WithOutputTensorIndex< T > Struct Template Reference

WithOutputTensorIndex wraps a value and an index. More...

#include <TfParser.hpp>

Public Member Functions
	WithOutputTensorIndex (const T &value, unsigned int index)
	WithOutputTensorIndex (T &&value, unsigned int index)

Public Attributes
T	m_IndexedValue
unsigned int	m_Index

Detailed Description

template<typename T>
struct armnnTfParser::WithOutputTensorIndex< T >

WithOutputTensorIndex wraps a value and an index.

The purpose of this template is to signify that, in Tensorflow, the input name of a layer has the convention of 'inputTensorName:#index', where the #index can be omitted and it implicitly means the 0 output of the referenced layer. By supporting this notation we can handle layers with multiple outputs, such as Split.

Definition at line 46 of file TfParser.hpp.

Constructor & Destructor Documentation

WithOutputTensorIndex() [1/2]

WithOutputTensorIndex ( const T &  value, unsigned int  index  )

inline

Definition at line 51 of file TfParser.hpp.

    : m_IndexedValue{value}
    , m_Index{index} {}

WithOutputTensorIndex() [2/2]

WithOutputTensorIndex ( T &&  value, unsigned int  index  )

inline

Definition at line 55 of file TfParser.hpp.

    : m_IndexedValue{value}
    , m_Index{index} {}

Member Data Documentation

m_Index

unsigned int m_Index

Definition at line 49 of file TfParser.hpp.

Referenced by armnnTfParser::OutputShapeOfExpandDims(), and TfParser::TfParser().

m_IndexedValue

T m_IndexedValue

Definition at line 48 of file TfParser.hpp.

Referenced by armnnTfParser::OutputShapeOfExpandDims(), and TfParser::TfParser().

---

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

• src/armnnTfParser/TfParser.hpp