Skip to content

P
Paddle

PaddlePaddle / Paddle
大约 1 年 前同步成功

通知 2299
Star 20931
Fork 5422
P
Paddle
未验证 提交 581b2c64 编写于 作者: F From00 提交者： GitHub
浏览文件
操作

Move GumbelSoftmax OP to phi (#39873) 

* Move GumbelSoftmax OP to phi

* platform::errors -> phi::errors; GumbelSoftmaxGradInferMeta -> backend.h/cc

* Use axis util in kernel impl

* Remove namespace platform::errors

* Use GetCPUEngine in Device Context
上级 caea126c
隐藏空白更改
内联 并排
Showing with 658 addition and 283 deletion
paddle/fluid/operators/gumbel_softmax_op.cc
浏览文件 @ 581b2c64
...@@ -12,10 +12,11 @@ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. ...@@ -12,10 +12,11 @@ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and See the License for the specific language governing permissions and
limitations under the License. */ limitations under the License. */
#include "paddle/fluid/operators/gumbel_softmax_op.h" #include "paddle/fluid/framework/infershape_utils.h"
#include <string> #include "paddle/fluid/framework/op_registry.h"
#include <unordered_map> #include "paddle/phi/core/infermeta_utils.h"
#include "paddle/fluid/operators/common_infer_shape_functions.h" #include "paddle/phi/infermeta/backward.h"
#include "paddle/phi/infermeta/unary.h"
namespace paddle { namespace paddle {
namespace operators { namespace operators {
...@@ -23,10 +24,6 @@ class GumbelSoftmaxOp : public framework::OperatorWithKernel { ...@@ -23,10 +24,6 @@ class GumbelSoftmaxOp : public framework::OperatorWithKernel {
public: public:
using framework::OperatorWithKernel::OperatorWithKernel; using framework::OperatorWithKernel::OperatorWithKernel;
void InferShape(framework::InferShapeContext* ctx) const override {
return UnaryOpUnchangedInferShapeCheckAxis(ctx);
}
protected: protected:
framework::OpKernelType GetExpectedKernelType( framework::OpKernelType GetExpectedKernelType(
const framework::ExecutionContext& ctx) const override { const framework::ExecutionContext& ctx) const override {
...@@ -71,20 +68,6 @@ Samples from the Gumbel-Softmax distribution and optionally discretizes. ...@@ -71,20 +68,6 @@ Samples from the Gumbel-Softmax distribution and optionally discretizes.
class GumbelSoftmaxGradOp : public framework::OperatorWithKernel { class GumbelSoftmaxGradOp : public framework::OperatorWithKernel {
public: public:
using framework::OperatorWithKernel::OperatorWithKernel; using framework::OperatorWithKernel::OperatorWithKernel;
void InferShape(framework::InferShapeContext* ctx) const override {
OP_INOUT_CHECK(ctx->HasInput("Out"), "Input", "Out", "gumbel_softmax_grad");
OP_INOUT_CHECK(ctx->HasInput(framework::GradVarName("Out")), "Input",
"Out@GRAD", "gumbel_softmax_grad");
PADDLE_ENFORCE_EQ(
ctx->GetInputDim("Out"),
ctx->GetInputDim(framework::GradVarName("Out")),
platform::errors::InvalidArgument("Input(Out) and its gradients "
"should have the same shape."));
ctx->SetOutputDim(framework::GradVarName("X"),
ctx->GetInputDim(framework::GradVarName("Out")));
}
}; };
template <typename T> template <typename T>
...@@ -107,17 +90,16 @@ class GumbelSoftmaxGradOpMaker : public framework::SingleGradOpMaker<T> { ...@@ -107,17 +90,16 @@ class GumbelSoftmaxGradOpMaker : public framework::SingleGradOpMaker<T> {
namespace ops = paddle::operators; namespace ops = paddle::operators;
DELCARE_INFER_SHAPE_FUNCTOR(gumbel_softmax, GumbelSoftmaxInferShapeFunctor,
PT_INFER_META(phi::GumbelSoftmaxInferMeta));
DELCARE_INFER_SHAPE_FUNCTOR(gumbel_softmax_grad,
GumbelSoftmaxGradInferShapeFunctor,
PT_INFER_META(phi::GumbelSoftmaxGradInferMeta));
REGISTER_OPERATOR(gumbel_softmax, ops::GumbelSoftmaxOp, REGISTER_OPERATOR(gumbel_softmax, ops::GumbelSoftmaxOp,
ops::GumbelSoftmaxOpMaker, ops::GumbelSoftmaxOpMaker,
ops::GumbelSoftmaxGradOpMaker<paddle::framework::OpDesc>, ops::GumbelSoftmaxGradOpMaker<paddle::framework::OpDesc>,
ops::GumbelSoftmaxGradOpMaker<paddle::imperative::OpBase>); ops::GumbelSoftmaxGradOpMaker<paddle::imperative::OpBase>,
REGISTER_OPERATOR(gumbel_softmax_grad, ops::GumbelSoftmaxGradOp); GumbelSoftmaxInferShapeFunctor);
REGISTER_OPERATOR(gumbel_softmax_grad, ops::GumbelSoftmaxGradOp,
REGISTER_OP_CPU_KERNEL( GumbelSoftmaxGradInferShapeFunctor);
gumbel_softmax,
ops::GumbelSoftmaxKernel<paddle::platform::CPUDeviceContext, float>,
ops::GumbelSoftmaxKernel<paddle::platform::CPUDeviceContext, double>);
REGISTER_OP_CPU_KERNEL(
gumbel_softmax_grad,
ops::GumbelSoftmaxGradKernel<paddle::platform::CPUDeviceContext, float>,
ops::GumbelSoftmaxGradKernel<paddle::platform::CPUDeviceContext, double>);
paddle/fluid/operators/gumbel_softmax_op.h 已删除 100644 → 0
浏览文件 @ caea126c
/* Copyright (c) 2021 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. */
#pragma once
#include "paddle/fluid/framework/generator.h"
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/fluid/operators/math/softmax.h"
