mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-24 14:54:47 +00:00
Fusion Conv+Bias+ReLU(+Add) (#62)
* fix relu
* clean up
* clean up
* adding 1x1 conv
* adding 1x1 conv
* added 1x1 conv
* refactor
* refactor
* refactor
* added profiler for conv+bias+relu+add
* clean up
* adding conv+bias+relu
* adding conv+bias+relu
* added conv+bias+relu
* Update README.md
* update cpu verification
* adding c shuffle
* update static_tensor for dealing with invalid element
* adding c shuffle
* debugging
* fix bug
* convert to fp16 before shuffle
* shuffle more than one M/NRepeat
* clean up
* remove coordinate step hack from GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v3r1
* clean up
* remove coordinate step hack from all gridwise gemm xdl
* clean up coordinate step hack
* clean up coordinate step hack
* ThreadwiseTensorSliceTransfer_v3r2 support pointwise op on both src and dst
* adding output shuffle in conv+bias+relu+add
* update
* added conv+bias+relu+add with c shuffle
* added conv+bias+relu+add with c shuffle
* fix forward_sweep bugs in threadwise copy
* clean up
* refactor
* clean up
* clean up
* added conv_c_shuffle+bias_relu
* clean up
* added conv+bias+relu+atomic_add
* clean up
* clean up
* clean up
* clean up
* clean up
* clean up
* misc fixes; add 1x1 specialization
* clean up
* delete unused device op
* clean up
* add support for odd C value
[ROCm/composable_kernel commit: acbd7bd7c5]
This commit is contained in:
Chao Liu
305
profiler/include/profile_conv_fwd_bias_relu_add_impl.hpp
Normal file
305
profiler/include/profile_conv_fwd_bias_relu_add_impl.hpp
Normal file
@@ -0,0 +1,305 @@
|
||||
#pragma once
|
||||
#include "config.hpp"
|
||||
#include "device.hpp"
|
||||
#include "host_tensor.hpp"
|
||||
#include "host_tensor_generator.hpp"
|
||||
#include "host_conv.hpp"
|
||||
#include "tensor_layout.hpp"
|
||||
#include "device_tensor.hpp"
|
||||
#include "device_conv_fwd_bias_activation_add.hpp"
|
||||
#include "element_wise_operation.hpp"
|
||||
|
||||
namespace ck {
|
||||
namespace tensor_operation {
|
||||
namespace device {
|
||||
namespace device_conv2d_fwd_bias_activation_add_instance {
|
||||
|
||||
using DeviceConvFwdBiasReluAddPtr =
|
||||
DeviceConvFwdBiasActivationAddPtr<ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::AddReluAdd>;
|
||||
|
||||
void add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instances(
|
||||
std::vector<DeviceConvFwdBiasReluAddPtr>&);
|
||||
|
||||
} // namespace device_conv2d_fwd_bias_activation_add_instance
|
||||
} // namespace device
|
||||
} // namespace tensor_operation
|
||||
} // namespace ck
|
||||
|
||||
namespace ck {
|
||||
namespace profiler {
|
||||
|
||||
template <typename TIn,
|
||||
typename TWei,
|
||||
typename TOut,
|
||||
typename InElementOp,
|
||||
typename WeiElementOp,
|
||||
typename OutElementOp>
|
||||
void host_reference_calculation(const Tensor<TIn>& in_n_c_hi_wi,
|
||||
const Tensor<TWei>& wei_k_c_y_x,
|
||||
Tensor<TOut>& out_n_k_ho_wo,
|
||||
const Tensor<TOut>& bias_k,
|
||||
const Tensor<TOut>& resi_n_k_ho_wo,
|
||||
const std::vector<ck::index_t>& conv_strides,
|
||||
const std::vector<ck::index_t>& conv_dilations,
|
||||
const std::vector<ck::index_t>& in_left_pads,
|
||||
const std::vector<ck::index_t>& /* in_right_pads */,
|
||||
const InElementOp& in_element_op,
|
||||
const WeiElementOp& wei_element_op,
|
||||
const OutElementOp& out_element_op)
|
||||
{
|
||||
auto f_nchw = [&](auto n, auto k, auto ho, auto wo) {
|
||||
double v = 0;
|
||||
for(int c = 0; c < wei_k_c_y_x.mDesc.GetLengths()[1]; ++c)
|
||||
{
|
||||
for(int y = 0; y < wei_k_c_y_x.mDesc.GetLengths()[2]; ++y)
|
||||
{
|
||||
int hi = ho * conv_strides[0] + y * conv_dilations[0] - in_left_pads[0];
|
||||
for(int x = 0; x < wei_k_c_y_x.mDesc.GetLengths()[3]; ++x)
|
||||
{
|
||||
int wi = wo * conv_strides[1] + x * conv_dilations[1] - in_left_pads[1];
|
||||
if(hi >= 0 && hi < in_n_c_hi_wi.mDesc.GetLengths()[2] && wi >= 0 &&
|
||||
wi < in_n_c_hi_wi.mDesc.GetLengths()[3])
|
||||
{
|
||||
v += in_element_op(static_cast<const double>(in_n_c_hi_wi(n, c, hi, wi))) *
|
||||
wei_element_op(static_cast<const double>(wei_k_c_y_x(k, c, y, x)));
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
out_n_k_ho_wo(n, k, ho, wo) = out_element_op(v, bias_k(k), resi_n_k_ho_wo(n, k, ho, wo));
|
||||
};
|
||||
|
||||
make_ParallelTensorFunctor(f_nchw,
|
||||
out_n_k_ho_wo.mDesc.GetLengths()[0],
|
||||
out_n_k_ho_wo.mDesc.GetLengths()[1],
|
||||
out_n_k_ho_wo.mDesc.GetLengths()[2],
|
||||
out_n_k_ho_wo.mDesc.GetLengths()[3])(
|
||||
std::thread::hardware_concurrency());
|
||||
}
|
||||
|
||||
template <int NDimSpatial,
|
||||
typename InDataType,
|
||||
typename WeiDataType,
|
||||
typename OutDataType,
|
||||
typename InLayout,
|
||||
typename WeiLayout,
|
||||
typename OutLayout>
|
||||
void profile_conv_fwd_bias_relu_add_impl(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
ck::index_t N,
|
||||
ck::index_t K,
|
||||
ck::index_t C,
|
||||
std::vector<ck::index_t> input_spatial_lengths,
|
||||
std::vector<ck::index_t> filter_spatial_lengths,
|
||||
std::vector<ck::index_t> output_spatial_lengths,
|
||||
std::vector<ck::index_t> conv_filter_strides,
|
||||
std::vector<ck::index_t> conv_filter_dilations,
|
||||
std::vector<ck::index_t> input_left_pads,
|
||||
std::vector<ck::index_t> input_right_pads)
|
||||
{
|
||||
const ck::index_t Y = filter_spatial_lengths[0];
|
||||
const ck::index_t X = filter_spatial_lengths[1];
|
||||
|
||||
const ck::index_t Hi = input_spatial_lengths[0];
|
||||
