grad_op_desc_maker.h 6.4 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
2

L
Luo Tao 已提交
3 4 5
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
6

L
Luo Tao 已提交
7
    http://www.apache.org/licenses/LICENSE-2.0
8

L
Luo Tao 已提交
9 10 11 12 13
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. */
14 15

#pragma once
16
#include <algorithm>
17 18
#include <string>
#include <unordered_set>
Y
Yu Yang 已提交
19
#include <vector>
Y
Yi Wang 已提交
20 21
#include "paddle/fluid/framework/op_desc.h"
#include "paddle/fluid/framework/operator.h"
22 23 24 25

namespace paddle {
namespace framework {

26 27 28 29 30 31 32 33
/*
  This functor class is responsible for creating the gradient ops for the given
  operator fwd_op. After it is called (through operator()), the pairs of
  (gradient variable, corresponding input variable of fwd_op) will be added to
  grad_to_var. If an input variable of fwd_op is contained in no_grad_set, its
  gradient varialbe will be ignored or kEmptyVarName depending on the template
  argument DropEmptyIG in the derived classes.
 */
34 35
class GradOpDescMakerBase {
 public:
36
  explicit GradOpDescMakerBase(
Y
Yu Yang 已提交
37
      const OpDesc& fwd_op, const std::unordered_set<std::string>& no_grad_set,
Y
Yu Yang 已提交
38
      std::unordered_map<std::string, std::string>* grad_to_var,
Y
Yu Yang 已提交
39
      const std::vector<BlockDesc*>& grad_block = std::vector<BlockDesc*>())
Y
Yu Yang 已提交
40 41 42 43
      : fwd_op_(fwd_op),
        no_grad_set_(no_grad_set),
        grad_to_var_(grad_to_var),
        grad_block_(grad_block) {}
44 45

  virtual ~GradOpDescMakerBase() = default;
Y
Yu Yang 已提交
46
  virtual std::vector<std::unique_ptr<OpDesc>> operator()() const = 0;
47 48

 protected:
49 50
  std::vector<std::string> InputGrad(const std::string& name,
                                     bool drop_empty_grad = true) const {
51
    std::vector<std::string> ret_val;
52
    auto var_names = this->Input(name);
53
    ret_val.reserve(var_names.size());
54 55 56 57 58 59 60 61 62 63 64
    std::transform(var_names.begin(), var_names.end(),
                   std::back_inserter(ret_val),
                   [this](const std::string& fwd_var_name) -> std::string {
                     auto g_name = GradVarName(fwd_var_name);
                     if (no_grad_set_.count(g_name)) {
                       return kEmptyVarName;
                     } else {
                       (*this->grad_to_var_)[g_name] = fwd_var_name;
                       return g_name;
                     }
                   });
65 66 67
    if (!drop_empty_grad) {
      return ret_val;
    }
68 69 70 71 72
    PADDLE_ENFORCE_LE(var_names.size(), 1UL,
                      "BUG from operator developer:"
                      " for input argument with a list of variables, "
                      " drop_empty_grad is not allowed because it makes"
                      " the correspondence bewteen a variable and its gradient"
73
                      " ambiguous."
74 75 76
                      " Op type %s",
                      fwd_op_.Type());

77 78 79 80 81 82
    std::vector<std::string> dropped_ret_val;
    dropped_ret_val.reserve(ret_val.size());
    std::copy_if(ret_val.begin(), ret_val.end(),
                 std::back_inserter(dropped_ret_val),
                 [](const std::string& str) { return str != kEmptyVarName; });
    return dropped_ret_val;
83 84 85
  }

  std::vector<std::string> OutputGrad(const std::string& name) const {
86 87 88 89
    std::vector<std::string> ret_val;
    auto onames = this->Output(name);
    ret_val.reserve(onames.size());
    std::transform(onames.begin(), onames.end(), std::back_inserter(ret_val),
90 91 92 93 94
                   [this](const std::string& fwd_var_name) -> std::string {
                     auto g_name = GradVarName(fwd_var_name);
                     (*this->grad_to_var_)[g_name] = fwd_var_name;
                     return g_name;
                   });
95
    return ret_val;
96 97
  }

