang/log1p.cc
View File

@ -40,7 +40,7 @@ class Log1pCnnl : public BangKernelWithoutConfig {
checkCnnlError(cnnlDestroyTensorDescriptor(cDesc));
}
};
REGISTER_KERNEL(Device::BANG, OpType::Log1p, DataType::Float32, Log1pCnnl,
"Log1p_cnnl_BANG_Float32");

22
src/kernels/bang/muln.cc
View File

@ -1,6 +1,6 @@
#include "operators/element_wise.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/element_wise.h"
namespace infini {
class MulNCnnl : public BangKernelWithoutConfig {
@ -10,8 +10,8 @@ class MulNCnnl : public BangKernelWithoutConfig {
auto context = dynamic_cast<const BangRuntimeObj *>(_context);
int num = op->numInputs();
void *argv[num];
for(int i = 0; i < num; ++i) {
argv[i] = op->getInputs(i)->getRawDataPtr<void *>();
for (int i = 0; i < num; ++i) {
argv[i] = op->getInputs(i)->getRawDataPtr<void *>();
}
void *const cData = (op->getOutput()->getRawDataPtr<void *>());
@ -24,20 +24,22 @@ class MulNCnnl : public BangKernelWithoutConfig {
checkCnnlError(cnnlCreateTensorDescriptor(&desc));
checkCnnlError(cnnlSetTensorDescriptor(desc, CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlTensorDescriptor_t descArray[num];
for(int i = 0; i < num; ++i) {
checkCnnlError(cnnlCreateTensorDescriptor(&descArray[i]));
checkCnnlError(cnnlSetTensorDescriptor(descArray[i], CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlTensorDescriptor_t descArray[num];
for (int i = 0; i < num; ++i) {
checkCnnlError(cnnlCreateTensorDescriptor(&descArray[i]));
checkCnnlError(
cnnlSetTensorDescriptor(descArray[i], CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
}
cnnlStatus_t stat = cnnlMulN(context->cnnlHandle(), descArray, argv, num, desc, cData);
cnnlStatus_t stat =
cnnlMulN(context->cnnlHandle(), descArray, argv, num, desc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
// Destories in BANG does not require sync. But cnnl does not state
// whether sync is required before destories.
for(int i = 0; i < num; ++i) {
for (int i = 0; i < num; ++i) {
checkCnnlError(cnnlDestroyTensorDescriptor(descArray[i]));
}
checkCnnlError(cnnlDestroyTensorDescriptor(desc));

4
src/kernels/bang/negtensor.cc
View File

@ -40,7 +40,7 @@ class NegTensorCnnl : public BangKernelWithoutConfig {
}
};
REGISTER_KERNEL(Device::BANG, OpType::NegTensor, DataType::Float32, NegTensorCnnl,
"NegTensor_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::NegTensor, DataType::Float32,
NegTensorCnnl, "NegTensor_cnnl_BANG_Float32");
}; // namespace infini

