bfloat16 (Brain Floating Point) data type More...

#include <onnxruntime_cxx_api.h>

Inheritance diagram for Ort::BFloat16_t:

Public Types
using	Base = onnxruntime_float16::BFloat16Impl< BFloat16_t >

Public Member Functions
	BFloat16_t ()=default

	BFloat16_t (float v) noexcept
	__ctor from float. Float is converted into bfloat16 16-bit representation. More...

float	ToFloat () const noexcept
	Converts bfloat16 to float More...

	operator float () const noexcept
	User defined conversion operator. Converts BFloat16_t to float. More...

bool	operator== (const BFloat16_t &rhs) const noexcept

bool	operator!= (const BFloat16_t &rhs) const noexcept

bool	operator< (const BFloat16_t &rhs) const noexcept

Public Member Functions inherited from onnxruntime_float16::BFloat16Impl< BFloat16_t >
	BFloat16Impl ()=default

bool	IsNegative () const noexcept
	Checks if the value is negative More...

bool	IsNaN () const noexcept
	Tests if the value is NaN More...

bool	IsFinite () const noexcept
	Tests if the value is finite More...

bool	IsPositiveInfinity () const noexcept
	Tests if the value represents positive infinity. More...

bool	IsNegativeInfinity () const noexcept
	Tests if the value represents negative infinity More...

bool	IsInfinity () const noexcept
	Tests if the value is either positive or negative infinity. More...

bool	IsNaNOrZero () const noexcept
	Tests if the value is NaN or zero. Useful for comparisons. More...

bool	IsNormal () const noexcept
	Tests if the value is normal (not zero, subnormal, infinite, or NaN). More...

bool	IsSubnormal () const noexcept
	Tests if the value is subnormal (denormal). More...

BFloat16_t	Abs () const noexcept
	Creates an instance that represents absolute value. More...

BFloat16_t	Negate () const noexcept
	Creates a new instance with the sign flipped. More...

Static Public Member Functions
static constexpr BFloat16_t	FromBits (uint16_t v) noexcept
	Explicit conversion to uint16_t representation of bfloat16. More...

Static Public Member Functions inherited from onnxruntime_float16::BFloat16Impl< BFloat16_t >
static bool	AreZero (const BFloat16Impl &lhs, const BFloat16Impl &rhs) noexcept
	IEEE defines that positive and negative zero are equal, this gives us a quick equality check for two values by or'ing the private bits together and stripping the sign. They are both zero, and therefore equivalent, if the resulting value is still zero. More...

Additional Inherited Members
Public Attributes inherited from onnxruntime_float16::BFloat16Impl< BFloat16_t >
uint16_t	val

Static Public Attributes inherited from onnxruntime_float16::BFloat16Impl< BFloat16_t >
static constexpr uint16_t	kSignMask

static constexpr uint16_t	kBiasedExponentMask

static constexpr uint16_t	kPositiveInfinityBits

static constexpr uint16_t	kNegativeInfinityBits

static constexpr uint16_t	kPositiveQNaNBits

static constexpr uint16_t	kNegativeQNaNBits

static constexpr uint16_t	kSignaling_NaNBits

static constexpr uint16_t	kEpsilonBits

static constexpr uint16_t	kMinValueBits

static constexpr uint16_t	kMaxValueBits

static constexpr uint16_t	kRoundToNearest

static constexpr uint16_t	kOneBits

static constexpr uint16_t	kMinusOneBits

Protected Member Functions inherited from onnxruntime_float16::BFloat16Impl< BFloat16_t >
float	ToFloatImpl () const noexcept
	Converts bfloat16 to float More...

uint16_t	AbsImpl () const noexcept
	Creates an instance that represents absolute value. More...

uint16_t	NegateImpl () const noexcept
	Creates a new instance with the sign flipped. More...

Static Protected Member Functions inherited from onnxruntime_float16::BFloat16Impl< BFloat16_t >
static uint16_t	ToUint16Impl (float v) noexcept
	Converts from float to uint16_t float16 representation More...

Detailed Description

bfloat16 (Brain Floating Point) data type

This struct is used for converting float to bfloat16 and back so the user could feed inputs and fetch outputs using these type.

The size of the structure should align with uint16_t and one can freely cast uint16_t buffers to/from Ort::BFloat16_t to feed and retrieve data.

// This example demonstrates converion from float to float16
constexpr float values[] = {1.f, 2.f, 3.f, 4.f, 5.f};
std::vector<Ort::BFloat16_t> bfp16_values;
bfp16_values.reserve(std::size(values));
std::transform(std::begin(values), std::end(values), std::back_inserter(bfp16_values),
    [](float value) { return Ort::BFloat16_t(value); });

Definition at line 306 of file onnxruntime_cxx_api.h.

Member Typedef Documentation

using Ort::BFloat16_t::Base = onnxruntime_float16::BFloat16Impl<BFloat16_t>

Definition at line 318 of file onnxruntime_cxx_api.h.

Constructor & Destructor Documentation

Ort::BFloat16_t::BFloat16_t ( )

default

Ort::BFloat16_t::BFloat16_t ( float v )

inlineexplicitnoexcept

__ctor from float. Float is converted into bfloat16 16-bit representation.

Parameters

v	float value

Definition at line 333 of file onnxruntime_cxx_api.h.

Member Function Documentation

static constexpr BFloat16_t Ort::BFloat16_t::FromBits ( uint16_t v )

inlinestaticnoexcept

Explicit conversion to uint16_t representation of bfloat16.

Parameters

v	uint16_t bit representation of bfloat16

Returns: new instance of BFloat16_t

Definition at line 327 of file onnxruntime_cxx_api.h.

Ort::BFloat16_t::operator float ( ) const

inlineexplicitnoexcept

User defined conversion operator. Converts BFloat16_t to float.

Definition at line 420 of file onnxruntime_cxx_api.h.

bool Ort::BFloat16_t::operator!= ( const BFloat16_t & rhs ) const

inlinenoexcept

Definition at line 425 of file onnxruntime_cxx_api.h.

bool Ort::BFloat16_t::operator< ( const BFloat16_t & rhs ) const

inlinenoexcept

Definition at line 153 of file onnxruntime_cxx_inline.h.

bool Ort::BFloat16_t::operator== ( const BFloat16_t & rhs ) const

inlinenoexcept

Definition at line 145 of file onnxruntime_cxx_inline.h.

float Ort::BFloat16_t::ToFloat ( ) const

inlinenoexcept

Converts bfloat16 to float

Returns: float representation of bfloat16 value

Definition at line 339 of file onnxruntime_cxx_api.h.

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

onnxruntime/core/session/onnxruntime_cxx_api.h
onnxruntime/core/session/onnxruntime_cxx_inline.h

Public Types

Public Member Functions

Static Public Member Functions

Additional Inherited Members

Detailed Description

Member Typedef Documentation

Constructor & Destructor Documentation

Member Function Documentation