Bfloat16Layout

public final class Bfloat16Layout

Data layout that converts 32-bit floats from/to 16-bit, truncating their mantissa to 7 bits but preserving the 8-bit exponent with the same bias.

Public Constructors

Bfloat16Layout()

Public Methods

float	readFloat(ShortDataBuffer buffer, long index)
void	writeFloat(ShortDataBuffer buffer, float value, long index)

Inherited Methods

From class java.lang.Object

boolean	equals(Object arg0)
final Class<?>	getClass()
int	hashCode()
final void	notify()
final void	notifyAll()
String	toString()
final void	wait(long arg0, int arg1)
final void	wait(long arg0)
final void	wait()

From interface org.tensorflow.ndarray.buffer.layout.FloatDataLayout

abstract FloatDataBuffer	applyTo(ShortDataBuffer buffer) Apply this layout to the provided buffer.
abstract float	readFloat(ShortDataBuffer buffer, long index) Reads `n = scale()` values from the buffer at the given index and return them as a float.
abstract Float	readObject(ShortDataBuffer buffer, long index) Reads `n = scale()` values from the buffer at the given index and return them as a single value in the user type.
abstract void	writeFloat(ShortDataBuffer buffer, float value, long index) Writes a float into the buffer at the given index after converting it to the buffer type.
abstract void	writeObject(ShortDataBuffer buffer, Float value, long index)

From interface org.tensorflow.ndarray.buffer.layout.DataLayout

abstract DataBuffer<Float>	applyTo(ShortDataBuffer buffer) Apply this layout to the provided buffer.
abstract Float	readObject(ShortDataBuffer buffer, long index) Reads `n = scale()` values from the buffer at the given index and return them as a single value in the user type.
abstract int	scale() Indicates the number of buffer values are required to represent a single user value, default is 1.
abstract void	writeObject(ShortDataBuffer buffer, Float value, long index) Writes a user value into the buffer at the given index after converting it to the buffer type.

Public Constructors

public Bfloat16Layout ()

Public Methods

public float readFloat (ShortDataBuffer buffer, long index)

public void writeFloat (ShortDataBuffer buffer, float value, long index)

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2021-11-29 UTC.