#include "paddle/phi/kernels/funcs/math_function.h"
namespace paddle {
namespace operators {
using Tensor = framework::Tensor;
using DDim = framework::DDim;
template <typename T, int MajorType = Eigen::RowMajor,
typename IndexType = Eigen::DenseIndex>
using EigenMatrix = framework::EigenMatrix<T, MajorType, IndexType>;
template <typename T, size_t D, int MajorType = Eigen::RowMajor,
typename IndexType = Eigen::DenseIndex>
using EigenTensor = framework::EigenTensor<T, D>;
static inline int CanonicalAxis(const int axis, const int rank) {
if (axis < 0) {
return axis + rank;
}
return axis;
}
static inline int SizeToAxis(const int axis, DDim dims) {
int size = 1;
for (int i = 0; i < axis; i++) {
size *= dims[i];
}
return size;
}
static inline int SizeFromAxis(const int axis, DDim dims) {
int size = 1;
for (int i = axis; i < dims.size(); i++) {
size *= dims[i];
}
return size;
}
static inline int SizeOutAxis(const int axis, DDim dims) {
int size = 1;
for (int i = axis + 1; i < dims.size(); i++) {
size *= dims[i];
}
return size;
}
template <typename DeviceContext, typename T, int64_t Rank>
struct ArgMaxFunctor {
void operator()(const DeviceContext& ctx, const Tensor& in,
Tensor* index_tensor, const int64_t& axis) {
auto in_eigen = EigenTensor<T, Rank>::From(in, in.dims());
auto index_eigen = EigenTensor<int, Rank - 1>::From(*index_tensor);
index_eigen = in_eigen.argmax(axis).template cast<int>();
}
};
template <typename DeviceContext, typename T>
struct GumbleNoiseGenerator;
template <typename DeviceContext, typename T>
struct OneHotGenerator;
template <typename T>
struct GumbleNoiseGenerator<platform::CPUDeviceContext, T> {
static void Transform(const platform::CPUDeviceContext& context,
const T* input_data, T* output_data, int size_to_axis,
int size_from_axis, const float temperature) {
// generate uniform random number
const int size = size_to_axis * size_from_axis;
std::uniform_real_distribution<T> dist(0.00001, 1);
auto engine = paddle::framework::GetCPURandomEngine(0);
Tensor random_tensor;
auto* random_data =
random_tensor.mutable_data<T>({size}, platform::CPUPlace());
for (int64_t i = 0; i < size; ++i) {
random_data[i] = dist(*engine);
}
// generate gumbel noise
framework::DDim dim_2d{size_to_axis, size_from_axis};
auto gumbel_noise_eigen = EigenMatrix<T>::From(random_tensor, dim_2d);
gumbel_noise_eigen = -(((-(gumbel_noise_eigen.log())).log()));
// add noise
for (int64_t i = 0; i < size_to_axis * size_from_axis; i++) {
output_data[i] = (input_data[i] + random_data[i]) / temperature;
}
}
};
template <typename T>
struct OneHotGenerator<platform::CPUDeviceContext, T> {
static void Transform(const platform::CPUDeviceContext& context,
const Tensor& X, Tensor* Out, int axis) {
Tensor index;
std::vector<int> index_dim;
const auto rank = X.dims().size();
const int size_to_axis = SizeToAxis(axis, X.dims());
const int size_from_axis = SizeFromAxis(axis, X.dims());
const int size_out_axis = SizeOutAxis(axis, X.dims());
for (int i = 0; i < X.dims().size(); i++) {
if (i != axis) index_dim.push_back(X.dims().Get()[i]);
}
DDim index_ddim(index_dim.data(), rank - 1);
index.Resize(index_ddim);
auto* index_data = index.mutable_data<int>(context.GetPlace());
#define CALL_ARG_MINMAX_FUNCTOR(rank) \
ArgMaxFunctor<platform::CPUDeviceContext, T, rank> functor##rank; \
functor##rank(context, *Out, &index, axis);
switch (Out->dims().size()) {
case 1:
CALL_ARG_MINMAX_FUNCTOR(1);
break;
case 2:
CALL_ARG_MINMAX_FUNCTOR(2);
break;
case 3:
CALL_ARG_MINMAX_FUNCTOR(3);
break;
case 4:
CALL_ARG_MINMAX_FUNCTOR(4);
break;
case 5:
CALL_ARG_MINMAX_FUNCTOR(5);
break;
case 6:
CALL_ARG_MINMAX_FUNCTOR(6);
break;
default:
PADDLE_ENFORCE_LE(Out->dims().size(), 6,
platform::errors::InvalidArgument(
"gumbel_softmax operator doesn't supports "
"tensors whose ranks are greater "
"than 6 in CPU mode."));
break;
#undef CALL_ARG_MINMAX_FUNCTOR
}
phi::funcs::set_constant(context, Out, 0.0);
for (int i = 0; i < size_to_axis; i++) {
for (int j = 0; j < size_out_axis; j++) {
*(Out->data<T>() + i * size_from_axis + j +
index_data[i * size_out_axis + j] * size_out_axis) = 1.0;
}
}
}
};
template <typename DeviceContext, typename T>
class GumbelSoftmaxKernel : public framework::OpKernel<T> {
public:
void Compute(const framework::ExecutionContext& context) const override {
auto* X = context.Input<Tensor>("X");
auto* Out = context.Output<Tensor>("Out");
const int rank = X->dims().size();
const int axis = CanonicalAxis(context.Attr<int>("axis"), rank);
int axis_dim = X->dims()[axis];
const bool is_hard = context.Attr<bool>("hard");
const float temperature = context.Attr<float>("temperature");
PADDLE_ENFORCE_GT(temperature, 0,
platform::errors::InvalidArgument(
"The temperature must be greater than 0. But "
"received temperature = %f",
temperature));
// allocate memory on device.
