|
|
|
|
@ -0,0 +1,140 @@
|
|
|
|
|
## 在Paddle中如何使用Eigen
|
|
|
|
|
|
|
|
|
|
神经网络本质上是一个计算图，计算需要的数据存放在`Tensor`中，而计算过程是由`Operartor`来描述的。在执行时，`Operator`调用对应`OpKernel`中的`Compute`接口，实现对`Tensor`的操作。
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
### Eigen Tensor模块
|
|
|
|
|
|
|
|
|
|
Eigen Tensor模块对element-wise计算提供了强大的支持，并且书写一份代码，可以同时在CPU、GPU执行。但Eigen Tensor是一个正在开发中的模块，因此可能测试不够完备，文档较少。
|
|
|
|
|
|
|
|
|
|
关于Eigen Tensor模块的详细介绍请参考[文档](https://github.com/RLovelett/eigen/blob/master/unsupported/Eigen/CXX11/src/Tensor/README.md)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
### paddle::framework::Tensor
|
|
|
|
|
|
|
|
|
|
Paddle Tensor定义在framework目录下，其主要接口如下：
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
class Tensor {
|
|
|
|
|
public:
|
|
|
|
|
/*! Return a pointer to mutable memory block. */
|
|
|
|
|
template <typename T>
|
|
|
|
|
inline T* data();
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* @brief Return a pointer to mutable memory block.
|
|
|
|
|
* @note If not exist, then allocation.
|
|
|
|
|
*/
|
|
|
|
|
template <typename T>
|
|
|
|
|
inline T* mutable_data(platform::Place place);
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* @brief Return a pointer to mutable memory block.
|
|
|
|
|
*
|
|
|
|
|
* @param[in] dims The dimensions of the memory block.
|
|
|
|
|
* @param[in] place The place of the memory block.
|
|
|
|
|
*
|
|
|
|
|
* @note If not exist, then allocation.
|
|
|
|
|
*/
|
|
|
|
|
template <typename T>
|
|
|
|
|
inline T* mutable_data(DDim dims, platform::Place place);
|
|
|
|
|
|
|
|
|
|
/*! Resize the dimensions of the memory block. */
|
|
|
|
|
inline Tensor& Resize(const DDim& dims);
|
|
|
|
|
|
|
|
|
|
/*! Return the dimensions of the memory block. */
|
|
|
|
|
inline const DDim& dims() const;
|
|
|
|
|
|
|
|
|
|
private:
|
|
|
|
|
/*! holds the memory block if allocated. */
|
|
|
|
|
std::shared_ptr<Placeholder> holder_;
|
|
|
|
|
|
|
|
|
|
/*! points to dimensions of memory block. */
|
|
|
|
|
DDim dim_;
|
|
|
|
|
};
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
`Placeholder`的作用的延迟分配内存，即我们可以先定义一个Tensor，然后使用Resize接口设置Tensor的大小，最后再调用mutable_data接口分配实际的内存。
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
paddle::framework::Tensor t;
|
|
|
|
|
paddle::platform::CPUPlace place;
|
|
|
|
|
// set size first
|
|
|
|
|
t.Resize({2, 3});
|
|
|
|
|
// allocate memory on CPU later
|
|
|
|
|
t.mutable_data(place);
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
下面以AddOp为例说明Tensor的使用过程：
|
|
|
|
|
|
|
|
|
|
- InferShape
|
|
|
|
|
|
|
|
|
|
在运行神经网络计算图时，我们先调用每个`Operator`的`InferShape`接口，根据输入Tensor的大小来设置输出Tensor的大小，`Resize`接口会被调用。
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
void InferShape(const framework::InferShapeContext &ctx) const override {
|
|
|
|
|
PADDLE_ENFORCE_EQ(ctx.Input<Tensor>("X")->dims(),
|
|
|
|
|
ctx.Input<Tensor>("Y")->dims(),
|
|
|
|
|
"Two input of Add Op's dimension must be same.");
|
|
|
|
|
ctx.Output<Tensor>("Out")->Resize(ctx.Input<Tensor>("X")->dims());
|
|
|
|
|
}
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- Run
|
|
|
|
|
|
|
|
|
|
`Operator`的`Run`接口最终会调用对应`OpKernel`的`Compute`接口，在这时真正的分配内存，`mutable_data`接口会被调用。
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
void Compute(const framework::ExecutionContext& context) const override {
|
|
|
|
|
auto* input0 = context.Input<Tensor>("X");
|
|
|
|
|
auto* input1 = context.Input<Tensor>("Y");
|
|
|
|
|
auto* output = context.Output<Tensor>("Out");
|
|
|
|
|
|
|
|
|
|
output->mutable_data<T>(context.GetPlace());
|
|
|
|
|
|
|
|
|
|
auto X = EigenVector<T>::Flatten(*input0);
|
|
|
|
|
auto Y = EigenVector<T>::Flatten(*input1);
|
|
|
|
|
auto Z = EigenVector<T>::Flatten(*output);
|
|
|
|
|
|
|
|
|
|
auto place = context.GetEigenDevice<Place>();
|
|
|
|
|
|
|
|
|
|
Z.device(place) = X + Y;
|
|
|
|
|
}
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
### paddle::framework::Tensor到EigenTensor的转换
|
|
|
|
|
|
|
|
|
|
如上一小节所示，在具体的计算中，我们需要先把输入Tensor和输出Tensor转换为Eigen支持的格式。我们在[eigen.h](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/eigen.h)中提供了一些全局函数用来实现paddle::framework::Tensor到EigenTensor/EigenMatrix/EigenVector/EigenScalar的转换。
|
|
|
|
|
|
|
|
|
|
以EigenTensor为例，做一个介绍
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
Tensor t;
|
|
|
|
|
float* p = t.mutable_data<float>(make_ddim({1, 2, 3}), platform::CPUPlace());
|
|
|
|
|
for (int i = 0; i < 1 * 2 * 3; i++) {
|
|
|
|
|
p[i] = static_cast<float>(i);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
EigenTensor<float, 3>::Type et = EigenTensor<float, 3>::From(t);
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
From是EigenTensor模板struct提供的一个接口，可以实现从paddle::framework::Tensor到对EigenTensor的转换。由于Tensor的rank是模板参数，因此在转换时需要显示的指定。
|
|
|
|
|
|
|
|
|
|
需要额外注意的是，EigenVector<T>::From方法是把paddle中的一维Tensor转为Eigen的一维Tensor，在这里用EigenVector来表示；而EigenVector<T>::Flatten方法是把paddle中的一个Tensor进行reshape操作，压扁成为Eigen的一维Tensor，类型仍然为EigenVector。
|
|
|
|
|
|
|
|
|
|
更多的转换方法请参考eigen_test.cc中的[单元测试](https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/eigen_test.cc)。
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
### 实现计算
|
|
|
|
|
|
|
|
|
|
当需要完成计算时，我们需要等式左边的EigenTensor调用device接口：
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
auto place = context.GetEigenDevice<Place>();
|
|
|
|
|
Z.device(place) = X + Y;
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
由于Eigen Tensor模块的文档较少，我们可以参考TensorFlow的[kernels](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/core/kernels)模块下的相关`OpKernel`的计算代码。
|
qijun
8 years ago