sum_op.cc 8.6 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
2 3 4 5 6 7 8 9 10 11
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. */

Y
Yi Wang 已提交
12
#include "paddle/fluid/operators/sum_op.h"
13

14 15
#include <algorithm>
#include <string>
16
#include <vector>
17

Y
Yi Wang 已提交
18 19
#include "paddle/fluid/framework/var_type_inference.h"
#include "paddle/fluid/operators/detail/safe_ref.h"
20

21 22 23 24
#ifdef PADDLE_WITH_MKLDNN
#include "paddle/fluid/platform/mkldnn_helper.h"
#endif

25 26 27 28 29 30 31 32
namespace paddle {
namespace operators {
using framework::Tensor;

class SumOp : public framework::OperatorWithKernel {
 public:
  using framework::OperatorWithKernel::OperatorWithKernel;

33
  void InferShape(framework::InferShapeContext* ctx) const override {
Q
qiaolongfei 已提交
34
    PADDLE_ENFORCE(ctx->HasInputs("X"), "Inputs(X) should not be null");
35

Q
Qiao Longfei 已提交
36 37
    PADDLE_ENFORCE(ctx->HasOutput("Out"),
                   "Output(Out) of SumOp should not be null.");
38 39
    if (ctx->IsRuntime() &&
        ctx->GetOutputsVarType("Out")[0] ==
40
            framework::proto::VarType::LOD_TENSOR_ARRAY) {
41 42
      return;  // skip runtime infershape when is tensor array;
    }
43

44
    auto x_dims = ctx->GetInputsDim("X");
Q
Qiao Longfei 已提交
45
    size_t N = x_dims.size();
46 47 48 49
    PADDLE_ENFORCE_GT(N, 0, "Input tensors count should > 0.");
    if (N == 1) {
      VLOG(3) << "Warning: sum have only one input, may waste memory";
    }
Q
qiaolongfei 已提交
50

51 52 53 54 55 56 57 58 59 60
    framework::DDim in_dim({0});
    for (auto& x_dim : x_dims) {
      if (framework::product(x_dim) == 0) {
        continue;
      }
      if (framework::product(in_dim) == 0) {
        in_dim = x_dim;
      } else {
        PADDLE_ENFORCE_EQ(in_dim, x_dim, "Input tensors must have same shape");
      }
Q
qijun 已提交
61
    }
Q
Qiao Longfei 已提交
62 63
    ctx->SetOutputDim("Out", in_dim);
    ctx->ShareLoD("X", /*->*/ "Out");
64
  }
65 66

 protected:
67
  framework::OpKernelType GetExpectedKernelType(
68 69
      const framework::ExecutionContext& ctx) const override {
    auto x_vars = ctx.MultiInputVar("X");
C
chengduo 已提交
70
    auto x_vars_name = ctx.Inputs("X");
71 72 73 74 75 76 77 78 79 80 81 82

    framework::LibraryType library{framework::LibraryType::kPlain};
    framework::DataLayout layout{framework::DataLayout::kAnyLayout};

#ifdef PADDLE_WITH_MKLDNN
    if (library == framework::LibraryType::kPlain &&
        platform::CanMKLDNNBeUsed(ctx)) {
      library = framework::LibraryType::kMKLDNN;
      layout = framework::DataLayout::kMKLDNN;
    }
#endif

83
    if (x_vars[0]->IsType<framework::LoDTensor>()) {
84
      int dtype = -1;
C
chengduo 已提交
85 86 87
      for (size_t idx = 0; idx < x_vars.size(); ++idx) {
        PADDLE_ENFORCE(x_vars[idx] != nullptr,
                       "Input var[%s] should not be nullptr", x_vars_name[idx]);
C
chengduo 已提交
88 89
        auto tensor =
            framework::GetLoDTensorOrSelectedRowsValueFromVar(*x_vars[idx]);
90
        if (tensor->numel() == 0) {
91 92 93
          continue;
        }
        if (dtype == -1) {
94
          dtype = framework::ToDataType(tensor->type());
95
        } else {
96
          PADDLE_ENFORCE_EQ(dtype, framework::ToDataType(tensor->type()));
97 98 99 100 101
        }
      }
      PADDLE_ENFORCE_NE(dtype, -1,
                        "Sum operator should have at least one tensor");

