SampleINT8 Class Reference

The SampleINT8 class implements the INT8 sample. More...

Collaboration diagram for SampleINT8:

Public Member Functions
	SampleINT8 (const SampleINT8Params &params)
bool	build (DataType dataType)
	Function builds the network engine. More...
bool	isSupported (DataType dataType)
	Checks if the platform supports the data type. More...
bool	infer (std::vector< float > &score, int firstScoreBatch, int nbScoreBatches)
	Runs the TensorRT inference engine for this sample. More...
bool	teardown ()
	Cleans up any state created in the sample class. More...

Private Types
template<typename T >
using	SampleUniquePtr = std::unique_ptr< T, samplesCommon::InferDeleter >

Private Member Functions
bool	constructNetwork (SampleUniquePtr< nvinfer1::IBuilder > &builder, SampleUniquePtr< nvinfer1::INetworkDefinition > &network, SampleUniquePtr< nvinfer1::IBuilderConfig > &config, SampleUniquePtr< nvcaffeparser1::ICaffeParser > &parser, DataType dataType)
	Parses a Caffe model and creates a TensorRT network. More...
bool	processInput (const samplesCommon::BufferManager &buffers, const float *data)
	Reads the input and stores it in a managed buffer. More...
int	calculateScore (const samplesCommon::BufferManager &buffers, float *labels, int batchSize, int outputSize, int threshold)
	Scores model. More...

Private Attributes
SampleINT8Params	mParams
	The parameters for the sample. More...
nvinfer1::Dims	mInputDims
	The dimensions of the input to the network. More...
std::shared_ptr< nvinfer1::ICudaEngine >	mEngine
	The TensorRT engine used to run the network. More...

Detailed Description

The SampleINT8 class implements the INT8 sample.

It creates the network using a caffe model

Member Typedef Documentation

SampleUniquePtr

template<typename T >

using SampleINT8::SampleUniquePtr = std::unique_ptr<T, samplesCommon::InferDeleter>

private

Constructor & Destructor Documentation

SampleINT8()

SampleINT8::SampleINT8 ( const SampleINT8Params &  params )

inline

Member Function Documentation

build()

bool SampleINT8::build

(

DataType

dataType

)

Function builds the network engine.

Creates the network, configures the builder and creates the network engine.

This function creates the network by parsing the caffe model and builds the engine that will be used to run the model (mEngine)

Returns

Returns true if the engine was created successfully and false otherwise

Here is the call graph for this function:

isSupported()

bool SampleINT8::isSupported

(

DataType

dataType

)

Checks if the platform supports the data type.

Returns

Returns true if the platform supports the data type.

infer()

bool SampleINT8::infer	(	std::vector< float > &	score,
		int	firstScoreBatch,
		int	nbScoreBatches
	)

Runs the TensorRT inference engine for this sample.

This function is the main execution function of the sample. It allocates the buffer, sets inputs and executes the engine.

Here is the call graph for this function:

teardown()

bool SampleINT8::teardown

(

)

Cleans up any state created in the sample class.

Clean up the libprotobuf files as the parsing is complete

Note

It is not safe to use any other part of the protocol buffers library after ShutdownProtobufLibrary() has been called.

Here is the call graph for this function:

constructNetwork()

bool SampleINT8::constructNetwork ( SampleUniquePtr< nvinfer1::IBuilder > &  builder, SampleUniquePtr< nvinfer1::INetworkDefinition > &  network, SampleUniquePtr< nvinfer1::IBuilderConfig > &  config, SampleUniquePtr< nvcaffeparser1::ICaffeParser > &  parser, DataType  dataType  )

private

Parses a Caffe model and creates a TensorRT network.

Uses a caffe parser to create the network and marks the output layers.

Parameters

network	Pointer to the network that will be populated with the network
builder	Pointer to the engine builder

Here is the call graph for this function:

Here is the caller graph for this function:

processInput()

bool SampleINT8::processInput ( const samplesCommon::BufferManager &  buffers, const float *  data  )

private

Reads the input and stores it in a managed buffer.

Here is the call graph for this function:

Here is the caller graph for this function:

calculateScore()

int SampleINT8::calculateScore ( const samplesCommon::BufferManager &  buffers, float *  labels, int  batchSize, int  outputSize, int  threshold  )

private

Scores model.

Here is the call graph for this function:

Here is the caller graph for this function:

Member Data Documentation

mParams

SampleINT8Params SampleINT8::mParams

private

The parameters for the sample.

mInputDims

nvinfer1::Dims SampleINT8::mInputDims

private

The dimensions of the input to the network.

mEngine

std::shared_ptr<nvinfer1::ICudaEngine> SampleINT8::mEngine

private

The TensorRT engine used to run the network.

---

The documentation for this class was generated from the following file:

• sampleINT8.cpp