Out->mutable_data<T>(context.GetPlace());
if (Out->numel() == 0) {
return;
}
const int size_to_axis = SizeToAxis(axis, X->dims());
const int size_from_axis = SizeFromAxis(axis, X->dims());
Tensor X_noise_2d, Out_2d;
X_noise_2d.Resize({size_to_axis, size_from_axis});
Out_2d.ShareDataWith(*Out).Resize({size_to_axis, size_from_axis});
// generate gumbel noise and add it to X
auto* x_noise_data = X_noise_2d.mutable_data<T>(context.GetPlace());
GumbleNoiseGenerator<DeviceContext, T>::Transform(
context.template device_context<DeviceContext>(), X->data<T>(),
x_noise_data, size_to_axis, size_from_axis, temperature);
#ifdef PADDLE_ON_INFERENCE
math::SoftmaxFunctor<DeviceContext, T, true>()(
context.template device_context<DeviceContext>(), axis_dim, &X_noise_2d,
&Out_2d);
#else
math::SoftmaxFunctor<DeviceContext, T, false>()(
context.template device_context<DeviceContext>(), axis_dim, &X_noise_2d,
&Out_2d);
#endif
if (is_hard) {
OneHotGenerator<DeviceContext, T>::Transform(
context.template device_context<DeviceContext>(), *X, Out, axis);
}
}
};
template <typename DeviceContext, typename T>
class GumbelSoftmaxGradKernel : public framework::OpKernel<T> {
public:
void Compute(const framework::ExecutionContext& context) const override {
auto* Out = context.Input<Tensor>("Out");
auto* dOut = context.Input<Tensor>(framework::GradVarName("Out"));
auto* dX = context.Output<Tensor>(framework::GradVarName("X"));
const int rank = dX->dims().size();
const int axis = CanonicalAxis(context.Attr<int>("axis"), rank);
int axis_dim = dX->dims()[axis];
// allocate memory on device.
dX->mutable_data<T>(context.GetPlace());
if (dX->numel() == 0) {
return;
}
const int size_to_axis = SizeToAxis(axis, dX->dims());
const int size_from_axis = SizeFromAxis(axis, dX->dims());
Tensor dX_2d, Out_2d, dOut_2d;
dX_2d.ShareDataWith(*dX).Resize({size_to_axis, size_from_axis});
Out_2d.ShareDataWith(*Out).Resize({size_to_axis, size_from_axis});
dOut_2d.ShareDataWith(*dOut).Resize({size_to_axis, size_from_axis});
math::SoftmaxGradFunctor<DeviceContext, T>()(
context.template device_context<DeviceContext>(), axis_dim, &Out_2d,
&dOut_2d, &dX_2d);
}
};
} // namespace operators
} // namespace paddle
paddle/phi/infermeta/backward.cc
浏览文件 @ 581b2c64
...@@ -76,4 +76,16 @@ void GeneralBinaryGradInferMeta(const MetaTensor& x, ...@@ -76,4 +76,16 @@ void GeneralBinaryGradInferMeta(const MetaTensor& x,
} }
} }
void GumbelSoftmaxGradInferMeta(const MetaTensor& out,
const MetaTensor& dout,
int axis,
MetaTensor* dx) {
PADDLE_ENFORCE_EQ(
out.dims(),
dout.dims(),
errors::InvalidArgument(
"Input(Out) and its gradients should have the same shape."));
dx->share_meta(dout);
}
} // namespace phi } // namespace phi
paddle/phi/infermeta/backward.h
浏览文件 @ 581b2c64
...@@ -34,4 +34,8 @@ void GeneralBinaryGradInferMeta(const MetaTensor& x, ...@@ -34,4 +34,8 @@ void GeneralBinaryGradInferMeta(const MetaTensor& x,
MetaTensor* dx, MetaTensor* dx,
MetaTensor* dy); MetaTensor* dy);
void GumbelSoftmaxGradInferMeta(const MetaTensor& out,
const MetaTensor& dout,
int axis,
MetaTensor* dx);
} // namespace phi } // namespace phi
paddle/phi/infermeta/unary.cc
浏览文件 @ 581b2c64
...@@ -27,6 +27,30 @@ void UnchangedInferMeta(const MetaTensor& x, MetaTensor* out) { ...@@ -27,6 +27,30 @@ void UnchangedInferMeta(const MetaTensor& x, MetaTensor* out) {
out->share_meta(x); out->share_meta(x);
} }
