conv_fusion_op.cc 5.3 KB
Newer Older
Q
qingqing01 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
/* Copyright (c) 2016 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 <string>
#include <vector>
#include "paddle/fluid/operators/conv_op.h"
18
#include "paddle/fluid/platform/device/gpu/gpu_dnn.h"
Q
qingqing01 已提交
19 20 21 22 23 24 25 26 27 28

namespace paddle {
namespace operators {

// This fused conv follows the equation:
//   y = act ( alpha1 * conv(x) + alpha2 * z + bias ).
//   here, y is Output,
//         x is Input,
//         z is ResidualData,
//         bias is Bias
T
tianshuo78520a 已提交
29
// When `split_channels` is set, y will be split into multiple outputs,
Q
qingqing01 已提交
30
// each output has split_channels[i] number of channels.
Q
qingqing01 已提交
31 32 33 34 35 36 37 38
class Conv2DFusionOpMaker : public Conv2DOpMaker {
 protected:
  void Apply() override {
    AddAttr<std::string>(
        "activation",
        "The activation type can be 'identity', 'sigmoid', 'relu', 'relu6' "
        "'relux' , 'tanh', 'band_pass'")
        .SetDefault("relu");
Q
qingqing01 已提交
39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
    AddAttr<std::vector<int>>(
        "split_channels",
        "When `split_channels` are set, there will be multiple outputs, the "
        "output size is equal to the number of `split_channels`.")
        .SetDefault({});
    AddOutput("Outputs",
              "This Outputs is used when setting `split_channels`."
              "Usually used to fuse conv with same input and same filter size, "
              "padding, stride, dilation size.")
        .AsDuplicable()
        .AsDispensable();
    AddInput("AlgoCache",
             "The cache of convolution algorithm, a RAW type variable.")
        .AsDispensable();
    AddAttr<int>(
        "search_times",
        "The number of exhaustive search times for convolution algorithm.")
        .SetDefault(-1);
Q
qingqing01 已提交
57 58
  }
};
Q
qingqing01 已提交
59

Z
Zeng Jinle 已提交
60
class Conv2DFusionOp : public operators::ConvOp {
Q
qingqing01 已提交
61
 public:
Z
Zeng Jinle 已提交
62 63 64 65
  using operators::ConvOp::ConvOp;

 protected:
  void InferShape(framework::InferShapeContext* ctx) const override {
66 67
    OP_INOUT_CHECK(ctx->HasInput("Input"), "Input", "Input", "Conv2DFusion");
    OP_INOUT_CHECK(ctx->HasInput("Bias"), "Input", "Bias", "Conv2DFusion");
68

69
    auto in_dims = ctx->GetInputDim("Input");
70
    PADDLE_ENFORCE_EQ(
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
        in_dims.size(), 4U,
        platform::errors::InvalidArgument(
            "The input's dimension of Operator(Conv2DFusion) is expected "
            "to be 4. But received: input's dimension = %u, shape = [%s].",
            in_dims.size(), in_dims));

    // In some case, attribute data_format is "AnyLayout".
    std::string data_format = ctx->Attrs().Get<std::string>("data_format");
    PADDLE_ENFORCE_NE(
        data_format, "NHWC",
        platform::errors::PermissionDenied(
            "Operator(Conv2DFusion) only supports data format of "
            "channel first (NCHW) now. But recieved: data_format = '%s'.",
            data_format));

    std::vector<int64_t> output_shape = ComputeOutputShape(ctx);
87
    ctx->SetOutputDim("Output", framework::make_ddim(output_shape));
88
    ctx->ShareLoD("Input", "Output");
89

90
    std::vector<int> split_channels =
Q
qingqing01 已提交
91
        ctx->Attrs().Get<std::vector<int>>("split_channels");
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111
    if (split_channels.size()) {
      OP_INOUT_CHECK(ctx->HasOutputs("Outputs"), "Output", "Outputs",
                     "Conv2DFusion");
      PADDLE_ENFORCE_EQ(
          ctx->Outputs("Outputs").size(), split_channels.size(),
          platform::errors::InvalidArgument(
              "The number of Output(Outputs) of operator 'Conv2DFusion' is "
              "expected to be equal to the length of Attr(split_channels). But "
              "reiceved: the number of Output(Outputs) = %u; the length of "
              "Attr(split_channels) = %u, the content = [%s].",
              ctx->Outputs("Outputs").size(), split_channels.size(),
              framework::make_ddim(split_channels)));

      int split_channels_sum = 0;
      std::vector<framework::DDim> output_shapes(split_channels.size());
      for (size_t i = 0; i < split_channels.size(); ++i) {
        split_channels_sum += split_channels[i];
        output_shapes[i] =
            framework::make_ddim({output_shape[0], split_channels[i],
                                  output_shape[2], output_shape[3]});
Q
qingqing01 已提交
112
      }
113 114 115 116 117 118 119 120 121
      PADDLE_ENFORCE_EQ(
          split_channels_sum, output_shape[1],
          platform::errors::InvalidArgument(
              "The sum of Attr(split_channels) is expected to be equal to the "
              "total output channels. But recieved: the sum of "
              "Attr(split_channels) = %d, the total output channels = %d.",
              split_channels_sum, output_shape[1]));

      ctx->SetOutputsDim("Outputs", output_shapes);
Q
qingqing01 已提交
122 123 124 125
    }
  }
};

Q
qingqing01 已提交
126 127 128 129 130 131
// TODO(qingqing): add gradient operator for conv2d_fusion

}  // namespace operators
}  // namespace paddle

namespace ops = paddle::operators;
H
hong 已提交
132
REGISTER_OPERATOR(
Z
Zeng Jinle 已提交
133 134
    conv2d_fusion, ops::Conv2DFusionOp, ops::Conv2DFusionOpMaker,
    ops::ConvOpInferVarType,
H
hong 已提交
135 136
    paddle::framework::EmptyGradOpMaker<paddle::framework::OpDesc>,
    paddle::framework::EmptyGradOpMaker<paddle::imperative::OpBase>);