public override apply(parameter as NDArray<T>, gradient as NDArray<T>, gradientScale as Double = 1.0) as Void
Updates moments and one parameter in place without altering the derivative. Uses bias-corrected moments: p -= rate * mHat / (sqrt(vHat) + epsilon).
| Parameter | Description |
|---|---|
parameter | Mutable parameter tensor of any rank. |
gradientScale | Positive finite loss scale removed before the update. |
gradient | Same-shaped mean-reduced derivative. |
optimizer.apply(weights, dw)