Struct BatchNorm2d 

pub struct BatchNorm2d<D: Dtype, IT, OT, const RANK: usize> {
    pub num_features: usize,
    pub eps: f64,
    pub momentum: f64,
    pub affine: bool,
    pub track_running_stats: bool,
    /* private fields */
}

Batch normalisation over a 4-D [N, C, H, W] input.

Applies the same statistics as BatchNorm1d but treats each (C) channel independently across the (N, H, W) axes.

Fields

num_features: usize

Number of channels/features.

eps: f64

Numerical stability constant added to the variance.

momentum: f64

Weight of the current batch in the running-stats exponential moving average.

affine: bool

Whether to learn per-channel scale (gamma) and shift (beta) parameters.

track_running_stats: bool

Whether to maintain running mean/variance across batches.

Implementations

impl<D: Dtype, IT, OT, const RANK: usize> BatchNorm2d<D, IT, OT, RANK>

pub fn new(num_features: usize) -> Self

Creates a new layer with default eps/momentum/affine/track_running_stats.

pub fn with_eps(self, eps: f64) -> Self

Sets the numerical stability constant.

pub fn with_momentum(self, momentum: f64) -> Self

Sets the running-stats EMA momentum.

pub fn with_affine(self, affine: bool) -> Self

Sets whether to learn per-channel scale/shift parameters.

pub fn with_track_running_stats(self, track: bool) -> Self

Sets whether to maintain running mean/variance across batches.

Trait Implementations

impl<D: Dtype, IT: Tensor<D, RANK> + EagerTensor, OT: Tensor<D, RANK>, const RANK: usize> Layer<IT> for BatchNorm2d<D, IT, OT, RANK>

type Output = OT

This layer’s output type.

fn call(&self, _input: IT) -> Self::Output

Applies this layer to input.

impl<D: Dtype, const RANK: usize> Layer<SymTensor> for BatchNorm2d<D, SymTensor, SymTensor, RANK>

type Output = SymTensor

This layer’s output type.

fn call(&self, input: SymTensor) -> SymTensor

Applies this layer to input.