21
src/kernels/bang/pad.cc
View File

@ -1,6 +1,6 @@
#include "operators/pad.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/pad.h"
namespace infini {
class PadCnnl : public BangKernelWithoutConfig {
@ -24,30 +24,31 @@ class PadCnnl : public BangKernelWithoutConfig {
if (pads.size() == 2 && dim_size != 1) {
for (int i = 0; i < dim_size * 2; i += 2) {
paddings[i] = pads[0];
paddings[i+1] = pads[1];
paddings[i + 1] = pads[1];
}
} else {
for (int i = 0; i < dim_size * 2; i += 2) {
paddings[i] = pads[i/2];
paddings[i+1] = pads[i/2 + dim_size];
paddings[i] = pads[i / 2];
paddings[i + 1] = pads[i / 2 + dim_size];
}
}
int dimout_array[dim_size];
for (int i = 0; i < dim_size; ++i) {
dimout_array[i] = dim[i] + paddings[2*i] + paddings[2*i+1];
dimout_array[i] = dim[i] + paddings[2 * i] + paddings[2 * i + 1];
}
float paddingValue = 0.0;
// input
checkCnnlError(cnnlCreateTensorDescriptor(&aDesc));
checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_ARRAY,
CNNL_DTYPE_FLOAT, dim_size, dim_array));
checkCnnlError(cnnlSetTensorDescriptor(
aDesc, CNNL_LAYOUT_ARRAY, CNNL_DTYPE_FLOAT, dim_size, dim_array));
// output
checkCnnlError(cnnlCreateTensorDescriptor(&cDesc));
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_ARRAY,
CNNL_DTYPE_FLOAT, dim_size, dimout_array));
CNNL_DTYPE_FLOAT, dim_size,
dimout_array));
cnnlStatus_t stat =
cnnlPad(context->cnnlHandle(), aDesc, aData, paddings, &paddingValue, cDesc, cData);
cnnlStatus_t stat = cnnlPad(context->cnnlHandle(), aDesc, aData,
paddings, &paddingValue, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

4
src/kernels/bang/reciprocal.cc
View File

@ -40,7 +40,7 @@ class ReciprocalCnnl : public BangKernelWithoutConfig {
}
};
REGISTER_KERNEL(Device::BANG, OpType::Reciprocal, DataType::Float32, ReciprocalCnnl,
"Reciprocal_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::Reciprocal, DataType::Float32,
ReciprocalCnnl, "Reciprocal_cnnl_BANG_Float32");
}; // namespace infini

3
src/kernels/bang/rsqrt.cc
View File

@ -29,7 +29,8 @@ class RsqrtCnnl : public BangKernelWithoutConfig {
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlStatus_t stat =
cnnlRsqrt_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc, aData, cDesc, cData);
cnnlRsqrt_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

3
src/kernels/bang/sqrt.cc
View File

@ -29,7 +29,8 @@ class SqrtCnnl : public BangKernelWithoutConfig {
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlStatus_t stat =
cnnlSqrt_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION, aDesc, aData, cDesc, cData);
cnnlSqrt_v2(context->cnnlHandle(), CNNL_COMPUTATION_HIGH_PRECISION,
aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;

8
src/kernels/bang/transform.cc
View File

@ -25,8 +25,8 @@ class TransformCnnl : public BangKernelWithoutConfig {
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_NCHW,
CNNL_DTYPE_FLOAT, 4, dim_array));
cnnlStatus_t stat =
cnnlTransform(context->cnnlHandle(), &alpha, cDesc, aData, &beta, cData);
cnnlStatus_t stat = cnnlTransform(context->cnnlHandle(), &alpha, cDesc,
aData, &beta, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -36,7 +36,7 @@ class TransformCnnl : public BangKernelWithoutConfig {
}
};
REGISTER_KERNEL(Device::BANG, OpType::Transform, DataType::Float32, TransformCnnl,
"Transform_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::Transform, DataType::Float32,
TransformCnnl, "Transform_cnnl_BANG_Float32");
}; // namespace infini

