o2::zdc::fastsim::ConditionalModelSimulation Class Reference

Derived class implementing interface for specific types of models. More...

#include <FastSimulations.h>

Inherits o2::zdc::fastsim::NeuralFastSimulation.

Public Member Functions
	ConditionalModelSimulation (const std::string &modelPath, int64_t batchSize)
	~ConditionalModelSimulation () override=default
bool	setInput (std::vector< std::vector< float > > &input) override
	Implements setInput.
void	run () override
	Implements run().
const std::vector< Ort::Value > &	getResult () override
	Returns single model output as const&. Returned vector is of size 1.
Public Member Functions inherited from o2::zdc::fastsim::NeuralFastSimulation
	NeuralFastSimulation (const std::string &modelPath, OrtAllocatorType allocatorType, OrtMemType memoryType, int64_t batchSize)
virtual	~NeuralFastSimulation ()=default
void	initRunSession ()
	(late) init session and provide mechanism to customize ONNX session with external options
size_t	getBatchSize () const

Additional Inherited Members
Protected Member Functions inherited from o2::zdc::fastsim::NeuralFastSimulation
void	setInputOutputData ()
	Sets models metadata (input/output layers names, inputs shape) in onnx session.
void	setTensors (std::vector< std::vector< float > > &input)
	Converts flattend input data to Ort::Value. Tensor shapes are taken from loaded model metadata.
Protected Attributes inherited from o2::zdc::fastsim::NeuralFastSimulation
std::string	mModelPath
	model path (where to find the ONNX model)
Ort::Env	mEnv
Ort::Session *	mSession = nullptr
Ort::AllocatorWithDefaultOptions	mAllocator
Ort::MemoryInfo	mMemoryInfo
std::vector< std::string >	mInputNames
	Input/Output names and input shape.
std::vector< std::string >	mOutputNames
std::vector< std::vector< int64_t > >	mInputShapes
int64_t	mBatchSize
std::vector< Ort::Value >	mInputTensors
	Container for input tensors.

Detailed Description

Derived class implementing interface for specific types of models.

Definition at line 100 of file FastSimulations.h.

Constructor & Destructor Documentation

ConditionalModelSimulation()

ConditionalModelSimulation::ConditionalModelSimulation	(	const std::string &	modelPath,
		int64_t	batchSize
	)

Definition at line 81 of file FastSimulations.cxx.

~ConditionalModelSimulation()

o2::zdc::fastsim::ConditionalModelSimulation::~ConditionalModelSimulation ( )

overridedefault

Member Function Documentation

getResult()

const std::vector< Ort::Value > & ConditionalModelSimulation::getResult ( )

overridevirtual

Returns single model output as const&. Returned vector is of size 1.

Returns

std::vector<Ort::Value> model output

Implements o2::zdc::fastsim::NeuralFastSimulation.

Definition at line 116 of file FastSimulations.cxx.

run()

void ConditionalModelSimulation::run ( )

overridevirtual

Implements run().

Implements o2::zdc::fastsim::NeuralFastSimulation.

Definition at line 96 of file FastSimulations.cxx.

setInput()

bool ConditionalModelSimulation::setInput ( std::vector< std::vector< float > > &  input )

overridevirtual

Implements setInput.

Parameters

input

flattend input

Returns

true on success

false on failure

Implements o2::zdc::fastsim::NeuralFastSimulation.

Definition at line 86 of file FastSimulations.cxx.

---

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

• /home/runner/work/AliceO2/AliceO2/Detectors/ZDC/fastsimulation/include/FastSimulations.h
• /home/runner/work/AliceO2/AliceO2/Detectors/ZDC/fastsimulation/src/FastSimulations.cxx