const ck::index_t Wi = input_spatial_lengths[1];
|
||||
|
||||
const ck::index_t Ho = output_spatial_lengths[0];
|
||||
const ck::index_t Wo = output_spatial_lengths[1];
|
||||
|
||||
auto f_host_tensor_descriptor =
|
||||
[](std::size_t N_, std::size_t C_, std::size_t H, std::size_t W, auto layout) {
|
||||
if constexpr(is_same<decltype(layout), ck::tensor_layout::convolution::NCHW>::value ||
|
||||
is_same<decltype(layout), ck::tensor_layout::convolution::KCYX>::value ||
|
||||
is_same<decltype(layout), ck::tensor_layout::convolution::NKHW>::value)
|
||||
{
|
||||
return HostTensorDescriptor(std::vector<std::size_t>({N_, C_, H, W}),
|
||||
std::vector<std::size_t>({C_ * H * W, H * W, W, 1}));
|
||||
}
|
||||
else if constexpr(is_same<decltype(layout), tensor_layout::convolution::NHWC>::value ||
|
||||
is_same<decltype(layout), tensor_layout::convolution::KYXC>::value ||
|
||||
is_same<decltype(layout), tensor_layout::convolution::NHWK>::value)
|
||||
{
|
||||
return HostTensorDescriptor(std::vector<std::size_t>({N_, C_, H, W}),
|
||||
std::vector<std::size_t>({C_ * H * W, 1, W * C_, C_}));
|
||||
}
|
||||
};
|
||||
|
||||
Tensor<InDataType> in_n_c_hi_wi(f_host_tensor_descriptor(N, C, Hi, Wi, InLayout{}));
|
||||
Tensor<WeiDataType> wei_k_c_y_x(f_host_tensor_descriptor(K, C, Y, X, WeiLayout{}));
|
||||
Tensor<OutDataType> out_n_k_ho_wo_host_result(
|
||||
f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
Tensor<OutDataType> out_n_k_ho_wo_device_result(
|
||||
f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
|
||||
// bias: assume contiguous 1d vector
|
||||
Tensor<OutDataType> bias_k(
|
||||
HostTensorDescriptor(std::vector<std::size_t>({static_cast<std::size_t>(K)})));
|
||||
|
||||
// residual: assume same layout as output tensor
|
||||
Tensor<OutDataType> resi_n_k_ho_wo(f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
|
||||
std::cout << "in_n_c_hi_wi: " << in_n_c_hi_wi.mDesc << std::endl;
|
||||
std::cout << "wei_k_c_y_x: " << wei_k_c_y_x.mDesc << std::endl;
|
||||
std::cout << "out_n_k_ho_wo: " << out_n_k_ho_wo_host_result.mDesc << std::endl;
|
||||
std::cout << "bias_k: " << bias_k.mDesc << std::endl;
|
||||
std::cout << "resi_n_k_ho_wo: " << resi_n_k_ho_wo.mDesc << std::endl;
|
||||
|
||||
switch(init_method)
|
||||
{
|
||||
case 0: break;
|
||||
case 1:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_2<InDataType>{-5, 5});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_2<WeiDataType>{-5, 5});
|
||||
bias_k.GenerateTensorValue(GeneratorTensor_2<OutDataType>{-5, 5});
|
||||
resi_n_k_ho_wo.GenerateTensorValue(GeneratorTensor_2<OutDataType>{-5, 5});
|
||||
break;
|
||||
default:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_3<InDataType>{0.0, 1.0});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_3<WeiDataType>{-0.5, 0.5});
|
||||
bias_k.GenerateTensorValue(GeneratorTensor_3<OutDataType>{0.0, 1.0});
|
||||
resi_n_k_ho_wo.GenerateTensorValue(GeneratorTensor_3<OutDataType>{0.0, 1.0});
|
||||
}
|
||||
|
||||
using InElementOp = ck::tensor_operation::element_wise::PassThrough;
|
||||
using WeiElementOp = ck::tensor_operation::element_wise::PassThrough;
|
||||
using OutElementOp = ck::tensor_operation::element_wise::AddReluAdd;
|
||||
|
||||
if(do_verification)
|
||||
{
|
||||
host_reference_calculation(in_n_c_hi_wi,
|
||||
wei_k_c_y_x,
|
||||
out_n_k_ho_wo_host_result,
|
||||
bias_k,
|
||||
resi_n_k_ho_wo,
|
||||
conv_filter_strides,
|
||||
conv_filter_dilations,
|
||||
input_left_pads,
|
||||
input_right_pads,
|
||||
InElementOp{},
|
||||
WeiElementOp{},
|
||||
OutElementOp{});
|
||||
}
|
||||
|
||||
DeviceMem in_device_buf(sizeof(InDataType) * in_n_c_hi_wi.mDesc.GetElementSpace());
|
||||
DeviceMem wei_device_buf(sizeof(WeiDataType) * wei_k_c_y_x.mDesc.GetElementSpace());
|
||||
DeviceMem out_device_buf(sizeof(OutDataType) *
|
||||
out_n_k_ho_wo_device_result.mDesc.GetElementSpace());
|
||||
DeviceMem bias_device_buf(sizeof(OutDataType) * bias_k.mDesc.GetElementSpace());
|
||||
DeviceMem resi_device_buf(sizeof(OutDataType) * resi_n_k_ho_wo.mDesc.GetElementSpace());
|
||||
|
||||
in_device_buf.ToDevice(in_n_c_hi_wi.mData.data());
|
||||
wei_device_buf.ToDevice(wei_k_c_y_x.mData.data());
|
||||
bias_device_buf.ToDevice(bias_k.mData.data());
|
||||
resi_device_buf.ToDevice(resi_n_k_ho_wo.mData.data());
|
||||
|
||||
using DeviceConvFwdBiasReluAddPtr = ck::tensor_operation::device::
|
||||
DeviceConvFwdBiasActivationAddPtr<InElementOp, WeiElementOp, OutElementOp>;
|
||||
|
||||
// add device operator instances
|
||||
std::vector<DeviceConvFwdBiasReluAddPtr> op_ptrs;
|
||||
|
||||
if constexpr(ck::is_same_v<ck::remove_cv_t<InDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<WeiDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<OutDataType>, ck::half_t>)
|
||||
{
|
||||
ck::tensor_operation::device::device_conv2d_fwd_bias_activation_add_instance::
|
||||
add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instances(op_ptrs);
|
||||
}
|
||||
|
||||
if(op_ptrs.size() <= 0)
|
||||
{
|
||||
throw std::runtime_error("wrong! no device Conv instance found");
|
||||
}
|
||||
|
||||
std::string best_conv_name;
|
||||
float best_ave_time = 0;
|
||||
float best_tflops = 0;
|
||||
float best_gb_per_sec = 0;
|
||||
|
||||
// profile device Conv instances
|
||||
for(auto& op_ptr : op_ptrs)
|
||||
{
|
||||
auto argument_ptr = op_ptr->MakeArgumentPointer(
|
||||
static_cast<const InDataType*>(in_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const WeiDataType*>(wei_device_buf.GetDeviceBuffer()),
|
||||