23
src/kernels/bang/transpose.cc
View File

@ -1,6 +1,6 @@
#include "operators/transpose.h"
#include "bang/bang_kernel_without_config.h"
#include "bang/bang_runtime.h"
#include "operators/transpose.h"
namespace infini {
class TransposeCnnl : public BangKernelWithoutConfig {
@ -22,13 +22,13 @@ class TransposeCnnl : public BangKernelWithoutConfig {
int dimout_array[4] = {dimout[0], dimout[1], dimout[2], dimout[3]};
// get inputs
checkCnnlError(cnnlCreateTensorDescriptor(&aDesc));
checkCnnlError(cnnlSetTensorDescriptor(aDesc, CNNL_LAYOUT_ARRAY,
CNNL_DTYPE_FLOAT, 4, dimin_array));
checkCnnlError(cnnlSetTensorDescriptor(
aDesc, CNNL_LAYOUT_ARRAY, CNNL_DTYPE_FLOAT, 4, dimin_array));
// get outputs
checkCnnlError(cnnlCreateTensorDescriptor(&cDesc));
checkCnnlError(cnnlSetTensorDescriptor(cDesc, CNNL_LAYOUT_ARRAY,
CNNL_DTYPE_FLOAT, 4, dimout_array));
checkCnnlError(cnnlSetTensorDescriptor(
cDesc, CNNL_LAYOUT_ARRAY, CNNL_DTYPE_FLOAT, 4, dimout_array));
// get op descriptor
auto permute = op->getPermute();
@ -37,12 +37,13 @@ class TransposeCnnl : public BangKernelWithoutConfig {
checkCnnlError(cnnlSetTransposeDescriptor(opDesc, 4, permute));
size_t wsSize;
cnnlGetTransposeWorkspaceSize(context->cnnlHandle(), aDesc, opDesc, &wsSize);
cnnlGetTransposeWorkspaceSize(context->cnnlHandle(), aDesc, opDesc,
&wsSize);
BangPtr wsData = context->getWorkspace(wsSize);
cnnlStatus_t stat = cnnlTranspose_v2(context->cnnlHandle(), opDesc,
aDesc, aData, cDesc, cData,
wsData, wsSize);
cnnlStatus_t stat =
cnnlTranspose_v2(context->cnnlHandle(), opDesc, aDesc, aData, cDesc,
cData, wsData, wsSize);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -54,6 +55,6 @@ class TransposeCnnl : public BangKernelWithoutConfig {
}
};
REGISTER_KERNEL(Device::BANG, OpType::Transpose, DataType::Float32, TransposeCnnl,
"Transpose_cnnl_BANG_Float32");
REGISTER_KERNEL(Device::BANG, OpType::Transpose, DataType::Float32,
TransposeCnnl, "Transpose_cnnl_BANG_Float32");
}; // namespace infini

32
src/kernels/bang/trigon.cc
View File

@ -33,12 +33,10 @@ class TrigonCnnl : public BangKernelWithoutConfig {
// get op descriptor
cnnlTrigonDescriptor_t opDesc;
checkCnnlError(cnnlCreateTrigonDescriptor(&opDesc));
checkCnnlError(cnnlSetTrigonDescriptor(
opDesc, getOpType()));
checkCnnlError(cnnlSetTrigonDescriptor(opDesc, getOpType()));
cnnlStatus_t stat =
cnnlTrigonForward(context->cnnlHandle(), opDesc, aDesc,
aData, cDesc, cData);
cnnlStatus_t stat = cnnlTrigonForward(context->cnnlHandle(), opDesc,
aDesc, aData, cDesc, cData);
if (stat != CNNL_STATUS_SUCCESS)
return;
@ -55,7 +53,7 @@ class SinCnnl : public TrigonCnnl {
return CNNL_TRIGON_SIN;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -64,7 +62,7 @@ class CosCnnl : public TrigonCnnl {
return CNNL_TRIGON_COS;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -73,7 +71,7 @@ class TanCnnl : public TrigonCnnl {
return CNNL_TRIGON_TAN;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -82,7 +80,7 @@ class ASinCnnl : public TrigonCnnl {
return CNNL_TRIGON_ASIN;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -91,7 +89,7 @@ class ACosCnnl : public TrigonCnnl {
return CNNL_TRIGON_ACOS;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -100,7 +98,7 @@ class ATanCnnl : public TrigonCnnl {
return CNNL_TRIGON_ATAN;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -109,7 +107,7 @@ class SinHCnnl : public TrigonCnnl {
return CNNL_TRIGON_SINH;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -118,7 +116,7 @@ class CosHCnnl : public TrigonCnnl {
return CNNL_TRIGON_COSH;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -127,7 +125,7 @@ class TanHCnnl : public TrigonCnnl {
return CNNL_TRIGON_TANH;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -136,7 +134,7 @@ class ASinHCnnl : public TrigonCnnl {
return CNNL_TRIGON_ASINH;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -145,7 +143,7 @@ class ACosHCnnl : public TrigonCnnl {
return CNNL_TRIGON_ACOSH;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};
@ -154,7 +152,7 @@ class ATanHCnnl : public TrigonCnnl {
return CNNL_TRIGON_ATANH;
}
cnnlComputationPreference_t getPrefer() const override {
return CNNL_COMPUTATION_HIGH_PRECISION;
return CNNL_COMPUTATION_HIGH_PRECISION;
}
};