// meta x -> out without change, check if axis in range [-Rank(x), Rank(x)-1]
void UnchangedInferMetaCheckAxis(const MetaTensor& x,
int axis,
MetaTensor* out) {
auto rank = x.dims().size();
PADDLE_ENFORCE_GE(
axis,
-rank,
errors::InvalidArgument(
"Attr(axis) value should be in range [-R, R-1], "
"R is the rank of Input(X). But received axis: %d, R: %d.",
axis,
rank));
PADDLE_ENFORCE_LT(
axis,
rank,
phi::errors::InvalidArgument(
"Attr(axis) value should be in range [-R, R-1], "
"R is the rank of Input(X). But received axis: %d, R: %d.",
axis,
rank));
out->share_meta(x);
}
void FlattenInferMeta(const MetaTensor& x, void FlattenInferMeta(const MetaTensor& x,
int start_axis, int start_axis,
int stop_axis, int stop_axis,
...@@ -75,6 +99,14 @@ void FlattenInferMeta(const MetaTensor& x, ...@@ -75,6 +99,14 @@ void FlattenInferMeta(const MetaTensor& x,
} }
} }
void GumbelSoftmaxInferMeta(const MetaTensor& x,
float temperature,
bool hard,
int axis,
MetaTensor* out) {
UnchangedInferMetaCheckAxis(x, axis, out);
}
void CastInferMeta(const MetaTensor& x, DataType out_dtype, MetaTensor* out) { void CastInferMeta(const MetaTensor& x, DataType out_dtype, MetaTensor* out) {
out->set_dims(x.dims()); out->set_dims(x.dims());
out->set_dtype(out_dtype); out->set_dtype(out_dtype);
......
paddle/phi/infermeta/unary.h
浏览文件 @ 581b2c64
...@@ -34,11 +34,22 @@ class MetaConfig; ...@@ -34,11 +34,22 @@ class MetaConfig;
void UnchangedInferMeta(const MetaTensor& x, MetaTensor* out); void UnchangedInferMeta(const MetaTensor& x, MetaTensor* out);
// meta x -> out without change, check if axis in range [-Rank(x), Rank(x)-1]
void UnchangedInferMetaCheckAxis(const MetaTensor& x,
int axis,
MetaTensor* out);
void FlattenInferMeta(const MetaTensor& x, void FlattenInferMeta(const MetaTensor& x,
int start_axis, int start_axis,
int stop_axis, int stop_axis,
MetaTensor* out); MetaTensor* out);
void GumbelSoftmaxInferMeta(const MetaTensor& x,
float temperature,
bool hard,
int axis,
MetaTensor* out);
void CastInferMeta(const MetaTensor& x, DataType out_dtype, MetaTensor* out); void CastInferMeta(const MetaTensor& x, DataType out_dtype, MetaTensor* out);
void CholeskyInferMeta(const MetaTensor& x, bool upper, MetaTensor* out); void CholeskyInferMeta(const MetaTensor& x, bool upper, MetaTensor* out);
......
paddle/phi/kernels/CMakeLists.txt
浏览文件 @ 581b2c64
...@@ -10,7 +10,7 @@ add_subdirectory(funcs) ...@@ -10,7 +10,7 @@ add_subdirectory(funcs)
set_property(GLOBAL PROPERTY PHI_KERNELS "") set_property(GLOBAL PROPERTY PHI_KERNELS "")
set(COMMON_KERNEL_DEPS dense_tensor sparse_coo_tensor sparse_csr_tensor kernel_context kernel_factory arg_map_context convert_utils lod_utils) set(COMMON_KERNEL_DEPS dense_tensor sparse_coo_tensor sparse_csr_tensor kernel_context kernel_factory arg_map_context convert_utils lod_utils)
set(COMMON_KERNEL_DEPS ${COMMON_KERNEL_DEPS} eigen_function blas math_function im2col concat_and_split_functor) set(COMMON_KERNEL_DEPS ${COMMON_KERNEL_DEPS} eigen_function blas math_function im2col concat_and_split_functor softmax)
# remove this dep after removing fluid deps on tensor creation # remove this dep after removing fluid deps on tensor creation
set(COMMON_KERNEL_DEPS ${COMMON_KERNEL_DEPS} phi_api_utils) set(COMMON_KERNEL_DEPS ${COMMON_KERNEL_DEPS} phi_api_utils)
set(COMMON_KERNEL_DEPS ${COMMON_KERNEL_DEPS} infermeta) set(COMMON_KERNEL_DEPS ${COMMON_KERNEL_DEPS} infermeta)
......