static_cast<OutDataType*>(out_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const OutDataType*>(bias_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const OutDataType*>(resi_device_buf.GetDeviceBuffer()),
|
||||
N,
|
||||
K,
|
||||
C,
|
||||
input_spatial_lengths,
|
||||
filter_spatial_lengths,
|
||||
output_spatial_lengths,
|
||||
conv_filter_strides,
|
||||
conv_filter_dilations,
|
||||
input_left_pads,
|
||||
input_right_pads,
|
||||
InElementOp{},
|
||||
WeiElementOp{},
|
||||
OutElementOp{});
|
||||
|
||||
auto invoker_ptr = op_ptr->MakeInvokerPointer();
|
||||
|
||||
if(op_ptr->IsSupportedArgument(argument_ptr.get()))
|
||||
{
|
||||
std::string conv_name = op_ptr->GetTypeString();
|
||||
|
||||
float ave_time = invoker_ptr->Run(argument_ptr.get(), nrepeat);
|
||||
|
||||
std::size_t flop = std::size_t(2) * N * K * Ho * Wo * C * Y * X;
|
||||
|
||||
std::size_t num_btype =
|
||||
sizeof(InDataType) * (N * C * Hi * Wi) + sizeof(WeiDataType) * (K * C * Y * X) +
|
||||
sizeof(OutDataType) * (N * K * Ho * Wo) + sizeof(OutDataType) * (K) +
|
||||
sizeof(OutDataType) * (N * K * Ho * Wo);
|
||||
|
||||
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, " << conv_name << std::endl;
|
||||
|
||||
if(tflops > best_tflops)
|
||||
{
|
||||
best_conv_name = conv_name;
|
||||
best_tflops = tflops;
|
||||
best_ave_time = ave_time;
|
||||
best_gb_per_sec = gb_per_sec;
|
||||
}
|
||||
|
||||
if(do_verification)
|
||||
{
|
||||
out_device_buf.FromDevice(out_n_k_ho_wo_device_result.mData.data());
|
||||
|
||||
check_error(out_n_k_ho_wo_host_result, out_n_k_ho_wo_device_result);
|
||||
|
||||
if(do_log)
|
||||
{
|
||||
LogRangeAsType<float>(std::cout << "in : ", in_n_c_hi_wi.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(std::cout << "wei: ", wei_k_c_y_x.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(
|
||||
std::cout << "out_host : ", out_n_k_ho_wo_host_result.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(
|
||||
std::cout << "out_device: ", out_n_k_ho_wo_device_result.mData, ",")
|
||||
<< std::endl;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
std::cout << "Best Perf: " << best_ave_time << " ms, " << best_tflops << " TFlops, "
|
||||
<< best_gb_per_sec << " GB/s, " << best_conv_name << std::endl;
|
||||
}
|
||||
|
||||
} // namespace profiler
|
||||
} // namespace ck
|
||||
328
profiler/include/profile_conv_fwd_bias_relu_atomic_add_impl.hpp
Normal file
328
profiler/include/profile_conv_fwd_bias_relu_atomic_add_impl.hpp
Normal file
@@ -0,0 +1,328 @@
|
||||
#pragma once
|
||||
#include "config.hpp"
|
||||
#include "device.hpp"
|
||||
#include "host_tensor.hpp"
|
||||
#include "host_tensor_generator.hpp"
|
||||
#include "host_conv.hpp"
|
||||
#include "tensor_layout.hpp"
|
||||
#include "device_tensor.hpp"
|
||||
#include "device_conv_fwd_bias_activation.hpp"
|
||||
#include "element_wise_operation.hpp"
|
||||
|
||||
namespace ck {
|
||||
namespace tensor_operation {
|
||||
namespace device {
|
||||
namespace device_conv2d_fwd_bias_activation_atomic_add_instance {
|
||||
|
||||
using DeviceConvFwdBiasReluPtr =
|
||||
DeviceConvFwdBiasActivationPtr<ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::AddRelu>;
|
||||
|
||||
void add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instances(
|
||||
std::vector<DeviceConvFwdBiasReluPtr>&);
|
||||
|
||||
} // namespace device_conv2d_fwd_bias_activation_atomic_add_instance
|
||||
} // namespace device
|
||||
} // namespace tensor_operation
|
||||
} // namespace ck
|
||||
|
||||
namespace ck {
|
||||
namespace profiler {
|
||||
|
||||
void cpu_conv_bias_relu_atomic_add(ck::half_t* in_ptr,
|
||||
ck::half_t* weight_ptr,
|
||||
ck::half_t* output_ptr,
|
||||
ck::half_t* bias_ptr,
|
||||
const ck::index_t N,
|
||||
const ck::index_t K,
|
||||
const ck::index_t C,
|
||||
const ck::index_t Y,
|
||||
const ck::index_t X,
|
||||
const ck::index_t Hi,
|
||||
const ck::index_t Wi,
|
||||
const ck::index_t Ho,
|
||||
const ck::index_t Wo,
|
||||
const ck::index_t Stride,
|
||||
const ck::index_t Dilation,
|
||||
const ck::index_t Pad)
|
||||
{
|
||||
|
||||
const auto in_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(N),
|
||||
static_cast<std::size_t>(Hi),
|
||||
static_cast<std::size_t>(Wi),
|
||||
static_cast<std::size_t>(C)});
|
||||
const auto wei_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(K),
|
||||
static_cast<std::size_t>(Y),
|
||||
static_cast<std::size_t>(X),
|
||||
static_cast<std::size_t>(C)});
|
||||
const auto out_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(N),
|
||||
static_cast<std::size_t>(Ho),
|
||||
static_cast<std::size_t>(Wo),
|
||||
static_cast<std::size_t>(K)});
|
||||
const auto bias_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(K)});
|
||||
|
||||
auto f_k = [&](auto k) {
|
||||
for(int n = 0; n < N; ++n)
|
||||
{
|
||||
for(int ho = 0; ho < Ho; ++ho)
|
||||
{
|
||||
for(int wo = 0; wo < Wo; ++wo)
|
||||
{
|
||||
double v = 0;
|
||||
for(int c = 0; c < C; ++c)
|
||||
{
|
||||
for(int y = 0; y < Y; ++y)
|
||||
{
|
||||
int hi = ho * Stride + y * Dilation - Pad;
|
||||
for(int x = 0; x < X; ++x)
|
||||
{
|
||||
int wi = wo * Stride + x * Dilation - Pad;
|
||||
if(hi >= 0 && hi < Hi && wi >= 0 && wi < Wi)
|
||||
{
|
||||
double in =
|
||||
in_ptr[in_desc.GetOffsetFromMultiIndex(n, hi, wi, c)];
|
||||
double wei =
|
||||
weight_ptr[wei_desc.GetOffsetFromMultiIndex(k, y, x, c)];
|
||||
|
||||
v += in * wei;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
v += bias_ptr[bias_desc.GetOffsetFromMultiIndex(k)];
|
||||
|
||||
v = v > 0 ? v : 0;
|
||||