10
src/operators/activation_backward.cc
View File

@ -1,17 +1,15 @@
#include "operators/activation_backward.h"
namespace infini {
ActivationBackwardObj::ActivationBackwardObj(OpType type,
GraphObj *graph,
Tensor y,
Tensor diff_y,
Tensor x,
ActivationBackwardObj::ActivationBackwardObj(OpType type, GraphObj *graph,
Tensor y, Tensor diff_y, Tensor x,
Tensor diff_x)
: OperatorObj(type, {y, diff_y, x}, {diff_x}) {
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>> ActivationBackwardObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>>
ActivationBackwardObj::inferShape(const TensorVec &inputs) const {
const auto A = inputs[0];
return {{A->getDims()}};
}

25
src/operators/conv.cc
View File

@ -183,9 +183,9 @@ void ConvTransposed2dObj::setAuxilaryAttributes(PaddingMode mode) {
void ConvBackwardFilterObj::setAuxilaryAttributes(PaddingMode mode) {
const Tensor &inputX = inputs[0];
const Tensor &diffY = inputs[1];
n = inputX->getDims()[0], c = inputX->getDims()[1], h = inputX->getDims()[2],
w = inputX->getDims()[3], f = diffY->getDims()[0], r = diffY->getDims()[2],
s = diffY->getDims()[3];
n = inputX->getDims()[0], c = inputX->getDims()[1],
h = inputX->getDims()[2], w = inputX->getDims()[3], f = diffY->getDims()[0],
r = diffY->getDims()[2], s = diffY->getDims()[3];
if (mode == PaddingMode::Same) {
int oh = h / sh;
int ow = w / sw;
@ -196,9 +196,10 @@ void ConvBackwardFilterObj::setAuxilaryAttributes(PaddingMode mode) {
}
}
ConvBackwardFilterObj::ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Tensor diffY, Tensor diffW,
int ph, int pw, int sh, int sw, int dh, int dw, Tensor bias,
ActType act)
ConvBackwardFilterObj::ConvBackwardFilterObj(GraphObj *graph, Tensor inputX,
Tensor diffY, Tensor diffW, int ph,
int pw, int sh, int sw, int dh,
int dw, Tensor bias, ActType act)
: ConvBaseObj(OpType::Conv, {inputX, diffY}, diffW, ph, pw, sh, sw, dh, dw,
inputX, diffY),
act(act) {
@ -208,9 +209,11 @@ ConvBackwardFilterObj::ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Ten
IT_ASSERT(checkValid(graph));
}
ConvBackwardFilterObj::ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Tensor diffY, Tensor diffW,
PaddingMode mode, int sh, int sw, int dh, int dw, Tensor bias,
ActType act)
ConvBackwardFilterObj::ConvBackwardFilterObj(GraphObj *graph, Tensor inputX,
Tensor diffY, Tensor diffW,
PaddingMode mode, int sh, int sw,
int dh, int dw, Tensor bias,
ActType act)
: ConvBaseObj(OpType::Conv, {inputX, diffY}, diffW, mode, sh, sw, dh, dw,
inputX, diffY),
act(act) {
@ -220,7 +223,8 @@ ConvBackwardFilterObj::ConvBackwardFilterObj(GraphObj *graph, Tensor inputX, Ten
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>> ConvBackwardFilterObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>>
ConvBackwardFilterObj::inferShape(const TensorVec &inputs) const {
const auto &inputX = inputs[0], &diffY = inputs[1];
auto n = inputX->getDims()[0];
auto h = inputX->getDims()[2];
@ -251,5 +255,4 @@ optional<vector<Shape>> ConvBackwardFilterObj::inferShape(const TensorVec &input
return {{{on, oc, oh, ow}}};
}
} // namespace infini

