Struct MixedPrecisionConfig

pub struct MixedPrecisionConfig {
    pub dynamic_loss_scaling: bool,
    pub initial_loss_scale: f64,
    pub scale_factor: f64,
    pub scale_window: usize,
    pub min_loss_scale: f64,
    pub max_loss_scale: f64,
    pub verbose: bool,
}

Mixed precision training configuration

Fields

dynamic_loss_scaling: bool

Whether to use dynamic loss scaling

initial_loss_scale: f64

Initial loss scale (usually a power of 2, like 2^16)

scale_factor: f64

Scale factor to increase loss scale after successful steps

scale_window: usize

Number of consecutive successful steps before increasing loss scale

min_loss_scale: f64

Minimum loss scale

max_loss_scale: f64

Maximum loss scale

verbose: bool

Whether to enable verbosity during training

Trait Implementations

impl Clone for MixedPrecisionConfig

fn clone(&self) -> MixedPrecisionConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for MixedPrecisionConfig

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

Formats the value using the given formatter.

impl Default for MixedPrecisionConfig

fn default() -> Self

Returns the “default value” for a type.