operator.cc 13.8 KB
Newer Older
Q
Qiao Longfei 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.

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. */

15
#include <algorithm>
T
tensor-tang 已提交
16
#include <atomic>
D
dzhwinter 已提交
17 18

#include "paddle/framework/executor.h"
19
#include "paddle/framework/lod_tensor_array.h"
D
dzhwinter 已提交
20
#include "paddle/framework/operator.h"
21
#include "paddle/framework/shape_inference.h"
22
#include "paddle/framework/var_type.h"
Q
Qiao Longfei 已提交
23 24 25 26

namespace paddle {
namespace framework {

27
std::string OperatorBase::Input(const std::string& name) const {
Y
Yu Yang 已提交
28
  auto& ins = Inputs(name);
Y
Yu Yang 已提交
29
  PADDLE_ENFORCE_LE(ins.size(), 1UL,
30 31
                    "Operator %s's input %s should contain only one variable.",
                    type_, name);
Y
Yu Yang 已提交
32
  return ins.empty() ? kEmptyVarName : ins[0];
Y
Yan Chunwei 已提交
33 34
}

Y
Yu Yang 已提交
35 36
const std::vector<std::string>& OperatorBase::Inputs(
    const std::string& name) const {
Y
Yu Yang 已提交
37
  auto it = inputs_.find(name);
38 39
  PADDLE_ENFORCE(it != inputs_.end(), "Operator %s does not have the input %s.",
                 type_, name);
Y
Yu Yang 已提交
40
  return it->second;
Y
Yan Chunwei 已提交
41 42
}

43
std::string OperatorBase::Output(const std::string& name) const {
Y
Yu Yang 已提交
44
  auto& outs = Outputs(name);
Y
Yu Yang 已提交
45
  PADDLE_ENFORCE_LE(outs.size(), 1UL,
46 47
                    "Operator %s's output %s should contain only one variable.",
                    type_, name);
Y
Yu Yang 已提交
48
  return outs.empty() ? kEmptyVarName : outs[0];
Y
Yan Chunwei 已提交
49 50
}

Y
Yu Yang 已提交
51 52
const std::vector<std::string>& OperatorBase::Outputs(
    const std::string& name) const {
Y
Yu Yang 已提交
53
  auto it = outputs_.find(name);
54 55
  PADDLE_ENFORCE(it != outputs_.end(),
                 "Operator %s does not have an output called %s.", type_, name);
Y
Yu Yang 已提交
56
  return it->second;
Y
Yan Chunwei 已提交
57 58
}

Q
Qiao Longfei 已提交
59 60
std::string OperatorBase::DebugString() const {
  std::stringstream ss;
Y
Yu Yang 已提交
61
  ss << "Op(" << type_ << "), inputs:{";
Y
Yu Yang 已提交
62 63
  for (auto it = inputs_.begin(); it != inputs_.end();) {
    auto& input = *it;
Y
Yu Yang 已提交
64 65 66 67 68 69
    ss << input.first << "[";
    for (size_t i = 0; i < input.second.size(); ++i) {
      ss << input.second[i];
      if (i != input.second.size() - 1) {
        ss << ", ";
      }
70
    }
Y
Yu Yang 已提交
71
    ss << "]";
Y
Yu Yang 已提交
72 73
    ++it;
    if (it != inputs_.end()) {
74 75
      ss << ", ";
    }
Q
Qiao Longfei 已提交
76
  }
Y
Yu Yang 已提交
77
  ss << "}, outputs:{";
Y
Yu Yang 已提交
78 79
  for (auto it = outputs_.begin(); it != outputs_.end();) {
    auto& output = *it;
Y
Yu Yang 已提交
80 81 82 83 84 85
    ss << output.first << "[";
    for (size_t i = 0; i < output.second.size(); ++i) {
      ss << output.second[i];
      if (i != output.second.size() - 1) {
        ss << ", ";
      }
86
    }
Y
Yu Yang 已提交
87
    ss << "]";
Y
Yu Yang 已提交
88 89
    ++it;
    if (it != outputs_.end()) {
90 91
      ss << ", ";
    }
Q
Qiao Longfei 已提交
92
  }
Y
Yu Yang 已提交
93
  ss << "}.";
Q
Qiao Longfei 已提交
94 95 96
  return ss.str();
}

D
dongzhihong 已提交
97 98
void OperatorBase::Rename(const std::string& old_name,
                          const std::string& new_name) {
Y
Yu Yang 已提交
99 100 101 102 103 104 105
  for (auto& input : inputs_) {
    std::replace(input.second.begin(), input.second.end(), old_name, new_name);
  }
  for (auto& output : outputs_) {
    std::replace(output.second.begin(), output.second.end(), old_name,
                 new_name);
  }
D
dongzhihong 已提交
106 107
}

