wehub-resource-sync
283 lines
10 KiB
Plaintext
283 lines
10 KiB
Plaintext
// Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
|
|
//
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
#include "paddle/common/enforce.h"
|
|
#include "paddle/phi/core/dense_tensor.h"
|
|
#include "paddle/phi/core/kernel_registry.h"
|
|
#include "paddle/phi/kernels/fusion/gpu/fused_dropout_add_utils.h"
|
|
|
|
#include "paddle/phi/kernels/funcs/dropout_impl_util.h"
|
|
#include "paddle/phi/kernels/funcs/functors.h"
|
|
#include "paddle/phi/kernels/primitive/compute_primitives.h"
|
|
|
|
#include "paddle/phi/kernels/funcs/dropout_impl.cu.h"
|
|
|
|
namespace phi {
|
|
namespace fusion {
|
|
|
|
template <typename T1, typename T2 = T1, typename OutT = T1>
|
|
struct NoMaskFwFunctor {
|
|
const float retain_prob_;
|
|
const bool is_upscale_in_train_;
|
|
using MT = typename phi::dtype::MPTypeTrait<T1>::Type;
|
|
MT factor;
|
|
HOSTDEVICE inline NoMaskFwFunctor(const float retain_prob,
|
|
const bool is_upscale_in_train)
|
|
: retain_prob_(retain_prob), is_upscale_in_train_(is_upscale_in_train) {
|
|
factor = static_cast<MT>(1.0f / retain_prob_);
|
|
}
|
|
|
|
HOSTDEVICE inline void operator()(OutT* dst,
|
|
const T1* src_val,
|
|
const T2* rand,
|
|
int num) const {
|
|
static constexpr int kCount =
|
|
funcs::uniform_distribution<T2>::kReturnsCount;
|
|
#pragma unroll
|
|
for (int i = 0; i < kCount; i++) {
|
|
if (rand[i] < retain_prob_) {
|
|
dst[i] = is_upscale_in_train_
|
|
? static_cast<T1>(static_cast<MT>(src_val[i]) * factor)
|
|
: static_cast<T1>(src_val[i]);
|
|
dst[i] += src_val[i + kCount];
|
|
} else {
|
|
dst[i] = src_val[i + kCount];
|
|
}
|
|
}
|
|
}
|
|
};
|
|
|
|
template <typename T>
|
|
struct ScaleAddFuctor {
|
|
using MT = typename phi::dtype::MPTypeTrait<T>::Type;
|
|
explicit ScaleAddFuctor(const MT factor, bool upscale_in_train)
|
|
: factor_(factor), upscale_in_train_(upscale_in_train) {}
|
|
|
|
__device__ __forceinline__ T operator()(const T src, const T res) const {
|
|
return upscale_in_train_
|
|
? src + res
|
|
: static_cast<T>(static_cast<MT>(src) * factor_) + res;
|
|
}
|
|
|
|
private:
|
|
MT factor_;
|
|
bool upscale_in_train_;
|
|
};
|
|
|
|
template <typename T, typename Functor>
|
|
__global__ void VectorizedDropoutForward(
|
|
/* This is used to relate kernel to cudaGraph nodes*/
|
|
unsigned int identifier,
|
|
const size_t n,
|
|
uint64_t seed,
|
|
const T* src,
|
|
const T* res,
|
|
T* dst,
|
|
uint64_t increment,
|
|
size_t main_offset,
|
|
Functor functor) {
|
|
size_t idx = static_cast<size_t>(BLOCK_ID_X * BLOCK_NUM_X);
|
|
static constexpr int kCount =
|
|
funcs::uniform_distribution<float>::kReturnsCount;
|
|
size_t stride = BLOCK_NUM_X * GRID_NUM_X * kCount;
|
|
#ifdef PADDLE_WITH_HIP
|
|
hiprandStatePhilox4_32_10_t state;
|
|
hiprand_init(seed, idx + THREAD_ID_X, increment, &state);
|
|
using SType = hiprandStatePhilox4_32_10_t;
|
|
#else
|
|
curandStatePhilox4_32_10_t state;
|
|
curand_init(seed, idx + THREAD_ID_X, increment, &state);