|
||||
output_ptr[out_desc.GetOffsetFromMultiIndex(n, ho, wo, k)] = v;
|
||||
}
|
||||
}
|
||||
}
|
||||
};
|
||||
|
||||
make_ParallelTensorFunctor(f_k, K)(std::thread::hardware_concurrency());
|
||||
}
|
||||
|
||||
template <int NDimSpatial,
|
||||
typename InDataType,
|
||||
typename WeiDataType,
|
||||
typename OutDataType,
|
||||
typename InLayout,
|
||||
typename WeiLayout,
|
||||
typename OutLayout>
|
||||
void profile_conv_fwd_bias_relu_atomic_add_impl(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
ck::index_t N,
|
||||
ck::index_t K,
|
||||
ck::index_t C,
|
||||
std::vector<ck::index_t> input_spatial_lengths,
|
||||
std::vector<ck::index_t> filter_spatial_lengths,
|
||||
std::vector<ck::index_t> output_spatial_lengths,
|
||||
std::vector<ck::index_t> conv_filter_strides,
|
||||
std::vector<ck::index_t> conv_filter_dilations,
|
||||
std::vector<ck::index_t> input_left_pads,
|
||||
std::vector<ck::index_t> input_right_pads)
|
||||
{
|
||||
const ck::index_t Y = filter_spatial_lengths[0];
|
||||
const ck::index_t X = filter_spatial_lengths[1];
|
||||
|
||||
const ck::index_t Hi = input_spatial_lengths[0];
|
||||
const ck::index_t Wi = input_spatial_lengths[1];
|
||||
|
||||
const ck::index_t Ho = output_spatial_lengths[0];
|
||||
const ck::index_t Wo = output_spatial_lengths[1];
|
||||
|
||||
auto f_host_tensor_descriptor =
|
||||
[](std::size_t N_, std::size_t C_, std::size_t H, std::size_t W, auto layout) {
|
||||
if constexpr(is_same<decltype(layout), ck::tensor_layout::convolution::NCHW>::value ||
|
||||
is_same<decltype(layout), ck::tensor_layout::convolution::KCYX>::value ||
|
||||
is_same<decltype(layout), ck::tensor_layout::convolution::NKHW>::value)
|
||||
{
|
||||
return HostTensorDescriptor(std::vector<std::size_t>({N_, C_, H, W}),
|
||||
std::vector<std::size_t>({C_ * H * W, H * W, W, 1}));
|
||||
}
|
||||
else if constexpr(is_same<decltype(layout), tensor_layout::convolution::NHWC>::value ||
|
||||
is_same<decltype(layout), tensor_layout::convolution::KYXC>::value ||
|
||||
is_same<decltype(layout), tensor_layout::convolution::NHWK>::value)
|
||||
{
|
||||
return HostTensorDescriptor(std::vector<std::size_t>({N_, C_, H, W}),
|
||||
std::vector<std::size_t>({C_ * H * W, 1, W * C_, C_}));
|
||||
}
|
||||
};
|
||||
|
||||
Tensor<InDataType> in_n_c_hi_wi(f_host_tensor_descriptor(N, C, Hi, Wi, InLayout{}));
|
||||
Tensor<WeiDataType> wei_k_c_y_x(f_host_tensor_descriptor(K, C, Y, X, WeiLayout{}));
|
||||
Tensor<OutDataType> out_n_k_ho_wo_host_result(
|
||||
f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
Tensor<OutDataType> out_n_k_ho_wo_device_result(
|
||||
f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
|
||||
// bias: assume contiguous 1d vector
|
||||
Tensor<OutDataType> bias_k(
|
||||
HostTensorDescriptor(std::vector<std::size_t>({static_cast<std::size_t>(K)})));
|
||||
|
||||
std::cout << "in_n_c_hi_wi: " << in_n_c_hi_wi.mDesc << std::endl;
|
||||
std::cout << "wei_k_c_y_x: " << wei_k_c_y_x.mDesc << std::endl;
|
||||
std::cout << "out_n_k_ho_wo: " << out_n_k_ho_wo_host_result.mDesc << std::endl;
|
||||
std::cout << "bias_k: " << bias_k.mDesc << std::endl;
|
||||
|
||||
switch(init_method)
|
||||
{
|
||||
case 0: break;
|
||||
case 1:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_2<InDataType>{-5, 5});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_2<WeiDataType>{-5, 5});
|
||||
bias_k.GenerateTensorValue(GeneratorTensor_2<OutDataType>{-5, 5});
|
||||
break;
|
||||
default:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_3<InDataType>{0.0, 1.0});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_3<WeiDataType>{-0.5, 0.5});
|
||||
bias_k.GenerateTensorValue(GeneratorTensor_3<OutDataType>{0.0, 1.0});
|
||||
}
|
||||
|
||||
using InElementOp = ck::tensor_operation::element_wise::PassThrough;
|
||||
using WeiElementOp = ck::tensor_operation::element_wise::PassThrough;
|
||||
using OutElementOp = ck::tensor_operation::element_wise::AddRelu;
|
||||
|
||||
if(do_verification)
|
||||
{
|
||||
cpu_conv_bias_relu_atomic_add(in_n_c_hi_wi.mData.data(),
|
||||
wei_k_c_y_x.mData.data(),
|
||||
out_n_k_ho_wo_host_result.mData.data(),
|
||||
bias_k.mData.data(),
|
||||
N,
|
||||
K,
|
||||
C,
|
||||
Y,
|
||||
X,
|
||||
Hi,
|
||||
Wi,
|
||||
Ho,
|
||||
Wo,
|
||||
conv_filter_strides[0],
|
||||
conv_filter_dilations[0],
|
||||
input_left_pads[0]);
|
||||
}
|
||||
|
||||
DeviceMem in_device_buf(sizeof(InDataType) * in_n_c_hi_wi.mDesc.GetElementSpace());
|
||||
DeviceMem wei_device_buf(sizeof(WeiDataType) * wei_k_c_y_x.mDesc.GetElementSpace());
|
||||
DeviceMem out_device_buf(sizeof(OutDataType) *
|
||||
out_n_k_ho_wo_device_result.mDesc.GetElementSpace());
|
||||
DeviceMem bias_device_buf(sizeof(OutDataType) * bias_k.mDesc.GetElementSpace());
|
||||
|
||||
in_device_buf.ToDevice(in_n_c_hi_wi.mData.data());
|
||||
wei_device_buf.ToDevice(wei_k_c_y_x.mData.data());
|
||||
bias_device_buf.ToDevice(bias_k.mData.data());
|
||||
|
||||
using DeviceConvFwdBiasReluPtr = ck::tensor_operation::device::
|
||||
DeviceConvFwdBiasActivationPtr<InElementOp, WeiElementOp, OutElementOp>;
|
||||
|
||||
// add device operator instances
|
||||
std::vector<DeviceConvFwdBiasReluPtr> op_ptrs;
|
||||
|
||||
if constexpr(ck::is_same_v<ck::remove_cv_t<InDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<WeiDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<OutDataType>, ck::half_t>)
|
||||
{
|
||||
ck::tensor_operation::device::device_conv2d_fwd_bias_activation_atomic_add_instance::
|
||||
add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instances(
