wehub-resource-sync
383 lines
14 KiB
C++
383 lines
14 KiB
C++
// 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/set_value_grad_kernel.h"
|
|
|
|
#include "paddle/phi/backends/cpu/cpu_context.h"
|
|
#include "paddle/phi/common/int_array.h"
|
|
#include "paddle/phi/core/dense_tensor.h"
|
|
#include "paddle/phi/core/kernel_registry.h"
|
|
#include "paddle/phi/core/tensor_utils.h"
|
|
#include "paddle/phi/kernels/full_kernel.h"
|
|
#include "paddle/phi/kernels/funcs/common_shape.h"
|
|
#include "paddle/phi/kernels/funcs/eigen/common.h"
|
|
#include "paddle/phi/kernels/funcs/eigen/eigen_function.h"
|
|
#include "paddle/phi/kernels/funcs/math_function.h"
|
|
#include "paddle/phi/kernels/funcs/strided_slice.h"
|
|
#include "paddle/phi/kernels/impl/share_data_kernel_impl.h"
|
|
#include "paddle/phi/kernels/reduce_sum_kernel.h"
|
|
#include "paddle/phi/kernels/reshape_kernel.h"
|
|
|
|
namespace phi {
|
|
|
|
inline void GetOffsets(const DDim& big_dim,
|
|
const DDim& small_dim,
|
|
DDim start_offset,
|
|
int cur_dim,
|
|
std::vector<DDim>* offsets) {
|
|
if (cur_dim == big_dim.size()) {
|
|
offsets->push_back(start_offset);
|
|
return;
|
|
}
|
|
if (small_dim[cur_dim] == big_dim[cur_dim]) {
|
|
GetOffsets(big_dim, small_dim, start_offset, cur_dim + 1, offsets);
|
|
} else {
|
|
for (int i = 0; i < big_dim[cur_dim]; i++) {
|
|
GetOffsets(big_dim, small_dim, start_offset, cur_dim + 1, offsets);
|
|
start_offset[cur_dim] += 1;
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context, size_t RANK>
|
|
void SetValueGradImpl(const Context& dev_ctx,
|
|
const DenseTensor& out_grad,
|
|
std::vector<int64_t>& starts_local, // NOLINT
|
|
std::vector<int64_t>& ends_local, // NOLINT
|
|
std::vector<int64_t>& steps_local, // NOLINT
|
|
const std::vector<int64_t>& axes,
|
|
const std::vector<int64_t>& decrease_axes,
|
|
const std::vector<int64_t>& none_axes UNUSED,
|
|
DenseTensor* x_grad,
|
|
DenseTensor* value_grad) {
|
|
PADDLE_ENFORCE_EQ(
|
|
out_grad.IsInitialized(),
|
|
true,
|
|
errors::PermissionDenied(
|
|
"The input of `set_value_grad`(out_grad) has not been initialized"));
|
|
|
|
auto in_dims = out_grad.dims();
|
|
|
|
std::vector<int> decrease_axis_int32(decrease_axes.begin(),
|
|
decrease_axes.end());
|
|
std::vector<int> axes_int32(axes.begin(), axes.end());
|
|
std::vector<int> infer_flags(axes.size(), 1);
|
|
std::vector<int64_t> out_dims_vector(in_dims.size(), -1);
|
|
funcs::StridedSliceOutDims(starts_local,
|
|
ends_local,
|
|
steps_local,
|
|
axes_int32,
|
|
infer_flags,
|
|
in_dims,
|
|
decrease_axis_int32,
|
|
out_dims_vector.data(),
|
|
axes.size(),
|
|
false);
|
|
|
|
DDim out_dims(make_ddim(out_dims_vector));
|
|
|
|
std::vector<int> reverse_vector(starts_local.size(), 0);