102
      return framework::OpKernelType(
103 104
          static_cast<framework::proto::VarType::Type>(dtype), ctx.GetPlace(),
          layout, library);
105
    } else if (x_vars[0]->IsType<framework::SelectedRows>()) {
106 107 108 109
      for (auto& var : x_vars) {
        auto& value = var->Get<framework::SelectedRows>().value();
        if (value.IsInitialized()) {
          return framework::OpKernelType(framework::ToDataType(value.type()),
110
                                         ctx.device_context(), layout, library);
111 112 113 114
        }
      }
      // if input sparse vars are not initialized, use an default kernel type.
      return framework::OpKernelType(framework::proto::VarType::FP32,
115
                                     ctx.device_context(), layout, library);
116
    } else if (x_vars[0]->IsType<framework::LoDTensorArray>()) {
Y
Yang Yang(Tony) 已提交
117 118 119 120 121
      for (auto& x_var : x_vars) {
        auto& array = x_var->Get<framework::LoDTensorArray>();
        for (auto& each : array) {
          if (each.numel() != 0) {
            return framework::OpKernelType(framework::ToDataType(each.type()),
122 123
                                           ctx.device_context(), layout,
                                           library);
Y
Yang Yang(Tony) 已提交
124
          }
125 126
        }
      }
Y
Yang Yang(Tony) 已提交
127
      PADDLE_THROW("Cannot find the input data type by all input data");
128 129 130 131
    }
    PADDLE_THROW("Unexpected branch. Input type is %s",
                 x_vars[0]->Type().name());
  }
132 133 134 135
};

class SumOpMaker : public framework::OpProtoAndCheckerMaker {
 public:
Y
Yu Yang 已提交
136
  void Make() override {
137 138
    AddInput("X", "(vector<Tensor>) The input tensors of sum operator.")
        .AsDuplicable();
139
    AddOutput("Out", "(Tensor) The output tensor of sum operator.");
140 141 142
    AddAttr<bool>("use_mkldnn",
                  "(bool, default false) Only used in mkldnn kernel")
        .SetDefault(false);
143
    AddComment(R"DOC(
144
Sum operator.
145

146 147
This operators sums the input tensors. All the inputs can carry the
LoD (Level of Details) information. However, the output only shares
148
the LoD information with the first input.
149
)DOC");
150 151 152
  }
};

Q
QI JUN 已提交
153 154
class SumOpVarTypeInference : public framework::VarTypeInference {
 public:
Y
Yu Yang 已提交
155 156
  void operator()(const framework::OpDesc& op_desc,
                  framework::BlockDesc* block) const override {
Q
QI JUN 已提交
157
    auto& inputs = op_desc.Input("X");
158
    auto var_type = framework::proto::VarType::SELECTED_ROWS;
Y
Yang Yang(Tony) 已提交
159 160
    for (auto& name : op_desc.Input("X")) {
      VLOG(10) << name << " "
Y
Yang Yu 已提交
161
               << block->FindRecursiveOrCreateVar(name).GetType();
Y
Yang Yang(Tony) 已提交
162 163
    }

