Paddle/paddle/fluid/inference/api/high_level_api.md

Inference High-level APIs

This document describes the high-level inference APIs, one can use them to deploy a Paddle model for an application quickly.

The APIs are described in paddle_inference_api.h, just one header file, and two libaries libpaddle_fluid.so and libpaddle_fluid_api.so are needed for a deployment.

PaddleTensor

We provide the PaddleTensor data structure to give a general tensor interface.

The definition is

struct PaddleTensor {
  std::string name;  // variable name.
  std::vector<int> shape;
  PaddleBuf data;  // blob of data.
  PaddleDType dtype;
};

The data is stored in a continuous memory PaddleBuf, and a PaddleDType specifies tensor's data type. The name field is used to specify the name of an input variable, that is important when there are multiple inputs and need to distinguish which variable to set.

engine

The inference APIs has two different underlying engines

• the native engine, which is consists of the native operators and framework,
• the Anakin engine, which has an Anakin library embedded.

The native engine takes a native Paddle model as input, and supports any model that trained by Paddle, the Anakin engine is faster for some model, but it can only take the Anakin model as input(user need to transform the format first manually) and currently not all Paddle models are supported.

enum class PaddleEngineKind {
  kNative = 0,  // Use the native Fluid facility.
  kAnakin,      // Use Anakin for inference.
};

PaddlePredictor and how to create one

The main interface is PaddlePredictor, there are following methods

• bool Run(const std::vector<PaddleTensor>& inputs, std::vector<PaddleTensor>* output_data)
  • take inputs and output output_data.
• Clone to clone a predictor from an existing one, with model parameter shared.

There is a factory method to help create a predictor, and the user takes the ownership of this object.

template <typename ConfigT, PaddleEngineKind engine = PaddleEngineKind::kNative>
std::unique_ptr<PaddlePredictor> CreatePaddlePredictor(const ConfigT& config);

By specifying the engine kind and config, one can get a specific implementation.

Reference

• paddle_inference_api.h
• some demos