|
||||
op_ptrs);
|
||||
}
|
||||
|
||||
if(op_ptrs.size() <= 0)
|
||||
{
|
||||
throw std::runtime_error("wrong! no device Conv instance found");
|
||||
}
|
||||
|
||||
std::string best_conv_name;
|
||||
float best_ave_time = 0;
|
||||
float best_tflops = 0;
|
||||
float best_gb_per_sec = 0;
|
||||
|
||||
// profile device Conv instances
|
||||
for(auto& op_ptr : op_ptrs)
|
||||
{
|
||||
auto argument_ptr = op_ptr->MakeArgumentPointer(
|
||||
static_cast<const InDataType*>(in_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const WeiDataType*>(wei_device_buf.GetDeviceBuffer()),
|
||||
static_cast<OutDataType*>(out_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const OutDataType*>(bias_device_buf.GetDeviceBuffer()),
|
||||
N,
|
||||
K,
|
||||
C,
|
||||
input_spatial_lengths,
|
||||
filter_spatial_lengths,
|
||||
output_spatial_lengths,
|
||||
conv_filter_strides,
|
||||
conv_filter_dilations,
|
||||
input_left_pads,
|
||||
input_right_pads,
|
||||
InElementOp{},
|
||||
WeiElementOp{},
|
||||
OutElementOp{});
|
||||
|
||||
auto invoker_ptr = op_ptr->MakeInvokerPointer();
|
||||
|
||||
if(op_ptr->IsSupportedArgument(argument_ptr.get()))
|
||||
{
|
||||
std::string conv_name = op_ptr->GetTypeString();
|
||||
|
||||
float ave_time = invoker_ptr->Run(argument_ptr.get(), nrepeat);
|
||||
|
||||
std::size_t flop = std::size_t(2) * N * K * Ho * Wo * C * Y * X;
|
||||
|
||||
std::size_t num_btype =
|
||||
sizeof(InDataType) * (N * C * Hi * Wi) + sizeof(WeiDataType) * (K * C * Y * X) +
|
||||
sizeof(OutDataType) * (N * K * Ho * Wo) + sizeof(OutDataType) * (K);
|
||||
|
||||
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, " << conv_name << std::endl;
|
||||
|
||||
if(tflops > best_tflops)
|
||||
{
|
||||
best_conv_name = conv_name;
|
||||
best_tflops = tflops;
|
||||
best_ave_time = ave_time;
|
||||
best_gb_per_sec = gb_per_sec;
|
||||
}
|
||||
|
||||
if(do_verification)
|
||||
{
|
||||
out_device_buf.FromDevice(out_n_k_ho_wo_device_result.mData.data());
|
||||
|
||||
check_error(out_n_k_ho_wo_host_result, out_n_k_ho_wo_device_result);
|
||||
|
||||
if(do_log)
|
||||
{
|
||||
LogRangeAsType<float>(std::cout << "in : ", in_n_c_hi_wi.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(std::cout << "wei: ", wei_k_c_y_x.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(
|
||||
std::cout << "out_host : ", out_n_k_ho_wo_host_result.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(
|
||||
std::cout << "out_device: ", out_n_k_ho_wo_device_result.mData, ",")
|
||||
<< std::endl;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
std::cout << "Best Perf: " << best_ave_time << " ms, " << best_tflops << " TFlops, "
|
||||
<< best_gb_per_sec << " GB/s, " << best_conv_name << std::endl;
|
||||
}
|
||||
|
||||
} // namespace profiler
|
||||
} // namespace ck
|
||||
327
profiler/include/profile_conv_fwd_bias_relu_impl.hpp
Normal file
327
profiler/include/profile_conv_fwd_bias_relu_impl.hpp
Normal file
@@ -0,0 +1,327 @@
|
||||
#pragma once
|
||||
#include "config.hpp"
|
||||
#include "device.hpp"
|
||||
#include "host_tensor.hpp"
|
||||
#include "host_tensor_generator.hpp"
|
||||
#include "host_conv.hpp"
|
||||
#include "tensor_layout.hpp"
|
||||
#include "device_tensor.hpp"
|
||||
#include "device_conv_fwd_bias_activation.hpp"
|
||||
#include "element_wise_operation.hpp"
|
||||
|
||||
namespace ck {
|
||||
namespace tensor_operation {
|
||||
namespace device {
|
||||
namespace device_conv2d_fwd_bias_activation_instance {
|
||||
|
||||
using DeviceConvFwdBiasReluPtr =
|
||||
DeviceConvFwdBiasActivationPtr<ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::AddRelu>;
|
||||
|
||||
void add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instances(
|
||||
std::vector<DeviceConvFwdBiasReluPtr>&);
|
||||
|
||||
} // namespace device_conv2d_fwd_bias_activation_instance
|
||||
} // namespace device
|
||||
} // namespace tensor_operation
|
||||
} // namespace ck
|
||||
|
||||
namespace ck {
|
||||
namespace profiler {
|
||||
|
||||
void cpu_conv_bias_relu(ck::half_t* in_ptr,
|
||||
ck::half_t* weight_ptr,
|
||||
ck::half_t* output_ptr,
|
||||
ck::half_t* bias_ptr,
|
||||
const ck::index_t N,
|
||||
const ck::index_t K,
|
||||
const ck::index_t C,
|
||||
const ck::index_t Y,
|
||||
const ck::index_t X,
|
||||
const ck::index_t Hi,
|
||||
const ck::index_t Wi,
|
||||
const ck::index_t Ho,
|
||||
const ck::index_t Wo,
|
||||
const ck::index_t Stride,
|
||||
const ck::index_t Dilation,
|
||||
const ck::index_t Pad)
|
||||
{
|
||||
|
||||
const auto in_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(N),
|
||||
static_cast<std::size_t>(Hi),
|
||||
static_cast<std::size_t>(Wi),
|
||||
static_cast<std::size_t>(C)});
|
||||
const auto wei_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(K),
|
||||
static_cast<std::size_t>(Y),
|
||||
static_cast<std::size_t>(X),
|
||||
static_cast<std::size_t>(C)});
|
||||
const auto out_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(N),
|
||||
static_cast<std::size_t>(Ho),
|
||||
static_cast<std::size_t>(Wo),
|
||||
static_cast<std::size_t>(K)});
|
||||
const auto bias_desc =
|
||||
HostTensorDescriptor(std::vector<std::size_t>{static_cast<std::size_t>(K)});
|
||||
|
||||
auto f_k = [&](auto k) {
|
||||
for(int n = 0; n < N; ++n)
|
||||
{
|
||||
for(int ho = 0; ho < Ho; ++ho)
|
||||
{
|
||||
for(int wo = 0; wo < Wo; ++wo)
|
||||
{
|
||||
double v = 0;
|
||||
for(int c = 0; c < C; ++c)
|
||||
{
|
||||
for(int y = 0; y < Y; ++y)
|
||||
{
|
||||
int hi = ho * Stride + y * Dilation - Pad;
