Module schedulers 

Structs

ConstantLR

Constant learning rate (no scheduling)

CosineAnnealingLR

Cosine annealing scheduler with warm restarts

CosineAnnealingWarmRestarts

Cosine annealing with warm restarts

CyclicalLR

Cyclical learning rate policy with different modes

ExponentialLR

Exponential decay scheduler: multiply LR by gamma every epoch

LRScheduleVisualizer

Learning rate schedule visualization helper

LinearLR

Linear learning rate schedule

MultiStepLR

Multi-step decay: multiply LR by gamma at specific milestones

OneCycleLR

One cycle learning rate policy (popular for modern deep learning)

PolynomialLR

Polynomial learning rate decay

ReduceLROnPlateau

Reduce learning rate on plateau (when validation loss stops improving)

StepLR

Step decay scheduler: multiply LR by gamma every step_size epochs

WarmupScheduler

Warmup scheduler that gradually increases learning rate

Enums

AnnealStrategy

CyclicalMode

ScaleMode

Traits

LearningRateScheduler

Learning rate scheduler trait for adaptive learning rate adjustment during training