2
src/operators/det.cc
View File

@ -11,7 +11,7 @@ optional<vector<Shape>> DetObj::inferShape(const TensorVec &inputs) const {
auto input = A->getDims();
int length = input.size();
if (length == 2) {
std::vector<int> output ={1};
std::vector<int> output = {1};
return {{output}};
} else {
std::vector<int> output(input.begin(), input.end() - 2);

58
src/operators/element_wise.cc
View File

@ -54,24 +54,24 @@ vector<int> ElementWiseObj::getOpAttrVector() const {
return {enum_to_underlying(type)};
}
MSELossObj::MSELossObj(GraphObj *graph, Tensor input0, Tensor input1, Reduction reduction, Tensor output)
: OperatorObj(OpType::MSELoss, {input0, input1}, {output}), reductionMode(reduction) {
MSELossObj::MSELossObj(GraphObj *graph, Tensor input0, Tensor input1,
Reduction reduction, Tensor output)
: OperatorObj(OpType::MSELoss, {input0, input1}, {output}),
reductionMode(reduction) {
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>>
MSELossObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>> MSELossObj::inferShape(const TensorVec &inputs) const {
const auto A = inputs[0], B = inputs[1];
if (A->getDims().size() != B->getDims().size() ||
A->getDims() != B->getDims())
return {};
if (reductionMode == None) {
return {{A->getDims()}};
return {{A->getDims()}};
} else {
Shape temp = { 1 };
return {{temp}};
Shape temp = {1};
return {{temp}};
}
}
@ -100,20 +100,19 @@ vector<int> MSELossObj::getOpAttrVector() const {
AddNObj::AddNObj(GraphObj *graph, int tensorNum, Tensor output, ...)
: OperatorObj(OpType::AddN), num(tensorNum) {
TensorVec temp;
Tensor *start = &output;
++start;
for(int i = 0; i < num; ++i) {
temp.push_back(*start);
start++;
}
setOutputs({output});
setInputs(temp);
TensorVec temp;
Tensor *start = &output;
++start;
for (int i = 0; i < num; ++i) {
temp.push_back(*start);
start++;
}
setOutputs({output});
setInputs(temp);
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>>
AddNObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>> AddNObj::inferShape(const TensorVec &inputs) const {
// For now,we only process the same dims here, broardcast will be considered
// in the opt layer.
const auto A = inputs[0];
@ -144,20 +143,19 @@ vector<int> AddNObj::getOpAttrVector() const {
MulNObj::MulNObj(GraphObj *graph, int tensorNum, Tensor output, ...)
: OperatorObj(OpType::MulN), num(tensorNum) {
TensorVec temp;
Tensor *start = &output;
++start;
for(int i = 0; i < num; ++i) {
temp.push_back(*start);
start++;
}
setOutputs({output});
setInputs(temp);
TensorVec temp;
Tensor *start = &output;
++start;
for (int i = 0; i < num; ++i) {
temp.push_back(*start);
start++;
}
setOutputs({output});
setInputs(temp);
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>>
MulNObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>> MulNObj::inferShape(const TensorVec &inputs) const {
// For now,we only process the same dims here, broardcast will be considered
// in the opt layer.
const auto A = inputs[0];