|
|
using SType = curandStatePhilox4_32_10_t;
|
|
#endif
|
|
T dst_res[kCount * 2];
|
|
float rands[kCount];
|
|
|
|
using Rand = funcs::uniform_distribution<float>;
|
|
int deal_size = BLOCK_NUM_X * kCount;
|
|
size_t fix = idx * kCount;
|
|
|
|
for (; fix < main_offset; fix += stride) {
|
|
kps::ReadData<T, kCount, 1, false>(&dst_res[0], src + fix, deal_size);
|
|
kps::ReadData<T, kCount, 1, false>(&dst_res[kCount], res + fix, deal_size);
|
|
|
|
kps::ElementwiseRandom<SType, float, kCount, Rand>(
|
|
&rands[0], Rand(), &state);
|
|
|
|
// dst
|
|
kps::OperatorTernary<T, float, T, Functor>(
|
|
&dst_res[0], &dst_res[0], &rands[0], functor, kCount);
|
|
|
|
kps::WriteData<T, kCount, 1, false>(dst + fix, &dst_res[0], deal_size);
|
|
|
|
if (fix > idx * kCount + 1) {
|
|
__syncthreads();
|
|
}
|
|
}
|
|
|
|
int remainder = n - fix;
|
|
if (remainder > 0) {
|
|
kps::ReadData<T, kCount, 1, true>(&dst_res[0], src + fix, remainder);
|
|
kps::ReadData<T, kCount, 1, true>(&dst_res[kCount], res + fix, remainder);
|
|
kps::ElementwiseRandom<SType, float, kCount, Rand>(
|
|
&rands[0], Rand(), &state);
|
|
// dst
|
|
kps::OperatorTernary<T, float, T, Functor>(
|
|
&dst_res[0], &dst_res[0], &rands[0], functor, kCount);
|
|
kps::WriteData<T, kCount, 1, true>(dst + fix, &dst_res[0], remainder);
|
|
__syncthreads();
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void FusedDropoutAddKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& y,
|
|
const optional<DenseTensor>& seed_tensor,
|
|
const Scalar& p,
|
|
bool is_test,
|
|
const std::string& mode,
|
|
int seed,
|
|
bool fix_seed,
|
|
DenseTensor* out,
|
|
DenseTensor* seed_offset) {
|
|
auto* out_data = dev_ctx.template Alloc<T>(out);
|
|
auto* seed_offset_data = dev_ctx.template HostAlloc<int64_t>(seed_offset);
|
|
int64_t numel = x.numel();
|
|
auto stream = dev_ctx.stream();
|
|
bool upscale_in_train = (mode == "upscale_in_train");
|
|
|
|
const auto* x_data = x.data<T>();
|
|
const auto* y_data = y.data<T>();
|
|
float dropout_rate = p.to<float>();
|
|
|
|
if (!is_test) {
|
|
if (dropout_rate == 1.0f) {
|
|
phi::Copy(dev_ctx, y, dev_ctx.GetPlace(), false, out);
|
|
return;
|
|
}
|
|
uint64_t seed_data;
|
|
uint64_t increment;
|
|
auto random_prop = GetRandomCudaProp(numel, dev_ctx);
|
|
size_t grid_size_64 = random_prop[0];
|
|
size_t block_size_64 = random_prop[1];
|
|
PADDLE_ENFORCE_LE_UINT32_MAX(grid_size_64, "fused_dropout_add grid.x");
|
|
PADDLE_ENFORCE_LE_UINT32_MAX(block_size_64, "fused_dropout_add block.x");
|
|
uint32_t grid_size = static_cast<uint32_t>(grid_size_64);
|
|
uint32_t block_size = static_cast<uint32_t>(block_size_64);
|
|
size_t offset = random_prop[2];
|
|
size_t main_offset = random_prop[3];
|
|
auto seed_tensor_ptr = seed_tensor.get_ptr();
|
|
funcs::GetSeedDataAndIncrement(dev_ctx,
|
|
seed_tensor_ptr,
|
|
fix_seed,
|
|
seed,
|
|
offset,
|
|
&seed_data,
|
|
&increment);
|
|
seed_offset_data[0] = static_cast<int64_t>(seed_data);
|
|