Y
Yu Yang 已提交
108
OperatorBase::OperatorBase(const std::string& type,
Y
Yu Yang 已提交
109 110
                           const VariableNameMap& inputs,
                           const VariableNameMap& outputs,
Y
Yu Yang 已提交
111 112
                           const AttributeMap& attrs)
    : type_(type), inputs_(inputs), outputs_(outputs), attrs_(attrs) {
113 114
  GenerateTemporaryNames();
  CheckAllInputOutputSet();
Y
Yu Yang 已提交
115
}
116

Q
qijun 已提交
117 118
std::vector<std::string> OperatorBase::InputVars() const {
  std::vector<std::string> ret_val;
Y
Yu Yang 已提交
119
  for (auto& o : inputs_) {
Q
qijun 已提交
120 121 122 123 124 125
    ret_val.reserve(ret_val.size() + o.second.size());
    ret_val.insert(ret_val.end(), o.second.begin(), o.second.end());
  }
  return ret_val;
}

Y
Yu Yang 已提交
126 127 128 129 130 131 132 133 134 135
std::vector<std::string> OperatorBase::OutputVars(bool has_intermediate) const {
  std::vector<std::string> ret_val;
  if (has_intermediate) {
    // push all outputs into ret_val
    for (auto& o : outputs_) {
      ret_val.reserve(ret_val.size() + o.second.size());
      ret_val.insert(ret_val.end(), o.second.begin(), o.second.end());
    }
    return ret_val;
  }
Y
Yu Yang 已提交
136
  auto& info = OpInfoMap::Instance().Get(Type());
Y
Yu Yang 已提交
137 138

  // get all OpProto::Var for outputs
Y
Yu Yang 已提交
139
  for (auto& o : info.Proto().outputs()) {
Y
Yu Yang 已提交
140 141 142 143 144 145 146 147 148
    // ignore all intermediate output
    if (o.intermediate()) continue;
    auto out = outputs_.find(o.name());
    if (out != outputs_.end()) {
      ret_val.reserve(ret_val.size() + out->second.size());
      ret_val.insert(ret_val.end(), out->second.begin(), out->second.end());
    }
  }
  return ret_val;
D
dongzhihong 已提交
149 150
}

151 152 153
void OperatorBase::CheckAllInputOutputSet() const {
  auto& info_map = OpInfoMap::Instance();
  auto* op_info = info_map.GetNullable(Type());
Y
Yu Yang 已提交
154
  if (op_info == nullptr || op_info->proto_ == nullptr) return;
155 156 157

  for (auto& in : op_info->Proto().inputs()) {
    PADDLE_ENFORCE(inputs_.find(in.name()) != inputs_.end(),
Y
Yu Yang 已提交
158
                   "Type %s's input %s is not set", Type(), in.name());
159 160 161 162
  }

  for (auto& out : op_info->Proto().outputs()) {
    PADDLE_ENFORCE(outputs_.find(out.name()) != outputs_.end(),
Y
Yu Yang 已提交
163
                   "Type %s's output %s is not set", Type(), out.name());
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179
  }
}

void OperatorBase::GenerateTemporaryNames() {
  static std::atomic<size_t> gUniqId(0UL);
  for (auto& output : outputs_) {
    for (auto& output_name : output.second) {
      if (output_name == kTempVarName) {
        output_name += type_;
        output_name += "@";
        output_name += std::to_string(gUniqId.fetch_add(1));
      }
    }
  }
}

Q
QI JUN 已提交
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
static const Tensor* GetTensorFromVar(const Variable* var) {
  const Tensor* t = nullptr;
  if (var->IsType<LoDTensor>()) {
    t = &(var->Get<LoDTensor>());
  } else if (var->IsType<SelectedRows>()) {
    t = &(var->Get<SelectedRows>().value());
  } else {
    PADDLE_THROW("Variable type must be LoDTensor/SelectedRows.");
  }
  return t;
}

static Tensor* GetMutableTensorFromVar(Variable* var) {
  Tensor* t = nullptr;
  if (var->IsType<LoDTensor>()) {
    t = var->GetMutable<LoDTensor>();
  } else if (var->IsType<SelectedRows>()) {
    t = var->GetMutable<SelectedRows>()->mutable_value();
  } else {
    PADDLE_THROW("Variable type must be LoDTensor/SelectedRows.");
  }
  return t;
}