18
src/operators/transpose.cc
View File

@ -1,21 +1,23 @@
#include "operators/transpose.h"
namespace infini {
TransposeObj::TransposeObj(GraphObj *graph, Tensor input, Tensor output, int permute[4])
TransposeObj::TransposeObj(GraphObj *graph, Tensor input, Tensor output,
int permute[4])
: OperatorObj(OpType::Transpose, {input}, {output}) {
transposePermute[0] = permute[0];
transposePermute[1] = permute[1];
transposePermute[2] = permute[2];
transposePermute[3] = permute[3];
transposePermute[0] = permute[0];
transposePermute[1] = permute[1];
transposePermute[2] = permute[2];
transposePermute[3] = permute[3];
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>> TransposeObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>>
TransposeObj::inferShape(const TensorVec &inputs) const {
const auto A = inputs[0];
auto input = A->getDims();
auto output = input;
for(int i = 0; i < 4; ++i){
for (int i = 0; i < 4; ++i) {
output[i] = input[transposePermute[i]];
}
return {{output}};

42
src/operators/unary.cc
View File

@ -32,8 +32,10 @@ vector<int> UnaryObj::getOpAttrVector() const {
return {enum_to_underlying(type)};
}
ClipObj::ClipObj(GraphObj *graph, Tensor input, Tensor output, float min, float max)
: OperatorObj(OpType::Clip, {input}, {output}), minValue(min), maxValue(max) {
ClipObj::ClipObj(GraphObj *graph, Tensor input, Tensor output, float min,
float max)
: OperatorObj(OpType::Clip, {input}, {output}), minValue(min),
maxValue(max) {
IT_ASSERT(checkValid(graph));
}
@ -64,7 +66,7 @@ vector<int> ClipObj::getOpAttrVector() const {
}
FillObj::FillObj(GraphObj *graph, Tensor input, Tensor output, float value)
: OperatorObj(OpType::Fill, {input}, {output}), setValue(value) {
: OperatorObj(OpType::Fill, {input}, {output}), setValue(value) {
IT_ASSERT(checkValid(graph));
}
@ -98,7 +100,7 @@ L2LossObj::L2LossObj(GraphObj *graph, Tensor input, Tensor output)
}
optional<vector<Shape>> L2LossObj::inferShape(const TensorVec &inputs) const {
Shape temp = { 1 };
Shape temp = {1};
return {{temp}};
}
@ -121,12 +123,15 @@ vector<int> L2LossObj::getOpAttrVector() const {
return {enum_to_underlying(type)};
}
TransformObj::TransformObj(GraphObj *graph, Tensor input, Tensor output, float alpha, float beta)
: OperatorObj(OpType::Transform, {input}, {output}), alphaValue(alpha), betaValue(beta) {
TransformObj::TransformObj(GraphObj *graph, Tensor input, Tensor output,
float alpha, float beta)
: OperatorObj(OpType::Transform, {input}, {output}), alphaValue(alpha),
betaValue(beta) {
IT_ASSERT(checkValid(graph));
}
optional<vector<Shape>> TransformObj::inferShape(const TensorVec &inputs) const {
optional<vector<Shape>>
TransformObj::inferShape(const TensorVec &inputs) const {
const auto A = inputs[0];
return {{A->getDims()}};
}
@ -179,8 +184,10 @@ vector<int> CastObj::getOpAttrVector() const {
return {enum_to_underlying(type)};
}
CumsumObj::CumsumObj(GraphObj *graph, Tensor input, Tensor output, int axis, bool exclusive, bool reverse)
: OperatorObj(OpType::Cumsum, {input}, {output}), axisValue(axis), exclusiveValue(exclusive), reverseValue(reverse) {
CumsumObj::CumsumObj(GraphObj *graph, Tensor input, Tensor output, int axis,
bool exclusive, bool reverse)
: OperatorObj(OpType::Cumsum, {input}, {output}), axisValue(axis),
exclusiveValue(exclusive), reverseValue(reverse) {
IT_ASSERT(checkValid(graph));
}
@ -208,16 +215,19 @@ vector<int> CumsumObj::getOpAttrVector() const {
return {enum_to_underlying(type)};
}
// CumprodObj::CumprodObj(GraphObj *graph, Tensor input, Tensor output, int axis, bool exclusive, bool reverse)
// : OperatorObj(OpType::Cumprod, {input}, {output}), axisValue(axis), exclusiveValue(exclusive), reverseValue(reverse) {
// CumprodObj::CumprodObj(GraphObj *graph, Tensor input, Tensor output, int
// axis, bool exclusive, bool reverse)
// : OperatorObj(OpType::Cumprod, {input}, {output}), axisValue(axis),
// exclusiveValue(exclusive), reverseValue(reverse) {
// IT_ASSERT(checkValid(graph));
// }
//
// optional<vector<Shape>> CumprodObj::inferShape(const TensorVec &inputs) const {
//
// optional<vector<Shape>> CumprodObj::inferShape(const TensorVec &inputs) const
// {
// const auto A = inputs[0];
// return {{A->getDims()}};
// }
//
//
// std::string CumprodObj::toString() const {
// std::ostringstream os;
// os << OpRegistry::getOpName(type) << "[" << getGuid() << "]";
@ -225,14 +235,14 @@ vector<int> CumsumObj::getOpAttrVector() const {
// os << "output=" << outputs[0]->getGuid() << ")";
// return os.str();
// }
//
//
// vector<int> CumprodObj::getWorkloadVector() const {
// vector<int> ret{enum_to_underlying(type)};
// const Shape shape = outputs[0]->getDims();
// ret.insert(ret.end(), shape.begin(), shape.end());
// return ret;
// }
//
//
// vector<int> CumprodObj::getOpAttrVector() const {
// return {enum_to_underlying(type)};
// }