Q
QI JUN 已提交
164 165
    bool any_input_is_lod_tensor = std::any_of(
        inputs.begin(), inputs.end(), [block](const std::string& name) {
Y
Yang Yu 已提交
166
          return block->FindRecursiveOrCreateVar(name).GetType() ==
167
                 framework::proto::VarType::LOD_TENSOR;
Q
QI JUN 已提交
168
        });
169 170

    auto is_tensor_array = [block](const std::string& name) {
Y
Yang Yu 已提交
171
      return block->FindRecursiveOrCreateVar(name).GetType() ==
172
             framework::proto::VarType::LOD_TENSOR_ARRAY;
173 174 175 176 177 178 179 180
    };

    bool any_input_is_tensor_array =
        std::any_of(inputs.begin(), inputs.end(), is_tensor_array);
    bool all_inputs_are_tensor_array =
        std::all_of(inputs.begin(), inputs.end(), is_tensor_array);

    if (any_input_is_tensor_array) {
Y
Yang Yang(Tony) 已提交
181 182 183 184
      if (!all_inputs_are_tensor_array) {
        std::ostringstream os;
        for (auto& each : inputs) {
          os << "    " << each << " type is "
Y
Yang Yu 已提交
185
             << block->FindRecursiveOrCreateVar(each).GetType() << "\n";
Y
Yang Yang(Tony) 已提交
186 187 188 189
        }
        PADDLE_ENFORCE(all_inputs_are_tensor_array,
                       "Not all inputs are tensor array:\n%s", os.str());
      }
190
      var_type = framework::proto::VarType::LOD_TENSOR_ARRAY;
191
    } else if (any_input_is_lod_tensor) {
192
      var_type = framework::proto::VarType::LOD_TENSOR;
Q
QI JUN 已提交
193 194 195
    }

    auto out_var_name = op_desc.Output("Out").front();
Y
Yang Yu 已提交
196
    auto& out_var = block->FindRecursiveOrCreateVar(out_var_name);
Y
Yang Yang(Tony) 已提交
197 198 199
    out_var.SetType(var_type);
    auto& in_var = detail::Ref(block->FindVarRecursive(inputs.front()));
    out_var.SetDataType(in_var.GetDataType());
Q
QI JUN 已提交
200 201 202
  }
};

203
class SumGradMaker : public framework::GradOpDescMakerBase {
204
 public:
205
  using framework::GradOpDescMakerBase::GradOpDescMakerBase;
206

Y
Yu Yang 已提交
207
  std::vector<std::unique_ptr<framework::OpDesc>> operator()() const override {
208
    auto x_grads = InputGrad("X", false);
Y
Yu Yang 已提交
209
    std::vector<std::unique_ptr<framework::OpDesc>> grad_ops;
210 211 212 213
    grad_ops.reserve(x_grads.size());
    auto og = OutputGrad("Out");
    std::transform(x_grads.begin(), x_grads.end(), std::back_inserter(grad_ops),
                   [&og](const std::string& x_grad) {
Y
Yu Yang 已提交
214
                     auto* grad_op = new framework::OpDesc();
Y
Yu Yang 已提交
215 216 217 218
                     grad_op->SetType("scale");
                     grad_op->SetInput("X", og);
                     grad_op->SetOutput("Out", {x_grad});
                     grad_op->SetAttr("scale", 1.0f);
Y
Yu Yang 已提交
219
                     return std::unique_ptr<framework::OpDesc>(grad_op);
220 221
                   });
    return grad_ops;
222 223 224 225 226 227 228
  }
};

}  // namespace operators
}  // namespace paddle

namespace ops = paddle::operators;
229

Q
QI JUN 已提交
230 231
REGISTER_OPERATOR(sum, ops::SumOp, ops::SumOpMaker, ops::SumGradMaker,
                  ops::SumOpVarTypeInference);
232

Q
QI JUN 已提交
233 234 235 236 237
REGISTER_OP_CPU_KERNEL(
    sum, ops::SumKernel<paddle::platform::CPUDeviceContext, float>,
    ops::SumKernel<paddle::platform::CPUDeviceContext, double>,
    ops::SumKernel<paddle::platform::CPUDeviceContext, int>,
    ops::SumKernel<paddle::platform::CPUDeviceContext, int64_t>);