pdft.optimizers.RiemannianAdam

class pdft.optimizers.RiemannianAdam(lr=0.001, beta1=0.9, beta2=0.999, eps=1e-08, max_grad_norm=None)

Bases: object

Riemannian Adam optimizer (Becigneul & Ganea, 2019).

Mirror of upstream src/optimizers.jl:173-183. