Enum candle_nn::init::FanInOut

pub enum FanInOut {
    FanIn,
    FanOut,
}

Number of features as input or output of a layer. In Kaiming initialization, choosing FanIn preserves the magnitude of the variance of the weights in the forward pass, choosing FanOut preserves this magnitude in the backward pass.

Variants

FanIn

FanOut

Implementations

impl FanInOut

pub fn for_shape(&self, shape: &Shape) -> usize

Compute the fan-in or fan-out value for a weight tensor of the specified dimensions. https://github.com/pytorch/pytorch/blob/dbeacf11820e336e803bb719b7aaaf2125ae4d9c/torch/nn/init.py#L284

Trait Implementations

impl Clone for FanInOut

fn clone(&self) -> FanInOut

Returns a copy of the value.

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

Performs copy-assignment from source.

impl Debug for FanInOut

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

Formats the value using the given formatter.

impl Copy for FanInOut