|
||||
for(int x = 0; x < X; ++x)
|
||||
{
|
||||
int wi = wo * Stride + x * Dilation - Pad;
|
||||
if(hi >= 0 && hi < Hi && wi >= 0 && wi < Wi)
|
||||
{
|
||||
double in =
|
||||
in_ptr[in_desc.GetOffsetFromMultiIndex(n, hi, wi, c)];
|
||||
double wei =
|
||||
weight_ptr[wei_desc.GetOffsetFromMultiIndex(k, y, x, c)];
|
||||
|
||||
v += in * wei;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
v += bias_ptr[bias_desc.GetOffsetFromMultiIndex(k)];
|
||||
|
||||
v = v > 0 ? v : 0;
|
||||
|
||||
output_ptr[out_desc.GetOffsetFromMultiIndex(n, ho, wo, k)] = v;
|
||||
}
|
||||
}
|
||||
}
|
||||
};
|
||||
|
||||
make_ParallelTensorFunctor(f_k, K)(std::thread::hardware_concurrency());
|
||||
}
|
||||
|
||||
template <int NDimSpatial,
|
||||
typename InDataType,
|
||||
typename WeiDataType,
|
||||
typename OutDataType,
|
||||
typename InLayout,
|
||||
typename WeiLayout,
|
||||
typename OutLayout>
|
||||
void profile_conv_fwd_bias_relu_impl(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
ck::index_t N,
|
||||
ck::index_t K,
|
||||
ck::index_t C,
|
||||
std::vector<ck::index_t> input_spatial_lengths,
|
||||
std::vector<ck::index_t> filter_spatial_lengths,
|
||||
std::vector<ck::index_t> output_spatial_lengths,
|
||||
std::vector<ck::index_t> conv_filter_strides,
|
||||
std::vector<ck::index_t> conv_filter_dilations,
|
||||
std::vector<ck::index_t> input_left_pads,
|
||||
std::vector<ck::index_t> input_right_pads)
|
||||
{
|
||||
const ck::index_t Y = filter_spatial_lengths[0];
|
||||
const ck::index_t X = filter_spatial_lengths[1];
|
||||
|
||||
const ck::index_t Hi = input_spatial_lengths[0];
|
||||
const ck::index_t Wi = input_spatial_lengths[1];
|
||||
|
||||
const ck::index_t Ho = output_spatial_lengths[0];
|
||||
const ck::index_t Wo = output_spatial_lengths[1];
|
||||
|
||||
auto f_host_tensor_descriptor =
|
||||
[](std::size_t N_, std::size_t C_, std::size_t H, std::size_t W, auto layout) {
|
||||
if constexpr(is_same<decltype(layout), ck::tensor_layout::convolution::NCHW>::value ||
|
||||
is_same<decltype(layout), ck::tensor_layout::convolution::KCYX>::value ||
|
||||
is_same<decltype(layout), ck::tensor_layout::convolution::NKHW>::value)
|
||||
{
|
||||
return HostTensorDescriptor(std::vector<std::size_t>({N_, C_, H, W}),
|
||||
std::vector<std::size_t>({C_ * H * W, H * W, W, 1}));
|
||||
}
|
||||
else if constexpr(is_same<decltype(layout), tensor_layout::convolution::NHWC>::value ||
|
||||
is_same<decltype(layout), tensor_layout::convolution::KYXC>::value ||
|
||||
is_same<decltype(layout), tensor_layout::convolution::NHWK>::value)
|
||||
{
|
||||
return HostTensorDescriptor(std::vector<std::size_t>({N_, C_, H, W}),
|
||||
std::vector<std::size_t>({C_ * H * W, 1, W * C_, C_}));
|
||||
}
|
||||
};
|
||||
|
||||
Tensor<InDataType> in_n_c_hi_wi(f_host_tensor_descriptor(N, C, Hi, Wi, InLayout{}));
|
||||
Tensor<WeiDataType> wei_k_c_y_x(f_host_tensor_descriptor(K, C, Y, X, WeiLayout{}));
|
||||
Tensor<OutDataType> out_n_k_ho_wo_host_result(
|
||||
f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
Tensor<OutDataType> out_n_k_ho_wo_device_result(
|
||||
f_host_tensor_descriptor(N, K, Ho, Wo, OutLayout{}));
|
||||
|
||||
// bias: assume contiguous 1d vector
|
||||
Tensor<OutDataType> bias_k(
|
||||
HostTensorDescriptor(std::vector<std::size_t>({static_cast<std::size_t>(K)})));
|
||||
|
||||
std::cout << "in_n_c_hi_wi: " << in_n_c_hi_wi.mDesc << std::endl;
|
||||
std::cout << "wei_k_c_y_x: " << wei_k_c_y_x.mDesc << std::endl;
|
||||
std::cout << "out_n_k_ho_wo: " << out_n_k_ho_wo_host_result.mDesc << std::endl;
|
||||
std::cout << "bias_k: " << bias_k.mDesc << std::endl;
|
||||
|
||||
switch(init_method)
|
||||
{
|
||||
case 0: break;
|
||||
case 1:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_2<InDataType>{-5, 5});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_2<WeiDataType>{-5, 5});
|
||||
bias_k.GenerateTensorValue(GeneratorTensor_2<OutDataType>{-5, 5});
|
||||
break;
|
||||
default:
|
||||
in_n_c_hi_wi.GenerateTensorValue(GeneratorTensor_3<InDataType>{0.0, 1.0});
|
||||
wei_k_c_y_x.GenerateTensorValue(GeneratorTensor_3<WeiDataType>{-0.5, 0.5});
|
||||
bias_k.GenerateTensorValue(GeneratorTensor_3<OutDataType>{0.0, 1.0});
|
||||
}
|
||||
|
||||
using InElementOp = ck::tensor_operation::element_wise::PassThrough;
|
||||
using WeiElementOp = ck::tensor_operation::element_wise::PassThrough;
|
||||
using OutElementOp = ck::tensor_operation::element_wise::AddRelu;
|
||||
|
||||
if(do_verification)
|
||||
{
|
||||
cpu_conv_bias_relu(in_n_c_hi_wi.mData.data(),
|
||||
wei_k_c_y_x.mData.data(),
|
||||
out_n_k_ho_wo_host_result.mData.data(),
|
||||
bias_k.mData.data(),
|
||||
N,
|
||||
K,
|
||||
C,
|
||||
Y,
|
||||
X,
|
||||
Hi,
|
||||
Wi,
|
||||
Ho,
|
||||
Wo,
|
||||
conv_filter_strides[0],
|
||||
conv_filter_dilations[0],
|
||||
input_left_pads[0]);
|
||||
}
|
||||
|
||||
DeviceMem in_device_buf(sizeof(InDataType) * in_n_c_hi_wi.mDesc.GetElementSpace());
|
||||
DeviceMem wei_device_buf(sizeof(WeiDataType) * wei_k_c_y_x.mDesc.GetElementSpace());
|
||||
DeviceMem out_device_buf(sizeof(OutDataType) *
|
||||
out_n_k_ho_wo_device_result.mDesc.GetElementSpace());
|
||||
DeviceMem bias_device_buf(sizeof(OutDataType) * bias_k.mDesc.GetElementSpace());
|
||||
|
||||
in_device_buf.ToDevice(in_n_c_hi_wi.mData.data());
|
||||
wei_device_buf.ToDevice(wei_k_c_y_x.mData.data());
|
||||
bias_device_buf.ToDevice(bias_k.mData.data());
|
||||
|
||||
using DeviceConvFwdBiasReluPtr = ck::tensor_operation::device::
|
||||
