Skip to content

GitLab

Commit bcc66c9b authored by Astha Rai's avatar Astha Rai
Browse files

updated column major implementation

parent f22f565c
Hide whitespace changes
Inline Side-by-side
Showing with 194 additions and 33 deletions
example/65_hip_tensor_permute/CMakeLists.txt
View file @ bcc66c9b
add_example_executable(example_elementwise_permute_4D_fp32_ht elementwise_permute_4D_fp32_ht.cpp)
add_example_executable(example_elementwise_permute_4D_fp16_ht elementwise_permute_4D_fp16_ht.cpp)
add_example_executable(example_elementwise_permute_4D_fp32_col elementwise_permute_4D_fp32_col.cpp)
add_example_executable(example_elementwise_permute_4D_fp16_col elementwise_permute_4D_fp16_col.cpp)
example/65_hip_tensor_permute/elementwise_permute_4D_fp16_col.cpp 0 → 100644
View file @ bcc66c9b
#include <iostream>
#include <cstdlib>
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/element/binary_element_wise_operation.hpp"
#include "ck/tensor_operation/gpu/device/impl/device_elementwise_impl_ht.hpp"
#include "ck/library/utility/algorithm.hpp"
#include "ck/library/utility/check_err.hpp"
#include "ck/library/utility/device_memory.hpp"
#include "ck/library/utility/host_tensor.hpp"
#include "ck/library/utility/host_tensor_generator.hpp"
using F16 = ck::half_t;
using F32 = float;
using ADataType = F16;
using BDataType = F16;
using PassThrough = ck::tensor_operation::element_wise::PassThrough;
using UnaryOp = ck::tensor_operation::element_wise::UnarySquare;
using Scale = ck::tensor_operation::element_wise::Scale;
using DeviceElementwisePermuteInstance =
ck::tensor_operation::device::DeviceElementwiseImpl<ck::Tuple<ADataType>, // InDataTypeTuple
ck::Tuple<BDataType>, // OutDataTypeTuple
PassThrough, // ElementwiseOp
UnaryOp, // UnaryOp
Scale, // Scalar
4, // NumDim
8, // MPerThread
ck::Sequence<1>, // InScalarPerVectorSeq
ck::Sequence<1>>; // OutScalarPerVectorSeq
template <typename HostTensorA, typename HostTensorB, typename FunctorA, typename FunctorB>
void host_elementwise4D(HostTensorB& B_nhwc,
const HostTensorA& A_nchw,
FunctorA functor_a,
FunctorB functor_b,
float scale)
{
for(std::size_t n = 0; n < A_nchw.mDesc.GetLengths()[0]; ++n)
for(std::size_t c = 0; c < A_nchw.mDesc.GetLengths()[1]; ++c)
for(std::size_t h = 0; h < A_nchw.mDesc.GetLengths()[2]; ++h)
for(std::size_t w = 0; w < A_nchw.mDesc.GetLengths()[3]; ++w)
{
ADataType tmp_val;
auto a_val = A_nchw(n, c, h, w);
functor_b(tmp_val, a_val);
functor_a(B_nhwc(n, h, w, c), scale * tmp_val);
}
}
int main()
{
bool do_verification = true;
bool time_kernel = true;
std::vector<std::size_t> nchw = {4, 2, 1, 8};
std::vector<std::size_t> nhwc = {4, 1, 8, 2};
Tensor<ADataType> a(nchw);
Tensor<BDataType> b(nhwc);
float scale = 1.f;
// a.GenerateTensorValue(GeneratorTensor_3<ADataType>{0.0, 1.0});
//
// for(std::size_t i = 0; i < a.mData.size(); i++){
// a.mData[i] = i;
// }
//
auto i = 0;
for(std::size_t w = 0; w < a.mDesc.GetLengths()[3]; ++w)
for(std::size_t h = 0; h < a.mDesc.GetLengths()[2]; ++h)
for(std::size_t c = 0; c < a.mDesc.GetLengths()[1]; ++c)
for(std::size_t n = 0; n < a.mDesc.GetLengths()[0]; ++n)
{
a.mData[(n * nchw[1] * nchw[2] * nchw[3]) + (c * nchw[2] * nchw[3]) +
(h * nchw[3]) + w] = i;
i++;
}
DeviceMem a_device_buf(sizeof(ADataType) * a.mDesc.GetElementSpaceSize());
DeviceMem b_device_buf(sizeof(BDataType) * b.mDesc.GetElementSpaceSize());
a_device_buf.ToDevice(a.mData.data());
std::array<const void*, 1> input = {a_device_buf.GetDeviceBuffer()};
