executor.h 4.7 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
Q
qijun 已提交
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

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. */

#pragma once

L
Liu Yiqun 已提交
17 18 19
#include <map>
#include <string>
#include <vector>
S
sneaxiy 已提交
20
#include "paddle/fluid/framework/garbage_collector.h"
Y
Yi Wang 已提交
21 22 23 24 25
#include "paddle/fluid/framework/op_info.h"
#include "paddle/fluid/framework/program_desc.h"
#include "paddle/fluid/framework/scope.h"
#include "paddle/fluid/framework/tensor.h"
#include "paddle/fluid/platform/device_context.h"
Q
qijun 已提交
26 27 28

namespace paddle {
namespace framework {
29
extern void InitializeVariable(Variable* var, proto::VarType::Type var_type);
Q
Qiao Longfei 已提交
30

S
sneaxiy 已提交
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70
template <typename T>
std::unordered_map<std::string, T> GetNonPersistableReferenceCount(
    const ProgramDesc& prog, size_t block_id) {
  auto& block = prog.Block(block_id);
  std::unordered_set<std::string> ignored_vars;
  std::unordered_map<std::string, T> ref_cnts;

  for (auto var_desc : block.AllVars()) {
    auto type = var_desc->Proto()->type().type();
    if (type != proto::VarType::LOD_TENSOR || var_desc->Persistable()) {
      ignored_vars.insert(var_desc->Name());  // ignore persistable vars
    }
  }

  for (auto op_desc : block.AllOps()) {
    for (auto& input : op_desc->Inputs()) {
      for (auto& input_name : input.second) {
        if (!ignored_vars.count(input_name)) {
          if (ref_cnts.count(input_name))
            ++ref_cnts[input_name];
          else
            ref_cnts[input_name] = 1;
        }
      }
    }

    for (auto& output : op_desc->Outputs()) {
      for (auto output_name : output.second) {
        if (!ignored_vars.count(output_name)) {
          if (ref_cnts.count(output_name))
            ++ref_cnts[output_name];
          else
            ref_cnts[output_name] = 1;
        }
      }
    }
  }
  return ref_cnts;
}

Q
Qiao Longfei 已提交
71 72 73 74 75 76 77
struct ExecutorPrepareContext {
  ExecutorPrepareContext(const framework::ProgramDesc& prog, size_t block_id);
  ~ExecutorPrepareContext();

  const framework::ProgramDesc& prog_;
  size_t block_id_;
  std::vector<std::unique_ptr<OperatorBase>> ops_;
S
sneaxiy 已提交
78 79

  std::unordered_map<std::string, int> ref_cnts_;
Q
Qiao Longfei 已提交
80 81
};

Q
qijun 已提交
82 83
class Executor {
 public:
D
dzhwinter 已提交
84 85
  // TODO(dzhwinter) : Do not rely on this function, it will be removed
  explicit Executor(const platform::DeviceContext& device)
D
dzhwinter 已提交
86
      : Executor(device.GetPlace()) {}
D
dzhwinter 已提交
87

D
dzhwinter 已提交
88
  explicit Executor(const platform::Place& place);
Y
Yang Yang 已提交
89

W
Wu Yi 已提交
90
  /*
Y
Yancey1989 已提交
91 92
   * Close this Executor.
   * Calling this method will send complete messages to all pserver instances.
W
Wu Yi 已提交
93
   */
Y
Yancey1989 已提交
94
  void Close();
W
Wu Yi 已提交
95

Y
Yang Yang 已提交
96 97 98 99 100 101 102
  /* @Brief
   * Runtime evaluation of the given ProgramDesc under certain Scope
   *
   * @param
   *  ProgramDesc
   *  Scope
   */
Y
Yu Yang 已提交
103 104
  void Run(const ProgramDesc& prog, Scope* scope, int block_id,
           bool create_local_scope = true, bool create_vars = true);
Q
qijun 已提交
105

X
fix  
Xin Pan 已提交
106
  // This API is very slow.
107
  void Run(const ProgramDesc& program, Scope* scope,
108 109
           std::map<std::string, const LoDTensor*>* feed_targets,
           std::map<std::string, LoDTensor*>* fetch_targets,
W
Wu Yi 已提交
110
           bool create_local_scope = true, bool create_vars = true,
111
           const std::string& feed_holder_name = "feed",
112
           const std::string& fetch_holder_name = "fetch");
113

Q
Qiao Longfei 已提交
114 115
  static std::unique_ptr<ExecutorPrepareContext> Prepare(
      const ProgramDesc& program, int block_id);
Y
Yu Yang 已提交
116

T
typhoonzero 已提交
117 118 119
  static std::vector<std::shared_ptr<ExecutorPrepareContext>> Prepare(
      const ProgramDesc& program, const std::vector<int>& block_ids);

L
Liu Yiqun 已提交
120
  void CreateVariables(const ProgramDesc& pdesc, Scope* scope, int block_id);
121

Y
Yu Yang 已提交
122 123
  void RunPreparedContext(ExecutorPrepareContext* ctx, Scope* scope,
                          bool create_local_scope = true,
Q
qiaolongfei 已提交
124
                          bool create_vars = true, bool keep_kids = false);
Y
Yu Yang 已提交
125

X
fix  
Xin Pan 已提交
126
  // This API is very slow.
127
  void RunPreparedContext(ExecutorPrepareContext* ctx, Scope* scope,
128 129
                          std::map<std::string, const LoDTensor*>* feed_targets,
                          std::map<std::string, LoDTensor*>* fetch_targets,
W
Wu Yi 已提交
130
                          bool create_local_scope = true,
L
Liu Yiqun 已提交
131
                          bool create_vars = true,
132
                          const std::string& feed_holder_name = "feed",
L
Liu Yiqun 已提交
133
                          const std::string& fetch_holder_name = "fetch");
134

135 136
  void EnableMKLDNN(const ProgramDesc& program);

Q
qijun 已提交
137
 private:
D
dzhwinter 已提交
138
  const platform::Place place_;
Q
qijun 已提交
139
};
Q
qijun 已提交
140 141 142

}  // namespace framework
}  // namespace paddle