paddle/phi/kernels/cpu/gumbel_softmax_grad_kernel.cc 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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/phi/kernels/gumbel_softmax_grad_kernel.h"
#include "paddle/phi/kernels/impl/gumbel_softmax_grad_kernel_impl.h"
#include "paddle/phi/core/kernel_registry.h"
PD_REGISTER_KERNEL(gumbel_softmax_grad,
CPU,
ALL_LAYOUT,
phi::GumbelSoftmaxGradKernel,
float,
double) {}
paddle/phi/kernels/cpu/gumbel_softmax_kernel.cc 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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/phi/kernels/gumbel_softmax_kernel.h"
#include "paddle/phi/kernels/impl/gumbel_softmax_kernel_impl.h"
#include "paddle/phi/core/kernel_registry.h"
#include "paddle/phi/kernels/funcs/axis_utils.h"
#include "paddle/phi/kernels/funcs/math_function.h"
namespace phi {
template <typename T>
struct GumbleNoiseGenerator<CPUContext, T> {
static void Transform(const CPUContext& ctx,
const T* input_data,
T* output_data,
int size_to_axis,
int size_from_axis,
const float temperature) {
// generate uniform random number
const int size = size_to_axis * size_from_axis;
std::uniform_real_distribution<T> dist(0.00001, 1);
auto engine = ctx.GetGenerator()->GetCPUEngine();
DenseTensor random_tensor;
random_tensor.Resize(make_ddim({size}));
auto* random_data = ctx.template Alloc<T>(&random_tensor);
for (int64_t i = 0; i < size; ++i) {
random_data[i] = dist(*engine);
}
// generate gumbel noise
DDim dim_2d{size_to_axis, size_from_axis};
auto gumbel_noise_eigen = EigenMatrix<T>::From(random_tensor, dim_2d);
gumbel_noise_eigen = -(((-(gumbel_noise_eigen.log())).log()));
// add noise
for (int64_t i = 0; i < size_to_axis * size_from_axis; i++) {
output_data[i] = (input_data[i] + random_data[i]) / temperature;
}
}
};
template <typename T>
struct OneHotGenerator<CPUContext, T> {
static void Transform(const CPUContext& ctx,
const DenseTensor& x,
DenseTensor* out,
int axis) {
DenseTensor index;
std::vector<int> index_dim;
const auto rank = x.dims().size();
const int size_to_axis = funcs::SizeToAxis(axis, x.dims());
const int size_from_axis = funcs::SizeFromAxis(axis, x.dims());
const int size_out_axis = funcs::SizeOutAxis(axis, x.dims());
for (int i = 0; i < x.dims().size(); i++) {
if (i != axis) index_dim.push_back(x.dims().Get()[i]);
}
DDim index_ddim(index_dim.data(), rank - 1);
index.Resize(index_ddim);
auto* index_data = ctx.template Alloc<int>(&index);
#define CALL_ARG_MINMAX_FUNCTOR(rank) \
ArgMaxFunctor<CPUContext, T, rank> functor##rank; \
functor##rank(ctx, *out, &index, axis);
switch (out->dims().size()) {
case 1:
CALL_ARG_MINMAX_FUNCTOR(1);
break;
case 2:
CALL_ARG_MINMAX_FUNCTOR(2);
break;
case 3:
CALL_ARG_MINMAX_FUNCTOR(3);
break;
case 4:
CALL_ARG_MINMAX_FUNCTOR(4);
break;
case 5:
CALL_ARG_MINMAX_FUNCTOR(5);
break;
case 6:
CALL_ARG_MINMAX_FUNCTOR(6);
break;
default:
PADDLE_ENFORCE_LE(
out->dims().size(),
6,
errors::InvalidArgument("gumbel_softmax operator doesn't supports "
"tensors whose ranks are greater "
"than 6 in CPU mode."));
break;
#undef CALL_ARG_MINMAX_FUNCTOR
}
funcs::set_constant(ctx, out, 0.0);
for (int i = 0; i < size_to_axis; i++) {
for (int j = 0; j < size_out_axis; j++) {
*(out->data<T>() + i * size_from_axis + j +
index_data[i * size_out_axis + j] * size_out_axis) = 1.0;
}
}
}
};
} // namespace phi
PD_REGISTER_KERNEL(
gumbel_softmax, CPU, ALL_LAYOUT, phi::GumbelSoftmaxKernel, float, double) {}
paddle/phi/kernels/gpu/gumbel_softmax_grad_kernel.cu 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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/phi/kernels/gumbel_softmax_grad_kernel.h"
#include "paddle/phi/kernels/impl/gumbel_softmax_grad_kernel_impl.h"
#include "paddle/phi/core/kernel_registry.h"
PD_REGISTER_KERNEL(gumbel_softmax_grad,
GPU,
ALL_LAYOUT,
phi::GumbelSoftmaxGradKernel,
float,
double) {}
paddle/fluid/operators/gumbel_softmax_op.cu → paddle/phi/kernels/gpu/gumbel_softmax_kernel.cu
浏览文件 @ 581b2c64
/* Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved. // Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
//
Licensed under the Apache License, Version 2.0 (the "License"); // Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License. // you may not use this file except in compliance with the License.
You may obtain a copy of the License at // You may obtain a copy of the License at
//
http://www.apache.org/licenses/LICENSE-2.0 // http://www.apache.org/licenses/LICENSE-2.0
//
Unless required by applicable law or agreed to in writing, software // Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS, // distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and // See the License for the specific language governing permissions and
limitations under the License. */ // limitations under the License.
