Struct DenseLayer 

pub struct DenseLayer {
    pub weights: Vec<f32>,
    pub biases: Vec<f32>,
    pub in_features: usize,
    pub out_features: usize,
    pub activation: ActivationFn,
}

A single fully-connected (dense) layer with an activation function.

Weights are stored in row-major order: weights[out * in_features + in].

Fields

weights: Vec<f32>

Weight matrix in row-major layout [out_features × in_features].

biases: Vec<f32>

Bias vector of length out_features.

in_features: usize

Number of input features.

out_features: usize

Number of output features.

activation: ActivationFn

Activation function applied after the affine transform.

Implementations

impl DenseLayer

pub fn new( in_features: usize, out_features: usize, activation: ActivationFn, ) -> Self

Create a new layer with zero-initialised weights and biases.

pub fn forward(&self, input: &[f32]) -> Vec<f32>

Compute activation(W * input + b).

Panics

Panics if input.len() != self.in_features.

pub fn set_weights(&mut self, w: &[f32])

Replace the weight matrix (must have length out_features * in_features).

Panics

Panics if w.len() does not match.

pub fn set_biases(&mut self, b: &[f32])

Replace the bias vector (must have length out_features).

Panics

Panics if b.len() does not match.

pub fn parameter_count(&self) -> usize

Total number of trainable parameters (weights + biases).

Trait Implementations

impl Clone for DenseLayer

fn clone(&self) -> DenseLayer

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for DenseLayer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.