std::array<void*, 1> output = {b_device_buf.GetDeviceBuffer()};
std::array<ck::index_t, 4> ab_lengths;
std::array<ck::index_t, 4> a_strides = {static_cast<int>(nchw[1] * nchw[2] * nchw[3]),
static_cast<int>(nchw[2] * nchw[3]),
static_cast<int>(nchw[3]),
1};
std::array<ck::index_t, 4> b_strides = {static_cast<int>(nhwc[1] * nhwc[2] * nhwc[3]),
1,
static_cast<int>(nhwc[2] * nhwc[3]),
static_cast<int>(nhwc[3])};
ck::ranges::copy(nchw, ab_lengths.begin());
auto broadcastPermute = DeviceElementwisePermuteInstance{};
auto argument = broadcastPermute.MakeArgumentPointer(ab_lengths,
{a_strides},
{b_strides},
input,
output,
PassThrough{},
UnaryOp{},
Scale{scale});
if(!broadcastPermute.IsSupportedArgument(argument.get()))
{
throw std::runtime_error(
"The runtime parameters seems not supported by the device instance, exiting!");
};
std::cout << "A (nchw): " << a.mDesc << std::endl;
std::cout << "B (nhwc): " << b.mDesc << std::endl;
auto broadcastPermute_invoker_ptr = broadcastPermute.MakeInvokerPointer();
float ave_time =
broadcastPermute_invoker_ptr->Run(argument.get(), StreamConfig{nullptr, time_kernel});
std::size_t flop = std::size_t(2) * nchw[0] * nchw[1] * nchw[2] * nchw[3];
std::size_t num_btype = sizeof(ADataType) * (nchw[0] * nchw[1] * nchw[2] * nchw[3]) +
sizeof(BDataType) * (nchw[0] * nchw[1] * nchw[2] * nchw[3]);
float tflops = static_cast<float>(flop) / 1.E9 / ave_time;
float gb_per_sec = num_btype / 1.E6 / ave_time;
std::cout << "Perf: " << ave_time << " ms, " << tflops << " TFlops, " << gb_per_sec << " GB/s"
<< std::endl;
bool pass = true;
LogRangeAsType<float>(std::cout << "A : ", a.mData, ",") << std::endl;
if(do_verification)
{
b_device_buf.FromDevice(b.mData.data());
Tensor<BDataType> host_b(nhwc);
host_elementwise4D(host_b, a, PassThrough{}, UnaryOp{}, scale);
LogRangeAsType<float>(std::cout << "B : ", b.mData, ",") << std::endl;
LogRangeAsType<float>(std::cout << "Host B : ", host_b.mData, ",") << std::endl;
pass &=
ck::utils::check_err(b.mData, host_b.mData, "Error: Incorrect results b", 1e-3, 1e-3);
}
return pass ? 0 : 1;
}
example/65_hip_tensor_permute/elementwise_permute_4D_fp32_col.cpp
View file @ bcc66c9b
......@@ -28,25 +28,25 @@ using DeviceElementwisePermuteInstance =
Scale, // Scalar
4, // NumDim
8, // MPerThread
ck::Sequence<8>, // InScalarPerVectorSeq
ck::Sequence<1>, // InScalarPerVectorSeq
ck::Sequence<1>>; // OutScalarPerVectorSeq
template <typename HostTensorA, typename HostTensorB, typename FunctorA, typename FunctorB>
void host_elementwise4D(HostTensorB& B_cwhn,
const HostTensorA& A_whcn,
void host_elementwise4D(HostTensorB& B_nhwc,
const HostTensorA& A_nchw,
FunctorA functor_a,
FunctorB functor_b,
float scale)
{
for(std::size_t w = 0; w < A_whcn.mDesc.GetLengths()[0]; ++w)
for(std::size_t h = 0; h < A_whcn.mDesc.GetLengths()[1]; ++h)
for(std::size_t c = 0; c < A_whcn.mDesc.GetLengths()[2]; ++c)
for(std::size_t n = 0; n < A_whcn.mDesc.GetLengths()[3]; ++n)
for(std::size_t n = 0; n < A_nchw.mDesc.GetLengths()[0]; ++n)
for(std::size_t c = 0; c < A_nchw.mDesc.GetLengths()[1]; ++c)
for(std::size_t h = 0; h < A_nchw.mDesc.GetLengths()[2]; ++h)
for(std::size_t w = 0; w < A_nchw.mDesc.GetLengths()[3]; ++w)
{
ADataType tmp_val;
auto a_val = A_whcn(w, h, c, n);
auto a_val = A_nchw(n, c, h, w);
functor_b(tmp_val, a_val);
functor_a(B_cwhn(c, w, h, n), scale * tmp_val);
functor_a(B_nhwc(n, h, w, c), scale * tmp_val);
}
}
......@@ -55,18 +55,27 @@ int main()
bool do_verification = true;
bool time_kernel = true;
std::vector<std::size_t> whcn = {4, 2, 1, 8};