|
|
funcs::StridedSliceFunctor(starts_local.data(),
|
|
ends_local.data(),
|
|
steps_local.data(),
|
|
axes_int32.data(),
|
|
reverse_vector.data(),
|
|
in_dims,
|
|
infer_flags,
|
|
decrease_axis_int32,
|
|
starts_local.size());
|
|
|
|
auto starts_indices = Eigen::DSizes<int64_t, RANK>();
|
|
auto ends_indices = Eigen::DSizes<int64_t, RANK>();
|
|
auto steps_indices = Eigen::DSizes<int64_t, RANK>();
|
|
auto reverse_axis = Eigen::array<bool, RANK>();
|
|
|
|
for (size_t axis = 0; axis < RANK; axis++) {
|
|
starts_indices[axis] = 0;
|
|
ends_indices[axis] = out_dims[axis];
|
|
steps_indices[axis] = 1;
|
|
reverse_axis[axis] = false;
|
|
}
|
|
|
|
for (size_t axis = 0; axis < axes.size(); axis++) {
|
|
int axis_index = axes[axis];
|
|
starts_indices[axis_index] = starts_local[axis];
|
|
ends_indices[axis_index] = ends_local[axis];
|
|
steps_indices[axis_index] = steps_local[axis];
|
|
reverse_axis[axis_index] = (reverse_vector[axis] == 1) ? true : false;
|
|
}
|
|
|
|
bool need_reverse = false;
|
|
for (size_t axis = 0; axis < axes.size(); axis++) {
|
|
if (reverse_vector[axis] == 1) {
|
|
need_reverse = true;
|
|
break;
|
|
}
|
|
}
|
|
|
|
auto& place = *dev_ctx.eigen_device();
|
|
funcs::SetConstant<Context, T> set_zero;
|
|
|
|
if (x_grad) {
|
|
// Set gradient of `Input`
|
|
Copy(dev_ctx, out_grad, dev_ctx.GetPlace(), false, x_grad);
|
|
|
|
auto x_grad_t = EigenTensor<T, RANK, Eigen::RowMajor>::From(*x_grad);
|
|
|
|
DenseTensor tmp = Full<T>(dev_ctx, out_dims_vector, static_cast<T>(0));
|
|
auto tmp_t = EigenTensor<T, RANK, Eigen::RowMajor>::From(tmp);
|
|
|
|
x_grad_t.stridedSlice(starts_indices, ends_indices, steps_indices)
|
|
.device(place) = tmp_t;
|
|
}
|
|
if (value_grad) {
|
|
dev_ctx.template Alloc<T>(value_grad);
|
|
set_zero(dev_ctx, value_grad, static_cast<T>(0));
|
|
|
|
auto in_t = EigenTensor<T, RANK, Eigen::RowMajor>::From(out_grad);
|
|
|
|
if (value_grad->dims() == out_dims) {
|
|
auto value_grad_t =
|
|
EigenTensor<T, RANK, Eigen::RowMajor>::From(*value_grad);
|
|
if (need_reverse) {
|
|
DenseTensor tmp = Full<T>(dev_ctx, out_dims_vector, static_cast<T>(0));
|
|
auto tmp_t = EigenTensor<T, RANK, Eigen::RowMajor>::From(tmp);
|
|
|
|
tmp_t.device(place) =
|
|
in_t.stridedSlice(starts_indices, ends_indices, steps_indices);
|
|
value_grad_t.device(place) = tmp_t.reverse(reverse_axis);
|
|
} else {
|
|
value_grad_t.device(place) =
|
|
in_t.stridedSlice(starts_indices, ends_indices, steps_indices);
|
|
}
|
|
} else {
|
|
int out_dims_size = out_dims.size();
|
|
auto value_grad_dims = value_grad->dims();
|
|
auto fake_value_grad_dims = out_dims;
|
|
|
|
// Create an extended shape according to the rules of broadcast.
