layer_normalization

Layer Normalization Layer API

This module provides classes to build layer normalization layers.

Classes

Name	Description
LayerNormalization1D	Layer normalization for 1D data
LayerNormalization2D	Layer normalization for 2D data

Machine-readable model

• LayerNormalization1DclassApply Layer Normalization to the input.
• LayerNormalization2DclassApply Layer Normalization to the input.

class

LayerNormalization1D

Python

Apply Layer Normalization to the input.

helia_edge/layers/preprocessing/layer_normalization.py:18

LayerNormalization1D(epsilon: float = 1e-06, name=None, **kwargs={})

Apply Layer Normalization to the input.

Base class: BaseAugmentation1D.

Inherited from BaseAugmentation: augment_masks(), augment_sample(), augment_targets(), batch_augment(), call(), get_random_transformations().

Parameters

Parameters of LayerNormalization1D

Name	Type	Default	Description
epsilon	float	1e-06	Small value to avoid division by zero.

attribute

training_only

Python

helia_edge/layers/preprocessing/layer_normalization.py:20

training_only = False

attribute

epsilon

Python

helia_edge/layers/preprocessing/layer_normalization.py:34

epsilon = epsilon

method

augment_samples

Python

Augment a batch of samples during training.

helia_edge/layers/preprocessing/layer_normalization.py:36

augment_samples(inputs) -> keras.KerasTensor

Augment a batch of samples during training.

method

get_config

Python

helia_edge/layers/preprocessing/layer_normalization.py:47

get_config()

class

LayerNormalization2D

Python

Apply Layer Normalization to the input.

helia_edge/layers/preprocessing/layer_normalization.py:51

LayerNormalization2D(epsilon: float = 1e-06, name=None, **kwargs={})

Apply Layer Normalization to the input.

Base class: BaseAugmentation2D.

Inherited from BaseAugmentation: augment_masks(), augment_sample(), augment_targets(), batch_augment(), call(), get_random_transformations().

Parameters

Parameters of LayerNormalization2D

Name	Type	Default	Description
epsilon	float	1e-06	Small value to avoid division by zero.

attribute

training_only

Python

helia_edge/layers/preprocessing/layer_normalization.py:53

training_only = False

attribute

epsilon

Python

helia_edge/layers/preprocessing/layer_normalization.py:67

epsilon = epsilon

method

augment_samples

Python

Augment a batch of samples during training.

helia_edge/layers/preprocessing/layer_normalization.py:69

augment_samples(inputs) -> keras.KerasTensor

Augment a batch of samples during training.

method

get_config

Python

helia_edge/layers/preprocessing/layer_normalization.py:80

get_config()