Struct AdaMaxOptimizer 

pub struct AdaMaxOptimizer { /* private fields */ }

AdaMax optimizer (variant of Adam with infinity norm).

Uses the infinity norm of gradients instead of L2 norm, making it more robust to large gradients and outliers.

Reference: Kingma & Ba, “Adam: A Method for Stochastic Optimization”, ICLR 2015

Implementations

impl AdaMaxOptimizer

pub fn new(config: OptimizerConfig) -> Self

Create a new AdaMax optimizer.

Trait Implementations

impl Debug for AdaMaxOptimizer

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

Formats the value using the given formatter.

impl Optimizer for AdaMaxOptimizer

fn step( &mut self, parameters: &mut HashMap<String, Array<f64, Ix2>>, gradients: &HashMap<String, Array<f64, Ix2>>, ) -> TrainResult<()>

Update parameters with computed gradients.

fn zero_grad(&mut self)

Zero all gradients.

fn get_lr(&self) -> f64

Get current learning rate.

fn set_lr(&mut self, lr: f64)

Set learning rate.

fn state_dict(&self) -> HashMap<String, Vec<f64>>

Get optimizer state for checkpointing.

fn load_state_dict(&mut self, state: HashMap<String, Vec<f64>>)

Load optimizer state from checkpoint.