pybind.cc 6.2 KB
Newer Older
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. */

L
Luo Tao 已提交
15
#include <Python.h>
Y
Yu Yang 已提交
16
#include <fstream>
Y
Yu Yang 已提交
17
#include <vector>
18

Y
Yu Yang 已提交
19 20 21 22 23 24 25 26 27
#include "paddle/framework/net.h"
#include "paddle/framework/op_registry.h"
#include "paddle/framework/operator.h"
#include "paddle/framework/scope.h"
#include "paddle/pybind/tensor_bind.h"
#include "pybind11/numpy.h"
#include "pybind11/pybind11.h"
#include "pybind11/stl.h"

28 29 30
namespace py = pybind11;
namespace pd = paddle::framework;

Y
Yu Yang 已提交
31
USE_OP(add_two);
Q
Qiao Longfei 已提交
32
USE_OP(onehot_cross_entropy);
Y
Yu Yang 已提交
33
USE_OP_WITHOUT_KERNEL(fc);
Q
Qiao Longfei 已提交
34
USE_OP(sgd);
Q
qijun 已提交
35
USE_OP(mul);
L
liaogang 已提交
36
USE_OP(mean);
Q
qijun 已提交
37 38 39
USE_OP(sigmoid);
USE_OP(softmax);
USE_OP(rowwise_add);
Y
Yan Chunwei 已提交
40
USE_OP_WITHOUT_KERNEL(recurrent_op);
Y
Yu Yang 已提交
41

Y
Yu Yang 已提交
42 43 44 45 46 47 48 49 50 51
template <typename ClassType>
void ExposeOperator(ClassType& m) {
  m.def("infer_shape", &ClassType::type::InferShape)
      .def("run", &ClassType::type::Run)
      .def("outputs",
           [](const typename ClassType::type& op) -> std::vector<std::string> {
             return op.outputs_;
           })
      .def("__str__", &ClassType::type::DebugString);
}
Y
Yu Yang 已提交
52

53 54 55 56 57
static size_t UniqueIntegerGenerator() {
  static std::atomic<size_t> generator;
  return generator.fetch_add(1);
}

58
PYBIND11_PLUGIN(core) {
Y
Yu Yang 已提交
59
  py::module m("core", "C++ core of PaddlePaddle");
60

Y
Yu Yang 已提交
61 62
  py::class_<pd::Tensor>(m, "Tensor", py::buffer_protocol())
      .def_buffer([](pd::Tensor& self) -> py::buffer_info {
63
        return paddle::pybind::CastToPyBuffer(self);
Y
Yu Yang 已提交
64 65 66 67 68
      })
      .def("get_dims",
           [](const pd::Tensor& self) { return pd::vectorize(self.dims()); })
      .def("set_dims",
           [](pd::Tensor& self, const std::vector<int>& dim) {
F
fengjiayi 已提交
69
             self.Resize(pd::make_ddim(dim));
Y
Yu Yang 已提交
70 71 72 73 74 75 76 77 78
           })
      .def("alloc_float",
           [](pd::Tensor& self) {
             self.mutable_data<float>(paddle::platform::CPUPlace());
           })
      .def("alloc_int",
           [](pd::Tensor& self) {
             self.mutable_data<int>(paddle::platform::CPUPlace());
           })
79
      .def("set", paddle::pybind::PyTensorSetFromArray<float>)
Y
Yu Yang 已提交
80 81 82
      .def("set", paddle::pybind::PyTensorSetFromArray<int>)
      .def("shape",
           [](pd::Tensor& self) { return pd::vectorize(self.dims()); });
Y
Yu Yang 已提交
83

84 85 86 87 88 89 90 91 92 93
  py::class_<pd::Variable>(m, "Variable", R"DOC(Variable Class.

All parameter, weight, gradient are variables in Paddle.
)DOC")
      .def("is_int", [](const pd::Variable& var) { return var.IsType<int>(); })
      .def("set_int",
           [](pd::Variable& var, int val) -> void {
             *var.GetMutable<int>() = val;
           })
      .def("get_int",
Y
Yu Yang 已提交
94 95 96 97 98
           [](const pd::Variable& var) -> int { return var.Get<int>(); })
      .def("get_tensor",
           [](pd::Variable& self) -> pd::Tensor* {
             return self.GetMutable<pd::Tensor>();
           },
Y
Yan Chunwei 已提交
99 100 101 102 103
           py::return_value_policy::reference)
      .def("get_net",
           [](pd::Variable& self) -> pd::NetOp* {
             return self.GetMutable<pd::NetOp>();
           },
Y
Yu Yang 已提交
104
           py::return_value_policy::reference);
105 106 107 108 109 110 111 112

  py::class_<pd::Scope, std::shared_ptr<pd::Scope>>(m, "Scope")
      .def(py::init<const std::shared_ptr<pd::Scope>&>())
      .def("get_var",
           &pd::Scope::GetVariable,
           py::return_value_policy::reference)
      .def("create_var",
           &pd::Scope::CreateVariable,
113 114
           py::return_value_policy::reference)
      .def("get_var_name", &pd::Scope::GetVariableName);
115