std::vector<std::size_t> cwhn = {1, 4, 2, 8};
Tensor<ADataType> a(whcn);
Tensor<BDataType> b(cwhn);
std::vector<std::size_t> nchw = {4, 2, 1, 8};
std::vector<std::size_t> nhwc = {4, 1, 8, 2};
Tensor<ADataType> a(nchw);
Tensor<BDataType> b(nhwc);
float scale = 1.f;
// a.GenerateTensorValue(GeneratorTensor_3<ADataType>{0.0, 1.0});
for(int i = 0; i < int(a.mData.size()); i++)
{
a.mData[i] = i;
}
//
// for(std::size_t i = 0; i < a.mData.size(); i++){
// a.mData[i] = i;
// }
//
auto i = 0;
for(std::size_t w = 0; w < a.mDesc.GetLengths()[3]; ++w)
for(std::size_t h = 0; h < a.mDesc.GetLengths()[2]; ++h)
for(std::size_t c = 0; c < a.mDesc.GetLengths()[1]; ++c)
for(std::size_t n = 0; n < a.mDesc.GetLengths()[0]; ++n)
{
a.mData[(n * nchw[1] * nchw[2] * nchw[3]) + (c * nchw[2] * nchw[3]) +
(h * nchw[3]) + w] = i;
i++;
}
DeviceMem a_device_buf(sizeof(ADataType) * a.mDesc.GetElementSpaceSize());
DeviceMem b_device_buf(sizeof(BDataType) * b.mDesc.GetElementSpaceSize());
......@@ -77,17 +86,16 @@ int main()
std::array<void*, 1> output = {b_device_buf.GetDeviceBuffer()};
std::array<ck::index_t, 4> ab_lengths;
std::array<ck::index_t, 4> a_strides = {static_cast<int>(whcn[0] * whcn[1] * whcn[2]),
static_cast<int>(whcn[0] * whcn[1]),
static_cast<int>(whcn[0]),
std::array<ck::index_t, 4> a_strides = {static_cast<int>(nchw[1] * nchw[2] * nchw[3]),
static_cast<int>(nchw[2] * nchw[3]),
static_cast<int>(nchw[3]),
1};
std::array<ck::index_t, 4> b_strides = {static_cast<int>(cwhn[0] * cwhn[1] * cwhn[2]),
std::array<ck::index_t, 4> b_strides = {static_cast<int>(nhwc[1] * nhwc[2] * nhwc[3]),
1,
static_cast<int>(cwhn[0] * cwhn[1]),
static_cast<int>(cwhn[0])};
static_cast<int>(nhwc[2] * nhwc[3]),
static_cast<int>(nhwc[3])};
ck::ranges::copy(whcn, ab_lengths.begin());
ck::ranges::copy(nchw, ab_lengths.begin());
auto broadcastPermute = DeviceElementwisePermuteInstance{};
auto argument = broadcastPermute.MakeArgumentPointer(ab_lengths,
......@@ -105,16 +113,16 @@ int main()
"The runtime parameters seems not supported by the device instance, exiting!");
};
std::cout << "A (whcn): " << a.mDesc << std::endl;
std::cout << "B (cwhn): " << b.mDesc << std::endl;
std::cout << "A (nchw): " << a.mDesc << std::endl;
std::cout << "B (nhwc): " << b.mDesc << std::endl;
auto broadcastPermute_invoker_ptr = broadcastPermute.MakeInvokerPointer();
float ave_time =
broadcastPermute_invoker_ptr->Run(argument.get(), StreamConfig{nullptr, time_kernel});
std::size_t flop = std::size_t(2) * whcn[0] * whcn[1] * whcn[2] * whcn[3];
std::size_t flop = std::size_t(2) * nchw[0] * nchw[1] * nchw[2] * nchw[3];
std::size_t num_btype = sizeof(ADataType) * (whcn[0] * whcn[1] * whcn[2] * whcn[3]) +
sizeof(BDataType) * (whcn[0] * whcn[1] * whcn[2] * whcn[3]);
std::size_t num_btype = sizeof(ADataType) * (nchw[0] * nchw[1] * nchw[2] * nchw[3]) +
sizeof(BDataType) * (nchw[0] * nchw[1] * nchw[2] * nchw[3]);
float tflops = static_cast<float>(flop) / 1.E9 / ave_time;
......@@ -129,11 +137,12 @@ int main()
if(do_verification)
{
b_device_buf.FromDevice(b.mData.data());
Tensor<BDataType> host_b(cwhn);
Tensor<BDataType> host_b(nhwc);
host_elementwise4D(host_b, a, PassThrough{}, UnaryOp{}, scale);
LogRangeAsType<float>(std::cout << "B : ", b.mData, ",") << std::endl;
LogRangeAsType<float>(std::cout << "Host B : ", host_b.mData, ",") << std::endl;
pass &=
ck::utils::check_err(b.mData, host_b.mData, "Error: Incorrect results b", 1e-3, 1e-3);
}
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment