Struct RandAugmentConfig 

pub struct RandAugmentConfig {
    pub n_ops: usize,
    pub magnitude: f32,
    pub fill_value: f32,
    pub ops: Vec<AugOp>,
}

Configuration for the RandAugment policy (Cubuk et al., NeurIPS 2020).

Fields

n_ops: usize

N: number of operations to sample and apply per image (default: 2).

magnitude: f32

M: shared magnitude on a 0–30 scale (default: 9.0).

fill_value: f32

Fill value for geometric transforms when sampling outside the image boundary.

ops: Vec<AugOp>

Pool of operations to sample from (default: all 14).

Implementations

impl RandAugmentConfig

pub fn validate(&self) -> SslResult<()>

Validate that the config is self-consistent.

Trait Implementations

impl Clone for RandAugmentConfig

fn clone(&self) -> RandAugmentConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RandAugmentConfig

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

Formats the value using the given formatter.

impl Default for RandAugmentConfig

fn default() -> Self

Returns the “default value” for a type.