wehub-resource-sync
92 lines
2.8 KiB
C++
92 lines
2.8 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/squeeze_grad_kernel.h"
|
|
|
|
#include "paddle/phi/backends/all_context.h"
|
|
#include "paddle/phi/core/kernel_registry.h"
|
|
#include "paddle/phi/core/tensor_utils.h"
|
|
|
|
namespace phi {
|
|
template <typename T, typename Context>
|
|
void SqueezeGradKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& dout,
|
|
const IntArray& axes UNUSED,
|
|
DenseTensor* dx) {
|
|
// NOTE: [Why not to use x.dims() ?]
|
|
// Because inplace strategy is different between old IR and PIR,
|
|
// we need fix it into x.dims() after cleaning old IR system.
|
|
auto x_dims = dx->dims();
|
|
|
|
dev_ctx.template Alloc<T>(dx);
|
|
phi::Copy(dev_ctx, dout, dev_ctx.GetPlace(), false, dx);
|
|
dx->Resize(x_dims);
|
|
}
|
|
} // namespace phi
|
|
|
|
PD_REGISTER_KERNEL(squeeze_grad,
|
|
CPU,
|
|
ALL_LAYOUT,
|
|
phi::SqueezeGradKernel,
|
|
float,
|
|
double,
|
|
bool,
|
|
int,
|
|
uint8_t,
|
|
int8_t,
|
|
int16_t,
|
|
int64_t,
|
|
phi::float16,
|
|
phi::bfloat16,
|
|
phi::complex64,
|
|
phi::complex128) {}
|
|
|
|
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
|
|
PD_REGISTER_KERNEL(squeeze_grad,
|
|
GPU,
|
|
ALL_LAYOUT,
|
|
phi::SqueezeGradKernel,
|
|
float,
|
|
double,
|
|
phi::float16,
|
|
phi::bfloat16,
|
|
bool,
|
|
int,
|
|
uint8_t,
|
|
int8_t,
|
|
int16_t,
|
|
int64_t,
|
|
phi::complex64,
|
|
phi::complex128) {}
|
|
|
|
#endif
|
|
|
|
#ifdef PADDLE_WITH_XPU
|
|
PD_REGISTER_KERNEL(squeeze_grad,
|
|
XPU,
|
|
ALL_LAYOUT,
|
|
phi::SqueezeGradKernel,
|
|
float,
|
|
double,
|
|
phi::float16,
|
|
phi::bfloat16,
|
|
bool,
|
|
int,
|
|
uint8_t,
|
|
int8_t,
|
|
int64_t) {}
|
|
|
|
#endif
|