18
test/kernels/bang/test_bang_activation_backward.cc
View File

@ -3,16 +3,17 @@
#include "core/kernel.h"
#include "core/runtime.h"
#include "operators/activation_backward.h"
#include "operators/unary.h"
#include "operators/element_wise.h"
#include "operators/unary.h"
#include "test.h"
namespace infini {
template <class T, class D>
void testActivationBackward(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testActivationBackward(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();
@ -39,14 +40,17 @@ void testActivationBackward(const std::function<void(void *, size_t, DataType)>
bangGraph->dataMalloc();
bangRuntime->run(bangGraph);
auto diffXGpu = gpuOp->getOutput();
EXPECT_TRUE(1);
}
TEST(cnnl_ActivationBackward, run) {
testActivationBackward<ReluBackwardObj, ReluObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});
testActivationBackward<SigmoidBackwardObj, SigmoidObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});
testActivationBackward<TanhBackwardObj, TanhObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});
testActivationBackward<ReluBackwardObj, ReluObj>(IncrementalGenerator(),
Shape{1, 2, 2, 3});
testActivationBackward<SigmoidBackwardObj, SigmoidObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3});
testActivationBackward<TanhBackwardObj, TanhObj>(IncrementalGenerator(),
Shape{1, 2, 2, 3});
}
} // namespace infini

5
test/kernels/bang/test_bang_addn.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testaddN(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testaddN(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_cast.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testCast(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_ceil.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testCeil(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_clip.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testClip(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_copy.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testCopy(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

4
test/kernels/bang/test_bang_cumsum.cc
View File

@ -9,8 +9,8 @@
namespace infini {
template <class T>
void testCumsum(const std::function<void(void *, size_t, DataType)> &generator, int axis,
const Shape &shape) {
void testCumsum(const std::function<void(void *, size_t, DataType)> &generator,
int axis, const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_det.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testDet(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_divdemo.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testDivDemo(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testDivDemo(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_erf.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testErf(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_exp.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testExp(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_fill.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testFill(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