204
template <>
205
const Tensor* ExecutionContext::Input<Tensor>(const std::string& name) const {
206
  auto* var = InputVar(name);
207
  return var == nullptr ? nullptr : GetTensorFromVar(var);
208 209 210
}

template <>
211
const std::vector<const Tensor*> ExecutionContext::MultiInput<Tensor>(
212 213 214 215
    const std::string& name) const {
  auto names = op().Inputs(name);
  std::vector<const Tensor*> res;
  res.reserve(names.size());
216 217 218 219 220
  std::transform(names.begin(), names.end(), std::back_inserter(res),
                 [&](const std::string& sub_name) {
                   auto var = scope_.FindVar(sub_name);
                   return var == nullptr ? nullptr : GetTensorFromVar(var);
                 });
221 222 223 224
  return res;
}

template <>
225
Tensor* ExecutionContext::Output<Tensor>(const std::string& name) const {
226
  auto var = OutputVar(name);
Q
QI JUN 已提交
227
  return var == nullptr ? nullptr : GetMutableTensorFromVar(var);
228 229 230
}

template <>
231
std::vector<Tensor*> ExecutionContext::MultiOutput<Tensor>(
232 233 234 235
    const std::string& name) const {
  auto names = op().Outputs(name);
  std::vector<Tensor*> res;
  res.reserve(names.size());
236 237
  std::transform(names.begin(), names.end(), std::back_inserter(res),
                 [&](const std::string& sub_name) {
238 239
                   auto var = scope_.FindVar(sub_name);
                   return var == nullptr ? nullptr
Q
QI JUN 已提交
240
                                         : GetMutableTensorFromVar(var);
241
                 });
242 243 244
  return res;
}

Y
Yu Yang 已提交
245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
bool OpSupportGPU(const std::string& op_type) {
  auto& all_kernels = OperatorWithKernel::AllOpKernels();
  auto it = all_kernels.find(op_type);
  if (it == all_kernels.end()) {
    // All control operator must support GPU
    return true;
  }
  for (auto& kern_pair : it->second) {
    if (platform::is_gpu_place(kern_pair.first.place_)) {
      return true;
    }
  }
  return false;
}

260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
class RuntimeInferShapeContext : public InferShapeContext {
 public:
  RuntimeInferShapeContext(const OperatorBase& op, const Scope& scope)
      : op_(op), scope_(scope) {}

  bool HasInput(const std::string& name) const override {
    auto& ins = Inputs(name);
    size_t length = ins.size();
    if (length == 0) {
      return false;
    }
    PADDLE_ENFORCE_EQ(length, 1UL, "Input %s should have more than one inputs",
                      name);
    auto ipt = ins[0];
    auto* var = ipt == kEmptyVarName ? nullptr : scope_.FindVar(ipt);
    return var != nullptr;
  }

  bool HasOutput(const std::string& name) const override {
    auto& outs = Outputs(name);
    size_t length = outs.size();
    if (length == 0) {
      return false;
    }
    PADDLE_ENFORCE_EQ(length, 1UL, "Output %s should have more than one inputs",
                      name);
    auto ipt = outs[0];
    auto* var = ipt == kEmptyVarName ? nullptr : scope_.FindVar(ipt);
    return var != nullptr;
  }

  bool HasInputs(const std::string& name) const override {
    auto inputs = op_.Inputs(name);
    if (inputs.empty()) {
      return false;
    }
    for (auto& input : inputs) {
      if (scope_.FindVar(input) == nullptr) {
        return false;
      }
    }
    return true;
  }

  bool HasOutputs(const std::string& name) const override {
    auto outputs = op_.Outputs(name);
    if (outputs.empty()) {
      return false;
    }
    for (auto& output : outputs) {
      if (scope_.FindVar(output) == nullptr) {
        return false;
      }
    }
    return true;
  }

  DDim GetInputDim(const std::string& name) const override {
    return GetDim(op_.Input(name));
  }

  void SetOutputDim(const std::string& name, const DDim& dim) override {
    SetDim(op_.Output(name), dim);
  }

  AttrReader Attrs() const override { return AttrReader(op_.Attrs()); }

  const std::vector<std::string>& Inputs(
      const std::string& name) const override {
    return op_.Inputs(name);
  }

  const std::vector<std::string>& Outputs(
      const std::string& name) const override {
    return op_.Outputs(name);
  }

Q
Qiao Longfei 已提交
337 338 339 340 341 342 343 344 345 346 347 348 349 350
  void ShareLoD(const std::string& in, const std::string& out, size_t i = 0,
                size_t j = 0) const override {
    PADDLE_ENFORCE_LT(i, Inputs(in).size());
    PADDLE_ENFORCE_LT(j, Outputs(out).size());
    Variable* in_var = scope_.FindVar(Inputs(in)[i]);
    Variable* out_var = scope_.FindVar(Outputs(out)[j]);
    if (!in_var->IsType<LoDTensor>()) return;
    PADDLE_ENFORCE(out_var->IsType<LoDTensor>(),
                   "The %d-th output of Output(%s) must be LoDTensor.", j, out);
    auto in_tensor = in_var->Get<LoDTensor>();
    auto* out_tensor = out_var->GetMutable<LoDTensor>();
    out_tensor->set_lod(in_tensor.lod());
  }