DeviceConvFwdBiasActivationPtr<InElementOp, WeiElementOp, OutElementOp>;
|
||||
|
||||
// add device operator instances
|
||||
std::vector<DeviceConvFwdBiasReluPtr> op_ptrs;
|
||||
|
||||
if constexpr(ck::is_same_v<ck::remove_cv_t<InDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<WeiDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<OutDataType>, ck::half_t>)
|
||||
{
|
||||
ck::tensor_operation::device::device_conv2d_fwd_bias_activation_instance::
|
||||
add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instances(op_ptrs);
|
||||
}
|
||||
|
||||
if(op_ptrs.size() <= 0)
|
||||
{
|
||||
throw std::runtime_error("wrong! no device Conv instance found");
|
||||
}
|
||||
|
||||
std::string best_conv_name;
|
||||
float best_ave_time = 0;
|
||||
float best_tflops = 0;
|
||||
float best_gb_per_sec = 0;
|
||||
|
||||
// profile device Conv instances
|
||||
for(auto& op_ptr : op_ptrs)
|
||||
{
|
||||
auto argument_ptr = op_ptr->MakeArgumentPointer(
|
||||
static_cast<const InDataType*>(in_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const WeiDataType*>(wei_device_buf.GetDeviceBuffer()),
|
||||
static_cast<OutDataType*>(out_device_buf.GetDeviceBuffer()),
|
||||
static_cast<const OutDataType*>(bias_device_buf.GetDeviceBuffer()),
|
||||
N,
|
||||
K,
|
||||
C,
|
||||
input_spatial_lengths,
|
||||
filter_spatial_lengths,
|
||||
output_spatial_lengths,
|
||||
conv_filter_strides,
|
||||
conv_filter_dilations,
|
||||
input_left_pads,
|
||||
input_right_pads,
|
||||
InElementOp{},
|
||||
WeiElementOp{},
|
||||
OutElementOp{});
|
||||
|
||||
auto invoker_ptr = op_ptr->MakeInvokerPointer();
|
||||
|
||||
if(op_ptr->IsSupportedArgument(argument_ptr.get()))
|
||||
{
|
||||
std::string conv_name = op_ptr->GetTypeString();
|
||||
|
||||
float ave_time = invoker_ptr->Run(argument_ptr.get(), nrepeat);
|
||||
|
||||
std::size_t flop = std::size_t(2) * N * K * Ho * Wo * C * Y * X;
|
||||
|
||||
std::size_t num_btype =
|
||||
sizeof(InDataType) * (N * C * Hi * Wi) + sizeof(WeiDataType) * (K * C * Y * X) +
|
||||
sizeof(OutDataType) * (N * K * Ho * Wo) + sizeof(OutDataType) * (K);
|
||||
|
||||
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, " << conv_name << std::endl;
|
||||
|
||||
if(tflops > best_tflops)
|
||||
{
|
||||
best_conv_name = conv_name;
|
||||
best_tflops = tflops;
|
||||
best_ave_time = ave_time;
|
||||
best_gb_per_sec = gb_per_sec;
|
||||
}
|
||||
|
||||
if(do_verification)
|
||||
{
|
||||
out_device_buf.FromDevice(out_n_k_ho_wo_device_result.mData.data());
|
||||
|
||||
check_error(out_n_k_ho_wo_host_result, out_n_k_ho_wo_device_result);
|
||||
|
||||
if(do_log)
|
||||
{
|
||||
LogRangeAsType<float>(std::cout << "in : ", in_n_c_hi_wi.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(std::cout << "wei: ", wei_k_c_y_x.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(
|
||||
std::cout << "out_host : ", out_n_k_ho_wo_host_result.mData, ",")
|
||||
<< std::endl;
|
||||
LogRangeAsType<float>(
|
||||
std::cout << "out_device: ", out_n_k_ho_wo_device_result.mData, ",")
|
||||
<< std::endl;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
std::cout << "Best Perf: " << best_ave_time << " ms, " << best_tflops << " TFlops, "
|
||||
<< best_gb_per_sec << " GB/s, " << best_conv_name << std::endl;
|
||||
}
|
||||
|
||||
} // namespace profiler
|
||||
} // namespace ck
|
||||
@@ -6,40 +6,26 @@
|
||||
#include "host_conv.hpp"
|
||||
#include "tensor_layout.hpp"
|
||||
#include "device_tensor.hpp"
|
||||
#include "device_conv.hpp"
|
||||
#include "device_conv_instance.hpp"
|
||||
#include "device_conv_fwd.hpp"
|
||||
#include "element_wise_operation.hpp"
|
||||
|
||||
namespace ck {
|
||||
namespace tensor_operation {
|
||||
namespace device {
|
||||
namespace device_conv_instance {
|
||||
namespace device_conv2d_fwd_instance {
|
||||
|
||||
using DeviceConvFwdNoOpPtr = DeviceConvFwdPtr<ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::PassThrough,
|
||||
ck::tensor_operation::element_wise::PassThrough>;
|
||||
|
||||
template <>
|
||||
void add_device_conv_fwd_instance<2,
|
||||
float,
|
||||
float,
|
||||
float,
|
||||
ck::tensor_layout::convolution::NHWC,
|
||||
ck::tensor_layout::convolution::KYXC,
|
||||
ck::tensor_layout::convolution::NHWK>(
|
||||
void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances(std::vector<DeviceConvFwdNoOpPtr>&);
|
||||
|
||||
void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances(std::vector<DeviceConvFwdNoOpPtr>&);
|
||||
|
||||
void add_device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances(
|
||||
std::vector<DeviceConvFwdNoOpPtr>&);
|
||||
|
||||
template <>
|
||||
void add_device_conv_fwd_instance<2,
|
||||
ck::half_t,
|
||||
ck::half_t,
|
||||
ck::half_t,
|
||||
ck::tensor_layout::convolution::NHWC,
|
||||
ck::tensor_layout::convolution::KYXC,
|
||||
ck::tensor_layout::convolution::NHWK>(
|
||||
std::vector<DeviceConvFwdNoOpPtr>&);
|
||||
|
||||
} // namespace device_conv_instance
|
||||
} // namespace device_conv2d_fwd_instance
|
||||
} // namespace device
|
||||
} // namespace tensor_operation
|
||||
} // namespace ck
|
||||
@@ -54,20 +40,20 @@ template <int NDimSpatial,
|
||||
typename InLayout,
|
||||
typename WeiLayout,
|
||||
typename OutLayout>
|
||||
void profile_conv(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
ck::index_t N,
|
||||
ck::index_t K,
|
||||
ck::index_t C,
|
||||
std::vector<ck::index_t> input_spatial_lengths,
|
||||
std::vector<ck::index_t> filter_spatial_lengths,
|
||||
std::vector<ck::index_t> output_spatial_lengths,
|
||||
std::vector<ck::index_t> conv_filter_strides,