seed_offset_data[1] = static_cast<int64_t>(increment);
|
|
|
|
auto dst_functor =
|
|
NoMaskFwFunctor<T, float>(1.0f - dropout_rate, upscale_in_train);
|
|
|
|
// we assume seed/offset is same across iterations
|
|
// seed_offset_data should preserved by cudaGraph pool
|
|
const GPUContext* dev_ctx_p = &dev_ctx;
|
|
|
|
// seed_offset_data should preserved by cudaGraph pool
|
|
auto gen_cuda = dev_ctx.GetGenerator();
|
|
auto state_index = gen_cuda->GetStateIndex();
|
|
auto parameterSetter = [dev_ctx_p,
|
|
offset,
|
|
seed_offset_data,
|
|
state_index,
|
|
seed_tensor_ptr,
|
|
fix_seed](backends::gpu::gpuKernelParams& params) {
|
|
if (!fix_seed) {
|
|
auto gen_cuda = dev_ctx_p->GetGenerator();
|
|
// ensure the generator use correct state index
|
|
gen_cuda->SetStateIndex(state_index);
|
|
|
|
// we assume seed is null pointer
|
|
// seed copy to cpu is meaningless here
|
|
assert(seed_tensor_ptr == nullptr);
|
|
|
|
uint64_t seed, increment;
|
|
std::tie(seed, increment) = gen_cuda->IncrementOffset(offset);
|
|
VLOG(10) << "CUDA_GRAPH seed = " << seed
|
|
<< ", increment = " << increment;
|
|
|
|
params.As<uint64_t>(2) = seed;
|
|
params.As<uint64_t>(6) = increment;
|
|
|
|
seed_offset_data[0] = static_cast<int64_t>(seed);
|
|
seed_offset_data[1] = static_cast<int64_t>(increment);
|
|
}
|
|
};
|
|
backends::gpu::CUDAGraphNodeLauncher::gpuKernelCallback_t
|
|
cudaKernelCallback = [=](unsigned int id) {
|
|
void* functionPtr = reinterpret_cast<void*>(
|
|
&(VectorizedDropoutForward<T, NoMaskFwFunctor<T, float>>));
|
|
#ifdef PADDLE_WITH_HIP
|
|
hipFunction_t cudaFunc = reinterpret_cast<hipFunction_t>(functionPtr);
|
|
#else
|
|
cudaFunction_t cudaFunc;
|
|
PADDLE_ENFORCE_GPU_SUCCESS(
|
|
cudaGetFuncBySymbol(&cudaFunc, functionPtr));
|
|
#endif
|
|
VLOG(10) << "[cudaKernelCallback] cudaFunc = " << cudaFunc
|
|
<< " functionPtr = " << functionPtr;
|
|
|
|
VectorizedDropoutForward<T, NoMaskFwFunctor<T, float>>
|
|
<<<grid_size, block_size, 0, stream>>>(id,
|
|
numel,
|
|
seed_data, // need save
|
|
x_data,
|
|
y_data,
|
|
out_data,
|
|
increment, // need save
|
|
main_offset,
|
|
dst_functor);
|
|
return cudaFunc;
|
|
};
|
|
backends::gpu::CUDAGraphNodeLauncher::Instance().KernelNodeLaunch(
|
|
parameterSetter, cudaKernelCallback);
|
|
|
|
VLOG(10) << "NON_CUDA_GRAPH seed = " << seed_data
|
|
<< ", increment = " << increment;
|
|
} else {
|
|
using MT = typename phi::dtype::MPTypeTrait<T>::Type;
|
|
MT factor = static_cast<MT>(1.0f - dropout_rate);
|
|
std::vector<DenseTensor*> outs = {out};
|
|
std::vector<const DenseTensor*> ins = {&x, &y};
|
|
|
|
funcs::ElementwiseKernel<T>(
|
|
dev_ctx, ins, &outs, ScaleAddFuctor<T>(factor, upscale_in_train));
|
|
}
|
|
}
|
|
|
|
} // namespace fusion
|
|
} // namespace phi
|
|
|
|
PD_REGISTER_KERNEL(fused_dropout_add,
|
|
GPU,
|
|
ALL_LAYOUT,
|
|
phi::fusion::FusedDropoutAddKernel,
|
|
float,
|
|
double,
|
|
phi::bfloat16,
|
|
phi::float16) {
|
|
kernel->OutputAt(1).SetDataType(phi::DataType::INT64);
|
|
}
|