LayerNorm

Function

Normalize all features of a single sample.

Prototype

torch.nn.LayerNorm(
    normalized_shape: Union[int, List[int], torch.Size],
    eps: float = 1e-5,
    elementwise_affine: bool = True,
    device=None,
    dtype=None
)

Parameters

**Table 1** Parameter description
Parameter	Type	Mandatory (Yes/No)	Description
normalized_shape	int/list/tuple	Yes	Shape to be normalized (for example, [C, H, W] or hidden_size)
eps	float	No	Numerical stability constant (default value: 1e-5)
elementwise_affine	bool	No	Whether to learn γ and β element-wise (default value: True)

Parent topic: Operator Interfaces