|
||||
std::vector<ck::index_t> conv_filter_dilations,
|
||||
std::vector<ck::index_t> input_left_pads,
|
||||
std::vector<ck::index_t> input_right_pads)
|
||||
void profile_conv_fwd_impl(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
ck::index_t N,
|
||||
ck::index_t K,
|
||||
ck::index_t C,
|
||||
std::vector<ck::index_t> input_spatial_lengths,
|
||||
std::vector<ck::index_t> filter_spatial_lengths,
|
||||
std::vector<ck::index_t> output_spatial_lengths,
|
||||
std::vector<ck::index_t> conv_filter_strides,
|
||||
std::vector<ck::index_t> conv_filter_dilations,
|
||||
std::vector<ck::index_t> input_left_pads,
|
||||
std::vector<ck::index_t> input_right_pads)
|
||||
{
|
||||
const ck::index_t Y = filter_spatial_lengths[0];
|
||||
const ck::index_t X = filter_spatial_lengths[1];
|
||||
@@ -146,20 +132,30 @@ void profile_conv(int do_verification,
|
||||
// add device Conv instances
|
||||
std::vector<DeviceConvFwdNoOpPtr> conv_ptrs;
|
||||
|
||||
ck::tensor_operation::device::device_conv_instance::add_device_conv_fwd_instance<2,
|
||||
InDataType,
|
||||
WeiDataType,
|
||||
OutDataType,
|
||||
InLayout,
|
||||
WeiLayout,
|
||||
OutLayout>(
|
||||
conv_ptrs);
|
||||
if constexpr(ck::is_same_v<ck::remove_cv_t<InDataType>, float> &&
|
||||
ck::is_same_v<ck::remove_cv_t<WeiDataType>, float> &&
|
||||
ck::is_same_v<ck::remove_cv_t<OutDataType>, float>)
|
||||
{
|
||||
ck::tensor_operation::device::device_conv2d_fwd_instance::
|
||||
add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances(conv_ptrs);
|
||||
}
|
||||
else if constexpr(ck::is_same_v<ck::remove_cv_t<InDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<WeiDataType>, ck::half_t> &&
|
||||
ck::is_same_v<ck::remove_cv_t<OutDataType>, ck::half_t>)
|
||||
{
|
||||
ck::tensor_operation::device::device_conv2d_fwd_instance::
|
||||
add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances(conv_ptrs);
|
||||
|
||||
ck::tensor_operation::device::device_conv2d_fwd_instance::
|
||||
add_device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances(conv_ptrs);
|
||||
}
|
||||
|
||||
if(conv_ptrs.size() <= 0)
|
||||
{
|
||||
throw std::runtime_error("wrong! no device Conv instance found");
|
||||
}
|
||||
|
||||
std::string best_conv_name;
|
||||
float best_ave_time = 0;
|
||||
float best_tflops = 0;
|
||||
float best_gb_per_sec = 0;
|
||||
@@ -189,6 +185,8 @@ void profile_conv(int do_verification,
|
||||
|
||||
if(conv_ptr->IsSupportedArgument(argument_ptr.get()))
|
||||
{
|
||||
std::string conv_name = conv_ptr->GetTypeString();
|
||||
|
||||
float ave_time = invoker_ptr->Run(argument_ptr.get(), nrepeat);
|
||||
|
||||
std::size_t flop = std::size_t(2) * N * K * Ho * Wo * C * Y * X;
|
||||
@@ -202,10 +200,11 @@ void profile_conv(int do_verification,
|
||||
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;
|
||||
<< " GB/s, " << conv_name << std::endl;
|
||||
|
||||
if(tflops > best_tflops)
|
||||
{
|
||||
best_conv_name = conv_name;
|
||||
best_tflops = tflops;
|
||||
best_ave_time = ave_time;
|
||||
best_gb_per_sec = gb_per_sec;
|
||||
@@ -235,7 +234,7 @@ void profile_conv(int do_verification,
|
||||
}
|
||||
|
||||
std::cout << "Best Perf: " << best_ave_time << " ms, " << best_tflops << " TFlops, "
|
||||
<< best_gb_per_sec << " GB/s" << std::endl;
|
||||
<< best_gb_per_sec << " GB/s, " << best_conv_name << std::endl;
|
||||
}
|
||||
|
||||
} // namespace profiler
|
||||
@@ -88,16 +88,16 @@ template <typename ADataType,
|
||||
typename ALayout,
|
||||
typename BLayout,
|
||||
typename CLayout>
|
||||
void profile_gemm(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
int M,
|
||||
int N,
|
||||
int K,
|
||||
int StrideA,
|
||||
int StrideB,
|
||||
int StrideC)
|
||||
void profile_gemm_impl(int do_verification,
|
||||
int init_method,
|
||||
bool do_log,
|
||||
int nrepeat,
|
||||
int M,
|
||||
int N,
|
||||
int K,
|
||||
int StrideA,
|
||||
int StrideB,
|
||||
int StrideC)
|
||||
{
|
||||
auto f_host_tensor_descriptor =
|
||||
[](std::size_t row, std::size_t col, std::size_t stride, auto layout) {
|
||||
@@ -164,6 +164,7 @@ void profile_gemm(int do_verification,
|
||||
throw std::runtime_error("wrong! no device GEMM instance found");
|
||||
}
|
||||
|
||||
std::string best_gemm_name;
|
||||
float best_ave_time = 0;
|
||||
float best_tflops = 0;
|
||||
float best_gb_per_sec = 0;
|
||||
@@ -189,9 +190,12 @@ void profile_gemm(int do_verification,
|
||||
|
||||
if(gemm_ptr->IsSupportedArgument(argument_ptr.get()))
|
||||
{
|
||||
std::string gemm_name = gemm_ptr->GetTypeString();
|
||||
|
||||
float ave_time = invoker_ptr->Run(argument_ptr.get(), nrepeat);
|
||||
|
||||
std::size_t flop = std::size_t(2) * M * N * K;
|
||||
|
||||
std::size_t num_btype =
|
||||
sizeof(ADataType) * M * K + sizeof(BDataType) * K * M + sizeof(CDataType) * M * N;
|
||||
|
||||
@@ -200,10 +204,11 @@ void profile_gemm(int do_verification,
|
||||
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;
|
||||
<< " GB/s, " << gemm_name << std::endl;
|
||||
|
||||
if(tflops > best_tflops)
|
||||
{
|
||||
best_gemm_name = gemm_name;
|
||||
best_tflops = tflops;
|
||||
best_ave_time = ave_time;
|
||||
best_gb_per_sec = gb_per_sec;
|
||||
@@ -234,7 +239,7 @@ void profile_gemm(int do_verification,
|
||||
}
|
||||
|
||||
std::cout << "Best Perf: " << best_ave_time << " ms, " << best_tflops << " TFlops, "
|
||||
<< best_gb_per_sec << " GB/s" << std::endl;
|
||||
<< best_gb_per_sec << " GB/s, " << best_gemm_name << std::endl;
|
||||
}
|
||||
|
||||
} // namespace profiler
|
||||
Reference in New Issue
Block a user