|
|
auto value_grad_dims_size = value_grad_dims.size();
|
|
|
|
int num_decrease = 0;
|
|
|
|
int decrease_axis_size = decrease_axes.size();
|
|
for (int i = 0; i < out_dims_size; i++) {
|
|
if (decrease_axes.end() !=
|
|
std::find(decrease_axes.begin(), decrease_axes.end(), i)) {
|
|
fake_value_grad_dims[i] = 1;
|
|
num_decrease++;
|
|
} else if (i < out_dims_size - (value_grad_dims_size +
|
|
decrease_axis_size - num_decrease)) {
|
|
fake_value_grad_dims[i] = 1;
|
|
} else {
|
|
auto index_grad =
|
|
i - (out_dims_size -
|
|
(value_grad_dims_size + decrease_axis_size - num_decrease));
|
|
fake_value_grad_dims[i] = value_grad_dims[index_grad];
|
|
|
|
PADDLE_ENFORCE_EQ(
|
|
(out_dims[i] == value_grad_dims[index_grad]) ||
|
|
(value_grad_dims[index_grad] == 1),
|
|
true,
|
|
errors::InvalidArgument("An error occurred while calculating %s: "
|
|
"[%s] can not be accumulated into [%s].",
|
|
"ValueTensor@GRAD",
|
|
out_dims,
|
|
value_grad_dims));
|
|
}
|
|
}
|
|
|
|
VLOG(3) << "Dimensions of "
|
|
<< "ValueTensor@GRAD"
|
|
<< "([" << value_grad_dims << "])is broadcasted into ["
|
|
<< fake_value_grad_dims << "].";
|
|
|
|
auto extent = Eigen::DSizes<int64_t, RANK>();
|
|
auto offset = out_dims;
|
|
for (int i = 0; i < out_dims_size; i++) {
|
|
offset[i] = 0;
|
|
extent[i] = fake_value_grad_dims[i];
|
|
}
|
|
std::vector<DDim> offsets;
|
|
GetOffsets(out_dims, fake_value_grad_dims, offset, 0, &offsets);
|
|
|
|
auto value_grad_t = EigenTensor<T, RANK, Eigen::RowMajor>::From(
|
|
*value_grad, fake_value_grad_dims);
|
|
|
|
DenseTensor tmp = Full<T>(dev_ctx, out_dims_vector, static_cast<T>(0));
|
|
auto tmp_t = EigenTensor<T, RANK, Eigen::RowMajor>::From(tmp);
|
|
|
|
tmp_t.device(place) =
|
|
in_t.stridedSlice(starts_indices, ends_indices, steps_indices);
|
|
|
|
// accumulate gradient
|
|
for (auto offset : offsets) {
|
|
value_grad_t.device(place) =
|
|
value_grad_t + tmp_t.slice(EigenDim<RANK>::From(offset), extent);
|
|
}
|
|
if (need_reverse) {
|
|
DenseTensor tmp_value =
|
|
Full<T>(dev_ctx,
|
|
{fake_value_grad_dims.Get(), fake_value_grad_dims.size()},
|
|
static_cast<T>(0));
|
|
auto tmp_value_t =
|
|
EigenTensor<T, RANK, Eigen::RowMajor>::From(tmp_value);
|
|
tmp_value_t.device(place) = value_grad_t.reverse(reverse_axis);
|
|
value_grad_t.device(place) = tmp_value_t;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void SetValueGradKernel(const Context& dev_ctx,
|
|
const DenseTensor& out_grad,
|
|
const IntArray& starts,
|
|
const IntArray& ends,
|
|
const IntArray& steps,
|
|
const std::vector<int64_t>& axes,
|
|
const std::vector<int64_t>& decrease_axes,
|
|
const std::vector<int64_t>& none_axes,
|
|
DenseTensor* x_grad,
|
|
DenseTensor* value_grad) {
|
|
if (out_grad.numel() == 0) {
|
|
if (x_grad) dev_ctx.template Alloc<T>(x_grad);
|
|
if (value_grad) dev_ctx.template Alloc<T>(value_grad);
|
|
return;
|
|
}
|
|
const int rank = out_grad.dims().size();
