Struct AlphaDropout 

pub struct AlphaDropout { /* private fields */ }

Alpha Dropout for SELU activations.

Maintains self-normalizing properties when used with SELU activation. Randomly sets elements to negative saturation value instead of zero.

Reference

• Klambauer, G., et al. (2017). Self-Normalizing Neural Networks. NeurIPS.

Implementations

impl AlphaDropout

pub fn new(p: f32) -> Self

Create a new AlphaDropout layer.

pub fn with_seed(p: f32, seed: u64) -> Self

Create AlphaDropout with a specific seed.

Trait Implementations

impl Debug for AlphaDropout

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

Formats the value using the given formatter.

impl Module for AlphaDropout

fn forward(&self, input: &Tensor) -> Tensor

Forward pass with alpha dropout (SELU-preserving).

Panics

Panics if the RNG mutex is poisoned (unrecoverable system state).

fn train(&mut self)

Set the module to training mode.

fn eval(&mut self)

Set the module to evaluation mode.

fn training(&self) -> bool

Check if the module is in training mode.

fn parameters(&self) -> Vec<&Tensor>

Get references to all learnable parameters.

fn parameters_mut(&mut self) -> Vec<&mut Tensor>

Get mutable references to all learnable parameters.

fn refresh_caches(&mut self)

Refresh any cached computations after parameters have been modified.

fn zero_grad(&mut self)

Zero out gradients for all parameters.

fn num_parameters(&self) -> usize

Get the number of learnable parameters.