351 352 353
  bool IsRuntime() const override { return true; }

 protected:
354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375
  DDim GetDim(const std::string& name) const override {
    Variable* var = scope_.FindVar(name);
    if (var->IsType<LoDTensor>()) {
      return var->Get<LoDTensor>().dims();
    } else if (var->IsType<SelectedRows>()) {
      return var->Get<SelectedRows>().GetCompleteDims();
    } else {
      PADDLE_THROW("Variable type must be LoDTensor/SelectedRows.");
    }
  }

  void SetDim(const std::string& name, const DDim& dim) override {
    Variable* var = scope_.FindVar(name);
    if (var->IsType<LoDTensor>()) {
      var->GetMutable<LoDTensor>()->Resize(dim);
    } else if (var->IsType<SelectedRows>()) {
      var->GetMutable<SelectedRows>()->set_height(dim[0]);
    } else {
      PADDLE_THROW("Variable type must be LoDTensor/SelectedRows.");
    }
  }

376
  proto::VarDesc::VarType GetVarType(const std::string& name) const override {
377 378 379 380 381
    auto* var = scope_.FindVar(name);
    return ToVarType(var->Type());
  }

 private:
382 383 384 385 386
  const OperatorBase& op_;
  const Scope& scope_;
};

void OperatorWithKernel::Run(const Scope& scope,
D
dzhwinter 已提交
387
                             const platform::Place& place) const {
388 389
  RuntimeInferShapeContext infer_shape_ctx(*this, scope);
  this->InferShape(&infer_shape_ctx);
D
dzhwinter 已提交
390 391
  platform::DeviceContextPool& pool = platform::DeviceContextPool::Get();
  auto dev_ctx = pool.Borrow(place);
392 393 394 395 396

  // check if op[type] has kernel registered.
  auto& all_op_kernels = AllOpKernels();
  auto kernels_iter = all_op_kernels.find(type_);
  if (kernels_iter == all_op_kernels.end()) {
Y
Yu Yang 已提交
397 398
    PADDLE_THROW(
        "There are no kernels which are registered in the %s operator.", type_);
399 400 401 402
  }

  // check if op[type] have kernel for kernel_key
  OpKernelMap& kernels = kernels_iter->second;
D
dzhwinter 已提交
403 404

  ExecutionContext ctx(*this, scope, *dev_ctx);
Y
Yu Yang 已提交
405
  auto kernel_key = GetKernelType(ctx);
406 407 408
  auto kernel_iter = kernels.find(kernel_key);

  if (kernel_iter == kernels.end()) {
Y
Yu Yang 已提交
409
    PADDLE_THROW("The operator %s does not support %s", type_, kernel_key);
410 411 412 413
  }

  kernel_iter->second->Compute(ctx);
}
Y
Yu Yang 已提交
414 415
OpKernelType OperatorWithKernel::GetKernelType(
    const ExecutionContext& ctx) const {
Q
QI JUN 已提交
416
  return OpKernelType(IndicateDataType(ctx), ctx.GetPlace());
Y
Yu Yang 已提交
417
}
418
proto::DataType OperatorWithKernel::IndicateDataType(
Y
Yu Yang 已提交
419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443
    const ExecutionContext& ctx) const {
  auto& scope = ctx.scope();
  int data_type = -1;
  for (auto& input : this->inputs_) {
    for (auto& ipt_name : input.second) {
      auto* var = scope.FindVar(ipt_name);
      if (var != nullptr) {
        const Tensor* t = nullptr;
        if (var->IsType<Tensor>()) {
          t = &var->Get<Tensor>();
        } else if (var->IsType<LoDTensor>()) {
          t = &var->Get<LoDTensor>();
        } else if (var->IsType<SelectedRows>()) {
          t = &(var->Get<SelectedRows>().value());
        }
        if (t != nullptr) {
          int tmp = static_cast<int>(ToDataType(t->type()));
          PADDLE_ENFORCE(tmp == data_type || data_type == -1,
                         "DataType of Paddle Op %s must be the same.", Type());
          data_type = tmp;
        }
      }
    }
  }
  PADDLE_ENFORCE(data_type != -1, "DataType should be indicated by input");
444
  return static_cast<proto::DataType>(data_type);
Y
Yu Yang 已提交
445
}
446

Q
Qiao Longfei 已提交
447
}  // namespace framework
L
liaogang 已提交
448
}  // namespace paddle