3
test/kernels/bang/test_bang_floordivtrunc.cc
View File

@ -43,7 +43,8 @@ void testFloorDivTrunc(
}
TEST(cnnl_FloorDivTrunc, run) {
testFloorDivTrunc<FloorDivTruncObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});
testFloorDivTrunc<FloorDivTruncObj>(IncrementalGenerator(),
Shape{1, 2, 2, 3});
}
} // namespace infini

2
test/kernels/bang/test_bang_l2loss.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testL2Loss(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_log.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testLog(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_maximum.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testMaximum(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testMaximum(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_minimum.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testMinimum(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testMinimum(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

14
test/kernels/bang/test_bang_mseloss.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testMSELoss(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testMSELoss(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();
@ -30,9 +29,12 @@ void testMSELoss(
Graph bangGraph = make_ref<GraphObj>(bangRuntime);
auto inputGpu1 = bangGraph->cloneTensor(inputCpu1);
auto inputGpu2 = bangGraph->cloneTensor(inputCpu2);
auto gpuOp1 = bangGraph->addOp<T>(inputGpu1, inputGpu2, MSELossObj::None, nullptr);
auto gpuOp2 = bangGraph->addOp<T>(inputGpu1, inputGpu2, MSELossObj::Sum, nullptr);
auto gpuOp3 = bangGraph->addOp<T>(inputGpu1, inputGpu2, MSELossObj::Mean, nullptr);
auto gpuOp1 =
bangGraph->addOp<T>(inputGpu1, inputGpu2, MSELossObj::None, nullptr);
auto gpuOp2 =
bangGraph->addOp<T>(inputGpu1, inputGpu2, MSELossObj::Sum, nullptr);
auto gpuOp3 =
bangGraph->addOp<T>(inputGpu1, inputGpu2, MSELossObj::Mean, nullptr);
bangGraph->dataMalloc();
bangRuntime->run(bangGraph);
auto outputGpu1 = gpuOp1->getOutput();

5
test/kernels/bang/test_bang_muln.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testmulN(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testmulN(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_negtensor.cc
View File

@ -9,8 +9,9 @@
namespace infini {
template <class T>
void testNegTensor(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testNegTensor(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_pad.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testPad(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();
@ -23,7 +23,8 @@ void testPad(const std::function<void(void *, size_t, DataType)> &generator,
// GPU
Graph bangGraph = make_ref<GraphObj>(bangRuntime);
auto inputGpu = bangGraph->cloneTensor(inputCpu);
auto gpuOp = bangGraph->addOp<T>(inputGpu, nullptr, vector<int>{1,1,1,1}, vector<int>{0,3});
auto gpuOp = bangGraph->addOp<T>(inputGpu, nullptr, vector<int>{1, 1, 1, 1},
vector<int>{0, 3});
bangGraph->dataMalloc();
bangRuntime->run(bangGraph);
auto outputGpu = gpuOp->getOutput();

5
test/kernels/bang/test_bang_pow.cc
View File

@ -9,9 +9,8 @@
namespace infini {
template <class T>
void testPow(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testPow(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_reciprocal.cc
View File

@ -9,8 +9,9 @@
namespace infini {
template <class T>
void testReciprocal(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testReciprocal(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_sqrt.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testSqrt(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_transform.cc
View File

@ -9,8 +9,9 @@
namespace infini {
template <class T>
void testTransform(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testTransform(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

5
test/kernels/bang/test_bang_transpose.cc
View File

@ -9,8 +9,9 @@
namespace infini {
template <class T>
void testTranspose(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
void testTranspose(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();

2
test/kernels/bang/test_bang_trigon.cc
View File

@ -10,7 +10,7 @@ namespace infini {
template <class T>
void testTrigon(const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape) {
const Shape &shape) {
// Runtime
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto bangRuntime = make_ref<BangRuntimeObj>();