StepLR class

Multiplies the LR by gamma every stepSize steps.

lr(t) = initialLr * gamma^floor(t / stepSize)

Inheritance

• Object
• LRScheduler
• StepLR

Constructors

StepLR(Optimizer optimizer, {required double initialLr, required int stepSize, double gamma = 0.1})

Properties

gamma → double

final

hashCode → int

The hash code for this object.

no setter, inherited

initialLr → double

final

lastLr → double?

The most recently applied LR (== optimizer.lr unless the caller mutated it externally). null before the first step().

no setter, inherited

optimizer → Optimizer

final, inherited

runtimeType → Type

A representation of the runtime type of the object.

no setter, inherited

stepCount ↔ int

Steps taken so far. 0 before the first step() call. After the first step(), stepCount == 1 and optimizer.lr reflects the LR for that first step.

getter/setter pair, inherited

stepSize → int

final

Methods

lrAt(int step) → double

The LR that will be applied on the next step() call.

override

noSuchMethod(Invocation invocation) → dynamic

Invoked when a nonexistent method or property is accessed.

inherited

step() → void

Advance one step and write the new LR into the wrapped optimizer. Call this AFTER optimizer.step().

inherited

toString() → String

A string representation of this object.

inherited

Operators

operator ==(Object other) → bool

The equality operator.

inherited