#pragma once
#include "paddle/phi/kernels/gumbel_softmax_kernel.h"
#include "paddle/fluid/framework/op_registry.h" #include "paddle/phi/kernels/impl/gumbel_softmax_kernel_impl.h"
#include "paddle/fluid/framework/operator.h"
#include "paddle/fluid/operators/gumbel_softmax_op.h" #include "paddle/phi/core/kernel_registry.h"
#include "paddle/phi/kernels/funcs/axis_utils.h"
#if defined(__NVCC__) || defined(__HIPCC__) #if defined(__NVCC__) || defined(__HIPCC__)
#ifdef __NVCC__ #ifdef __NVCC__
...@@ -31,11 +32,10 @@ namespace cub = hipcub; ...@@ -31,11 +32,10 @@ namespace cub = hipcub;
#include <thrust/random.h> #include <thrust/random.h>
#include <thrust/transform.h> #include <thrust/transform.h>
#include "paddle/fluid/framework/generator.h" #include "paddle/fluid/framework/generator.h"
#include "paddle/fluid/memory/memcpy.h" #include "paddle/fluid/framework/tensor_util.h"
#include "paddle/phi/kernels/funcs/math_function.h"
namespace paddle {
namespace operators {
namespace phi {
template <typename K, typename V> template <typename K, typename V>
using KeyValuePair = cub::KeyValuePair<K, V>; using KeyValuePair = cub::KeyValuePair<K, V>;
...@@ -46,7 +46,9 @@ struct UniformCUDAGenerator { ...@@ -46,7 +46,9 @@ struct UniformCUDAGenerator {
unsigned int offset_ = 0; unsigned int offset_ = 0;
HOSTDEVICE UniformCUDAGenerator(T min, T max, unsigned int seed) HOSTDEVICE UniformCUDAGenerator(T min, T max, unsigned int seed)
: min_(min), max_(max), seed_(seed) {} : min_(min), max_(max), seed_(seed) {}
HOSTDEVICE UniformCUDAGenerator(T min, T max, unsigned int seed, HOSTDEVICE UniformCUDAGenerator(T min,
T max,
unsigned int seed,
unsigned int offset) unsigned int offset)
: min_(min), max_(max), seed_(seed), offset_(offset) {} : min_(min), max_(max), seed_(seed), offset_(offset) {}
...@@ -60,9 +62,12 @@ struct UniformCUDAGenerator { ...@@ -60,9 +62,12 @@ struct UniformCUDAGenerator {
}; };
template <typename T, size_t BlockDim> template <typename T, size_t BlockDim>
__global__ void OneHotCUDAKernel(const int64_t height, const int64_t width, __global__ void OneHotCUDAKernel(const int64_t height,
const int64_t size_out_axis, const T init, const int64_t width,
const T* in, T* out) { const int64_t size_out_axis,
const T init,
const T* in,
T* out) {
typedef cub::BlockReduce<KeyValuePair<int, T>, BlockDim> BlockReduce; typedef cub::BlockReduce<KeyValuePair<int, T>, BlockDim> BlockReduce;
__shared__ typename BlockReduce::TempStorage temp_storage; __shared__ typename BlockReduce::TempStorage temp_storage;
...@@ -85,32 +90,40 @@ __global__ void OneHotCUDAKernel(const int64_t height, const int64_t width, ...@@ -85,32 +90,40 @@ __global__ void OneHotCUDAKernel(const int64_t height, const int64_t width,
} }
template <typename T> template <typename T>
struct OneHotGenerator<platform::CUDADeviceContext, T> { struct OneHotGenerator<GPUContext, T> {
static void Transform(const platform::CUDADeviceContext& context, static void Transform(const GPUContext& ctx,
const Tensor& X, Tensor* Out, int axis) { const DenseTensor& X,
const int size_to_axis = SizeToAxis(axis, X.dims()); DenseTensor* out,
const int size_from_axis = SizeFromAxis(axis, X.dims()); int axis) {
const int size_out_axis = SizeOutAxis(axis, X.dims()); const int size_to_axis = funcs::SizeToAxis(axis, X.dims());
const int size_from_axis = funcs::SizeFromAxis(axis, X.dims());
const int size_out_axis = funcs::SizeOutAxis(axis, X.dims());
constexpr int thread_size = 512; constexpr int thread_size = 512;
int64_t max_grid_dimx = context.GetCUDAMaxGridDimSize()[0]; int64_t max_grid_dimx = ctx.GetCUDAMaxGridDimSize()[0];
int64_t height = size_to_axis * size_out_axis; int64_t height = size_to_axis * size_out_axis;
int block_size = height < max_grid_dimx ? height : max_grid_dimx; int block_size = height < max_grid_dimx ? height : max_grid_dimx;
Tensor input_tensor; DenseTensor input_tensor;
input_tensor.mutable_data<T>(Out->dims(), platform::CUDAPlace()); input_tensor.Resize(out->dims());
paddle::framework::TensorCopy(*Out, context.GetPlace(), &input_tensor); ctx.template Alloc<T>(&input_tensor);
phi::funcs::set_constant(context, Out, 0.0); paddle::framework::TensorCopy(*out, ctx.GetPlace(), &input_tensor);
OneHotCUDAKernel< funcs::set_constant(ctx, out, 0.0);
T, thread_size><<<block_size, thread_size, 0, context.stream()>>>( OneHotCUDAKernel<T,
height, size_from_axis / size_out_axis, size_out_axis, thread_size><<<block_size, thread_size, 0, ctx.stream()>>>(
std::numeric_limits<T>::lowest(), input_tensor.data<T>(), height,
Out->data<T>()); size_from_axis / size_out_axis,
size_out_axis,
std::numeric_limits<T>::lowest(),
input_tensor.data<T>(),
out->data<T>());
} }
}; };