|
|
std::vector<int64_t> starts_local = starts.GetData();
|
|
std::vector<int64_t> ends_local = ends.GetData();
|
|
std::vector<int64_t> steps_local = steps.GetData();
|
|
|
|
bool ellipsis_flag = true;
|
|
for (size_t i = 0; i < axes.size(); i++) {
|
|
auto idx = axes[i];
|
|
if (!(starts_local[i] == 0 && ends_local[i] == out_grad.dims()[idx] &&
|
|
steps_local[i] == 1)) {
|
|
ellipsis_flag = false;
|
|
}
|
|
}
|
|
|
|
if (ellipsis_flag) {
|
|
if (x_grad) {
|
|
Full<T, Context>(dev_ctx, x_grad->dims(), Scalar(0), x_grad);
|
|
}
|
|
if (value_grad) {
|
|
if (value_grad->numel() == out_grad.numel()) {
|
|
if (value_grad->dims() != out_grad.dims()) {
|
|
DenseTensor out_grad_temp;
|
|
ShareDataKernel<T, Context>(dev_ctx, out_grad, &out_grad_temp);
|
|
out_grad_temp.Resize(value_grad->dims());
|
|
Copy(dev_ctx, out_grad_temp, dev_ctx.GetPlace(), false, value_grad);
|
|
} else {
|
|
Copy(dev_ctx, out_grad, dev_ctx.GetPlace(), false, value_grad);
|
|
}
|
|
} else {
|
|
auto reduce_dim = funcs::GetReduceDims(out_grad, *value_grad);
|
|
SumKernel<T, Context>(
|
|
dev_ctx, out_grad, reduce_dim, out_grad.dtype(), false, value_grad);
|
|
}
|
|
}
|
|
return;
|
|
}
|
|
|
|
switch (rank) {
|
|
#define CASE_RANK(__Rk) \
|
|
case __Rk: \
|
|
SetValueGradImpl<T, Context, __Rk>(dev_ctx, \
|
|
out_grad, \
|
|
starts_local, \
|
|
ends_local, \
|
|
steps_local, \
|
|
axes, \
|
|
decrease_axes, \
|
|
none_axes, \
|
|
x_grad, \
|
|
value_grad); \
|
|
break;
|
|
CASE_RANK(1);
|
|
CASE_RANK(2);
|
|
CASE_RANK(3);
|
|
CASE_RANK(4);
|
|
CASE_RANK(5);
|
|
CASE_RANK(6);
|
|
#undef CASE_RANK
|
|
default:
|
|
PADDLE_THROW(common::errors::InvalidArgument(
|
|
"The rank of set_value_grad's input should be less than 7, but "
|
|
"received %d.",
|
|
rank));
|
|
}
|
|
return;
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void SetValueWithScalarGradKernel(const Context& dev_ctx,
|
|
const DenseTensor& out_grad,
|
|
const IntArray& starts,
|
|
const IntArray& ends,
|
|
const IntArray& steps,
|
|
const std::vector<int64_t>& axes,
|
|
const std::vector<int64_t>& decrease_axes,
|
|
const std::vector<int64_t>& none_axes,
|
|
DenseTensor* x_grad) {
|
|
SetValueGradKernel<T, Context>(dev_ctx,
|
|
out_grad,
|
|
starts,
|
|
ends,
|
|
steps,
|
|
axes,
|
|
decrease_axes,
|
|
none_axes,
|
|
x_grad,
|
|
nullptr);
|
|
}
|
|
|
|
} // namespace phi
|
|
|
|
PD_REGISTER_KERNEL(set_value_grad,
|
|
CPU,
|
|
ALL_LAYOUT,
|
|
phi::SetValueGradKernel,
|
|
float,
|
|
double,
|
|
int,
|
|
int64_t,
|
|
bool,
|
|
int16_t,
|
|
uint8_t,
|
|
int8_t,
|
|
phi::bfloat16,
|
|
phi::float16,
|
|
phi::complex64,
|
|
phi::complex128) {}
|
|
|
|
PD_REGISTER_KERNEL(set_value_with_scalar_grad,
|
|
CPU,
|
|
ALL_LAYOUT,
|
|
phi::SetValueWithScalarGradKernel,
|
|
float,
|
|
double,
|
|
int,
|
|
int64_t,
|
|
bool,
|
|
int16_t,
|
|
uint8_t,
|
|
int8_t,
|
|
phi::bfloat16,
|
|
phi::float16,
|
|
phi::complex64,
|
|
phi::complex128) {}
|