Y
Yu Yang 已提交
116 117
  //! @note: Be careful! PyBind will return std::string as an unicode, not
  //! Python str. If you want a str object, you should cast them in Python.
Y
Yu Yang 已提交
118
  m.def("get_all_op_protos", []() -> std::vector<py::bytes> {
Y
Yu Yang 已提交
119
    auto& protos = pd::OpRegistry::protos();
Y
Yu Yang 已提交
120
    std::vector<py::bytes> ret_values;
Y
Yu Yang 已提交
121
    for (auto it = protos.begin(); it != protos.end(); ++it) {
Y
Yu Yang 已提交
122 123
      PADDLE_ENFORCE(it->second.IsInitialized(),
                     "OpProto must all be initialized");
Y
Yu Yang 已提交
124 125
      std::string str;
      PADDLE_ENFORCE(it->second.SerializeToString(&str),
Y
Yu Yang 已提交
126
                     "Serialize OpProto Error. This could be a bug of Paddle.");
Y
Yu Yang 已提交
127
      ret_values.push_back(py::bytes(str));
Y
Yu Yang 已提交
128 129 130
    }
    return ret_values;
  });
131 132 133 134 135 136
  m.def_submodule(
       "var_names",
       "The module will return special predefined variable name in Paddle")
      .def("empty", pd::OperatorBase::EMPTY_VAR_NAME)
      .def("temp", pd::OperatorBase::TMP_VAR_NAME);

Y
Yu Yang 已提交
137 138 139 140 141
  py::class_<paddle::platform::DeviceContext>(m, "DeviceContext")
      .def_static("cpu_context", []() -> paddle::platform::DeviceContext* {
        return new paddle::platform::CPUDeviceContext();
      });

Y
Yu Yang 已提交
142 143
  py::class_<pd::OperatorBase, std::shared_ptr<pd::OperatorBase>> operator_base(
      m, "Operator");
Y
Yu Yang 已提交
144

Y
Yu Yang 已提交
145
  operator_base.def_static("create", [](py::bytes protobin) {
Y
Yu Yang 已提交
146 147 148 149 150 151 152 153 154 155
    pd::OpDesc desc;
    PADDLE_ENFORCE(desc.ParsePartialFromString(protobin),
                   "Cannot parse user input to OpDesc");
    PADDLE_ENFORCE(desc.IsInitialized(),
                   "User OpDesc is not initialized, reason %s",
                   desc.InitializationErrorString());
    return pd::OpRegistry::CreateOp(desc);
  });
  ExposeOperator(operator_base);

Y
Yu Yang 已提交
156
  py::class_<pd::NetOp, std::shared_ptr<pd::NetOp>> net(m, "Net");
Y
Yu Yang 已提交
157 158

  net.def_static("create",
Y
Yu Yang 已提交
159 160
                 []() -> std::shared_ptr<pd::NetOp> {
                   auto retv = std::make_shared<pd::NetOp>();
Y
Yu Yang 已提交
161
                   retv->type_ = "plain_net";
Y
Yu Yang 已提交
162 163
                   return retv;
                 })
Y
Yu Yang 已提交
164
      .def("add_op", &pd::NetOp::AddOp)
Y
Yu Yang 已提交
165
      .def("add_op",
Y
Yu Yang 已提交
166 167
           [](pd::NetOp& self, const std::shared_ptr<pd::NetOp>& net) -> void {
             self.AddOp(std::static_pointer_cast<pd::OperatorBase>(net));
Y
Yu Yang 已提交
168
           })
Y
Yu Yang 已提交
169 170 171
      .def("complete_add_op", &pd::NetOp::CompleteAddOp)
      .def("complete_add_op",
           [](std::shared_ptr<pd::NetOp>& self) { self->CompleteAddOp(); });
Y
Yu Yang 已提交
172
  ExposeOperator(net);
Y
Yu Yang 已提交
173

174 175
  m.def("unique_integer", UniqueIntegerGenerator);

176
  return m.ptr();
L
Luo Tao 已提交
177
}