squeeze_kernel.cc 4.3 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
// 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_kernel.h"

#include "paddle/phi/backends/all_context.h"
#include "paddle/phi/core/kernel_registry.h"
#include "paddle/phi/core/tensor_utils.h"
#include "paddle/phi/kernels/funcs/unsqueeze.h"

namespace phi {
template <typename T, typename Context>
24 25
void SqueezeInferKernel(const Context& dev_ctx,
                        const DenseTensor& x,
26
                        const IntArray& axes UNUSED,
27
                        DenseTensor* out) {
H
heliqi 已提交
28
  auto out_dims = out->dims();
29
  dev_ctx.template Alloc<T>(out);
Z
zhupengyang 已提交
30 31 32
  if (x.Holder() == out->Holder()) {
    return;
  }
33 34 35 36 37
  phi::Copy(dev_ctx, x, dev_ctx.GetPlace(), false, out);
  out->Resize(out_dims);  // copy will reset the dims.
}

template <typename T, typename Context>
38 39 40 41
void SqueezeKernel(const Context& dev_ctx,
                   const DenseTensor& x,
                   const IntArray& axes,
                   DenseTensor* out,
42
                   DenseTensor* xshape UNUSED) {
43
  SqueezeInferKernel<T, Context>(dev_ctx, x, axes, out);
44 45 46 47
}

}  // namespace phi

48
PD_REGISTER_KERNEL(squeeze_infer,
49 50
                   CPU,
                   ALL_LAYOUT,
51
                   phi::SqueezeInferKernel,
52 53 54 55 56 57 58 59 60 61 62
                   float,
                   double,
                   phi::dtype::bfloat16,
                   bool,
                   int,
                   uint8_t,
                   int8_t,
                   int64_t,
                   phi::dtype::complex<float>,
                   phi::dtype::complex<double>) {}

63
PD_REGISTER_KERNEL(squeeze,
64 65
                   CPU,
                   ALL_LAYOUT,
66
                   phi::SqueezeKernel,
67 68 69 70 71 72 73 74 75 76 77
                   float,
                   double,
                   phi::dtype::bfloat16,
                   bool,
                   int,
                   uint8_t,
                   int8_t,
                   int64_t,
                   phi::dtype::complex<float>,
                   phi::dtype::complex<double>) {}
#if defined(PADDLE_WITH_CUDA) || defined(PADDLE_WITH_HIP)
78
PD_REGISTER_KERNEL(squeeze_infer,
79 80
                   GPU,
                   ALL_LAYOUT,
81
                   phi::SqueezeInferKernel,
82 83 84 85 86 87 88 89 90 91 92 93
                   float,
                   double,
                   phi::dtype::float16,
                   phi::dtype::bfloat16,
                   bool,
                   int,
                   uint8_t,
                   int8_t,
                   int64_t,
                   phi::dtype::complex<float>,
                   phi::dtype::complex<double>) {}

94
PD_REGISTER_KERNEL(squeeze,
95 96
                   GPU,
                   ALL_LAYOUT,
97
                   phi::SqueezeKernel,
98 99 100 101 102 103 104 105 106 107 108 109 110 111
                   float,
                   double,
                   phi::dtype::float16,
                   phi::dtype::bfloat16,
                   bool,
                   int,
                   uint8_t,
                   int8_t,
                   int64_t,
                   phi::dtype::complex<float>,
                   phi::dtype::complex<double>) {}
#endif

#ifdef PADDLE_WITH_XPU
112
PD_REGISTER_KERNEL(squeeze_infer,
113 114
                   XPU,
                   ALL_LAYOUT,
115
                   phi::SqueezeInferKernel,
116 117 118 119 120 121 122 123 124
                   float,
                   double,
                   phi::dtype::float16,
                   bool,
                   int,
                   uint8_t,
                   int8_t,
                   int64_t) {}

125
PD_REGISTER_KERNEL(squeeze,
126 127
                   XPU,
                   ALL_LAYOUT,
128
                   phi::SqueezeKernel,
129 130 131 132 133 134 135 136 137
                   float,
                   double,
                   phi::dtype::float16,
                   bool,
                   int,
                   uint8_t,
                   int8_t,
                   int64_t) {}
#endif