template <typename T> template <typename T>
__global__ void AddGumbelNoiseCUDAKernel(const T* input_data, T* output_data, __global__ void AddGumbelNoiseCUDAKernel(const T* input_data,
T* noise, const float temperature, T* output_data,
T* noise,
const float temperature,
int64_t n) { int64_t n) {
int index = threadIdx.x + blockIdx.x * blockDim.x; int index = threadIdx.x + blockIdx.x * blockDim.x;
int step = blockDim.x * gridDim.x; int step = blockDim.x * gridDim.x;
...@@ -121,29 +134,34 @@ __global__ void AddGumbelNoiseCUDAKernel(const T* input_data, T* output_data, ...@@ -121,29 +134,34 @@ __global__ void AddGumbelNoiseCUDAKernel(const T* input_data, T* output_data,
} }
template <typename T> template <typename T>
struct GumbleNoiseGenerator<platform::CUDADeviceContext, T> { struct GumbleNoiseGenerator<GPUContext, T> {
static void Transform(const platform::CUDADeviceContext& context, static void Transform(const GPUContext& ctx,
const T* input_data, T* output_data, int size_to_axis, const T* input_data,
int size_from_axis, const float temperature) { T* output_data,
Tensor random_tensor; int size_to_axis,
int size_from_axis,
const float temperature) {
DenseTensor random_tensor;
int64_t size = size_to_axis * size_from_axis; int64_t size = size_to_axis * size_from_axis;
T* random_data = random_tensor.Resize(make_ddim({size}));
random_tensor.mutable_data<T>({size}, platform::CUDAPlace()); auto* random_data = ctx.template Alloc<T>(&random_tensor);
thrust::counting_iterator<int64_t> index_sequence_begin(0); thrust::counting_iterator<int64_t> index_sequence_begin(0);
// generate gumbel noise // generate gumbel noise
int device_id = context.GetPlace().GetDeviceId(); int device_id = ctx.GetPlace().GetDeviceId();
auto gen_cuda = framework::GetDefaultCUDAGenerator(device_id); auto gen_cuda = paddle::framework::GetDefaultCUDAGenerator(device_id);
if (gen_cuda->GetIsInitPy()) { if (gen_cuda->GetIsInitPy()) {
auto seed_offset = gen_cuda->IncrementOffset(1); auto seed_offset = gen_cuda->IncrementOffset(1);
int64_t gen_offset = size * seed_offset.second; int64_t gen_offset = size * seed_offset.second;
thrust::transform( thrust::transform(
index_sequence_begin, index_sequence_begin + size, index_sequence_begin,
index_sequence_begin + size,
thrust::device_ptr<T>(random_data), thrust::device_ptr<T>(random_data),
UniformCUDAGenerator<T>(0.00001, 1, seed_offset.first, gen_offset)); UniformCUDAGenerator<T>(0.00001, 1, seed_offset.first, gen_offset));
} else { } else {
const unsigned int seed = std::random_device()(); const unsigned int seed = std::random_device()();
thrust::transform(index_sequence_begin, index_sequence_begin + size, thrust::transform(index_sequence_begin,
index_sequence_begin + size,
thrust::device_ptr<T>(random_data), thrust::device_ptr<T>(random_data),
UniformCUDAGenerator<T>(0.00001, 1, seed)); UniformCUDAGenerator<T>(0.00001, 1, seed));
} }
...@@ -151,22 +169,13 @@ struct GumbleNoiseGenerator<platform::CUDADeviceContext, T> { ...@@ -151,22 +169,13 @@ struct GumbleNoiseGenerator<platform::CUDADeviceContext, T> {
// add gumbel noise to X // add gumbel noise to X
const int thread_size = 512; const int thread_size = 512;
int64_t block_size = (size + thread_size) / thread_size; int64_t block_size = (size + thread_size) / thread_size;
AddGumbelNoiseCUDAKernel< AddGumbelNoiseCUDAKernel<T><<<block_size, thread_size, 0, ctx.stream()>>>(
T><<<block_size, thread_size, 0, context.stream()>>>(
input_data, output_data, random_data, temperature, size); input_data, output_data, random_data, temperature, size);
} }
}; };
} // namespace phi
#endif #endif
} // namespace operators
} // namespace paddle PD_REGISTER_KERNEL(
gumbel_softmax, GPU, ALL_LAYOUT, phi::GumbelSoftmaxKernel, float, double) {}
namespace ops = paddle::operators;
namespace plat = paddle::platform;
REGISTER_OP_CUDA_KERNEL(
gumbel_softmax, ops::GumbelSoftmaxKernel<plat::CUDADeviceContext, float>,
ops::GumbelSoftmaxKernel<plat::CUDADeviceContext, double>);
REGISTER_OP_CUDA_KERNEL(
gumbel_softmax_grad,
ops::GumbelSoftmaxGradKernel<plat::CUDADeviceContext, float>,
ops::GumbelSoftmaxGradKernel<plat::CUDADeviceContext, double>);
paddle/phi/kernels/gumbel_softmax_grad_kernel.h 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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.
#pragma once
#include "paddle/phi/core/dense_tensor.h"
namespace phi {
template <typename T, typename Context>
void GumbelSoftmaxGradKernel(const Context& dev_ctx,
const DenseTensor& out,
const DenseTensor& dout,
int axis,
DenseTensor* dx);
} // namespace phi
paddle/phi/kernels/gumbel_softmax_kernel.h 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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.
