Layernorm

Enumerations | Functions

Layernorm

Enumerations
enum	miopenNormMode_t {   MIOPEN_ELEMENTWISE_AFFINE = 0 ,   MIOPEN_WEIGHT_BIAS }

Functions
miopenStatus_t	miopenLayerNormForward (miopenHandle_t handle, miopenNormMode_t mode, const miopenTensorDescriptor_t xDesc, const void *x, const miopenTensorDescriptor_t weightDesc, const void *weight, const miopenTensorDescriptor_t biasDesc, const void *bias, const float epsilon, const int32_t normalized_dim, const miopenTensorDescriptor_t yDesc, void *y, const miopenTensorDescriptor_t meanDesc, void *mean, const miopenTensorDescriptor_t rstdDesc, void *rstd)
	Execute a layernorm forward layer. More...

Detailed Description

Enumeration Type Documentation

◆ miopenNormMode_t

enum miopenNormMode_t

LayerNorm mode

Enumerator
MIOPEN_ELEMENTWISE_AFFINE	initialized to ones for weights and zeros for biases
MIOPEN_WEIGHT_BIAS	learnable weights and biases of the module of shape normalized_shape

Examples

/home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-miopen/checkouts/latest/include/miopen/miopen.h.

Function Documentation

◆ miopenLayerNormForward()

miopenStatus_t miopenLayerNormForward	(	miopenHandle_t	handle,
		miopenNormMode_t	mode,
		const miopenTensorDescriptor_t	xDesc,
		const void *	x,
		const miopenTensorDescriptor_t	weightDesc,
		const void *	weight,
		const miopenTensorDescriptor_t	biasDesc,
		const void *	bias,
		const float	epsilon,
		const int32_t	normalized_dim,
		const miopenTensorDescriptor_t	yDesc,
		void *	y,
		const miopenTensorDescriptor_t	meanDesc,
		void *	mean,
		const miopenTensorDescriptor_t	rstdDesc,
		void *	rstd
	)

Execute a layernorm forward layer.

Parameters

handle	MIOpen handle (input)
mode	LayerNorm mode (input)
xDesc	Tensor descriptor for data input tensor x (input)
x	Data tensor x (input)
weightDesc	Tensor descriptor for data input tensor weight (input)
weight	Data tensor weight (input)
biasDesc	Tensor descriptor for data input tensor bias (input)
bias	Data tensor bias (input)
epsilon	Value to stablize inverse variance calculation (input)
normalized_dim	Nomalized dimensions in the input array (input)
yDesc	Tensor descriptor for output data tensor y (input)
y	Data tensor y (output)
meanDesc	Tensor descriptor for output data tensor mean (input)
mean	Data tensor mean (output)
rstdDesc	Tensor descriptor for output data tensor rstd (input)
rstd	Data tensor rstd (output)

Returns

miopenStatus_t

Examples

/home/docs/checkouts/readthedocs.org/user_builds/advanced-micro-devices-miopen/checkouts/latest/include/miopen/miopen.h.