IEEE 754 half-precision floating point data type. More...

#include <onnxruntime_cxx_api.h>

Inheritance diagram for Ort::Float16_t:

Public Types
using	Base = onnxruntime_float16::Float16Impl< Float16_t >

Public Member Functions
	Float16_t ()=default
	Default constructor.

	Float16_t (float v) noexcept
	__ctor from float. Float is converted into float16 16-bit representation.

float	ToFloat () const noexcept
	Converts float16 to float.

	operator float () const noexcept
	User defined conversion operator. Converts Float16_t to float.

Static Public Member Functions
static constexpr Float16_t	FromBits (uint16_t v) noexcept
	Explicit conversion to uint16_t representation of float16.

Detailed Description

IEEE 754 half-precision floating point data type.

This struct is used for converting float to float16 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::Float16_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::Float16_t> fp16_values;
fp16_values.reserve(std::size(values));
std::transform(std::begin(values), std::end(values), std::back_inserter(fp16_values),
    [](float value) { return Ort::Float16_t(value); });

Member Typedef Documentation

◆ Base

using Ort::Float16_t::Base = onnxruntime_float16::Float16Impl<Float16_t>

Constructor & Destructor Documentation

◆ Float16_t() [1/2]

Ort::Float16_t::Float16_t ( )

default

Default constructor.

◆ Float16_t() [2/2]

Ort::Float16_t::Float16_t ( float v )

inlineexplicitnoexcept

__ctor from float. Float is converted into float16 16-bit representation.

Parameters

v	float value

Member Function Documentation

◆ FromBits()

static constexpr Float16_t Ort::Float16_t::FromBits ( uint16_t v )

inlinestaticconstexprnoexcept

Explicit conversion to uint16_t representation of float16.

Parameters

v	uint16_t bit representation of float16

Returns: new instance of Float16_t

◆ operator float()

Ort::Float16_t::operator float ( ) const

inlineexplicitnoexcept

User defined conversion operator. Converts Float16_t to float.

◆ ToFloat()

float Ort::Float16_t::ToFloat ( ) const

inlinenoexcept

Converts float16 to float.

Returns: float representation of float16 value

Public Types

Public Member Functions

Static Public Member Functions