#pragma once
#include "paddle/phi/core/dense_tensor.h"
namespace phi {
template <typename T, typename Context>
void GumbelSoftmaxKernel(const Context& dev_ctx,
const DenseTensor& x,
float temperature,
bool hard,
int axis,
DenseTensor* out);
} // namespace phi
paddle/phi/kernels/impl/gumbel_softmax_grad_kernel_impl.h 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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.
#pragma once
#include "paddle/fluid/operators/math/softmax.h"
#include "paddle/fluid/operators/math/softmax_impl.h"
#include "paddle/phi/core/dense_tensor.h"
#include "paddle/phi/kernels/funcs/axis_utils.h"
namespace phi {
template <typename T, typename Context>
void GumbelSoftmaxGradKernel(const Context& ctx,
const DenseTensor& out,
const DenseTensor& dout,
int axis,
DenseTensor* dx) {
const int rank = dx->dims().size();
axis = funcs::CanonicalAxis(axis, rank);
int axis_dim = dx->dims()[axis];
// allocate memory on device.
ctx.template Alloc<T>(dx);
if (dx->numel() == 0) {
return;
}
const int size_to_axis = funcs::SizeToAxis(axis, dx->dims());
const int size_from_axis = funcs::SizeFromAxis(axis, dx->dims());
DenseTensor dx_2d(*dx), out_2d(out), dout_2d(dout);
dx_2d.Resize({size_to_axis, size_from_axis});
out_2d.Resize({size_to_axis, size_from_axis});
dout_2d.Resize({size_to_axis, size_from_axis});
paddle::operators::math::SoftmaxGradFunctor<Context, T>()(
ctx, axis_dim, &out_2d, &dout_2d, &dx_2d);
}
} // namespace phi
paddle/phi/kernels/impl/gumbel_softmax_kernel_impl.h 0 → 100644
浏览文件 @ 581b2c64
// Copyright (c) 2022 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.
#pragma once
#include <random>
#include "paddle/fluid/operators/math/softmax.h"
#include "paddle/fluid/operators/math/softmax_impl.h"
#include "paddle/phi/core/dense_tensor.h"
#include "paddle/phi/kernels/funcs/axis_utils.h"
#include "paddle/phi/kernels/funcs/eigen/common.h"
namespace phi {
template <typename Context, typename T, int64_t Rank>
struct ArgMaxFunctor {
void operator()(const Context& ctx,
const DenseTensor& in,
DenseTensor* index_tensor,
const int64_t& axis) {
auto in_eigen = EigenTensor<T, Rank>::From(in, in.dims());
auto index_eigen = EigenTensor<int, Rank - 1>::From(*index_tensor);
index_eigen = in_eigen.argmax(axis).template cast<int>();
}
};
template <typename Context, typename T>
struct GumbleNoiseGenerator;
template <typename Context, typename T>
struct OneHotGenerator;
template <typename T, typename Context>
void GumbelSoftmaxKernel(const Context& ctx,
const DenseTensor& x,
float temperature,
bool hard,
int axis,
DenseTensor* out) {
const int rank = x.dims().size();
axis = funcs::CanonicalAxis(axis, rank);
int axis_dim = x.dims()[axis];
PADDLE_ENFORCE_GT(temperature,
0,
phi::errors::InvalidArgument(
"The temperature must be greater than 0. But "
"received temperature = %f",
temperature));
// allocate memory on device.
ctx.template Alloc<T>(out);
if (out->numel() == 0) {
return;
}
const int size_to_axis = funcs::SizeToAxis(axis, x.dims());
const int size_from_axis = funcs::SizeFromAxis(axis, x.dims());
DenseTensor x_noise_2d, out_2d(*out);
x_noise_2d.Resize({size_to_axis, size_from_axis});
out_2d.Resize({size_to_axis, size_from_axis});
// generate gumbel noise and add it to X
auto* x_noise_data = ctx.template Alloc<T>(&x_noise_2d);
GumbleNoiseGenerator<Context, T>::Transform(ctx,
x.data<T>(),
x_noise_data,
size_to_axis,
size_from_axis,
temperature);
#ifdef PADDLE_ON_INFERENCE
paddle::operators::math::SoftmaxFunctor<Context, T, true>()(
ctx, axis_dim, &x_noise_2d, &out_2d);
#else
paddle::operators::math::SoftmaxFunctor<Context, T, false>()(
ctx, axis_dim, &x_noise_2d, &out_2d);
#endif
if (hard) {
OneHotGenerator<Context, T>::Transform(ctx, x, out, axis);
}
}
} // namespace phi
paddle/phi/ops/compat/gumbel_softmax_sig.cc 0 → 100644
浏览文件 @ 581b2c64
/* Copyright (c) 2022 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/phi/core/compat/op_utils.h"
namespace phi {
KernelSignature GumbelSoftmaxGradOpArgumentMapping(
const ArgumentMappingContext& ctx) {
return KernelSignature("gumbel_softmax_grad",
{"Out", GradVarName("Out")},
{"axis"},
{GradVarName("X")});
}
} // namespace phi
PD_REGISTER_ARG_MAPPING_FN(gumbel_softmax_grad,
phi::GumbelSoftmaxGradOpArgumentMapping);
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或