Y
Yu Yang 已提交
98 99
  std::vector<std::string> InputNames() const {
    return this->fwd_op_.InputNames();
100 101
  }

Y
Yu Yang 已提交
102 103
  std::vector<std::string> OutputNames() const {
    return this->fwd_op_.OutputNames();
104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124
  }

  std::vector<std::string> Input(const std::string& name) const {
    return fwd_op_.Input(name);
  }

  std::vector<std::string> Output(const std::string& name) const {
    return fwd_op_.Output(name);
  }

  const std::unordered_map<std::string, Attribute>& Attrs() const {
    return fwd_op_.GetAttrMap();
  }

  const Attribute& GetAttr(const std::string& name) const {
    auto& map = fwd_op_.GetAttrMap();
    auto it = map.find(name);
    PADDLE_ENFORCE(it != map.end(), "Cannot find attribute %s", name);
    return it->second;
  }

E
emailweixu 已提交
125 126 127 128 129
  template <typename T>
  inline const T& Attr(const std::string& name) const {
    return boost::get<T>(GetAttr(name));
  }

130 131 132
  std::string ForwardOpType() const { return this->fwd_op_.Type(); }

 private:
Y
Yu Yang 已提交
133
  const OpDesc& fwd_op_;
134
  const std::unordered_set<std::string>& no_grad_set_;
135
  std::unordered_map<std::string, std::string>* grad_to_var_;
Y
Yu Yang 已提交
136 137

 protected:
Y
Yu Yang 已提交
138
  std::vector<BlockDesc*> grad_block_;
139 140 141 142
};

class SingleGradOpDescMaker : public GradOpDescMakerBase {
 public:
Y
Yu Yang 已提交
143 144
  using GradOpDescMakerBase::GradOpDescMakerBase;

Y
Yu Yang 已提交
145 146
  std::vector<std::unique_ptr<OpDesc>> operator()() const {
    std::vector<std::unique_ptr<OpDesc>> retv;
Y
Yu Yang 已提交
147 148 149
    retv.emplace_back(this->Apply());
    return retv;
  }
150 151

 protected:
Y
Yu Yang 已提交
152
  virtual std::unique_ptr<OpDesc> Apply() const = 0;
153 154
};

155
template <bool DropEmptyIG = true>
156
class DefaultGradOpDescMaker : public SingleGradOpDescMaker {
Y
Yu Yang 已提交
157 158 159
 public:
  using SingleGradOpDescMaker::SingleGradOpDescMaker;

160
 protected:
Y
Yu Yang 已提交
161 162
  virtual std::unique_ptr<OpDesc> Apply() const {
    auto* grad = new OpDesc();
Y
Yu Yang 已提交
163
    grad->SetType(this->GradOpType());
164

Y
Yu Yang 已提交
165
    for (auto& input_param : this->InputNames()) {
Y
Yu Yang 已提交
166
      grad->SetInput(input_param, this->Input(input_param));
167 168
      grad->SetOutput(GradVarName(input_param),
                      this->InputGrad(input_param, DropEmptyIG));
169 170
    }

Y
Yu Yang 已提交
171
    for (auto& output_param : this->OutputNames()) {
Y
Yu Yang 已提交
172 173
      grad->SetInput(output_param, this->Output(output_param));
      grad->SetInput(GradVarName(output_param), this->OutputGrad(output_param));
174 175
    }

Y
Yu Yang 已提交
176
    grad->SetAttrMap(this->Attrs());
177

Y
Yu Yang 已提交
178
    return std::unique_ptr<OpDesc>(grad);
179 180 181 182 183 184 185
  }

  virtual std::string GradOpType() const {
    return this->ForwardOpType() + "_grad";
  }
};

Y
Yu Yang 已提交
186 187 188
class EmptyGradOpMaker : public GradOpDescMakerBase {
 public:
  using GradOpDescMakerBase::GradOpDescMakerBase;
Y
Yu Yang 已提交
189
  std::vector<std::unique_ptr<OpDesc>> operator()() const override {
Y
Yu Yang 已提交
190 191 192 193
    return {};
  }
};

194 195
}  // namespace framework
}  // namespace paddle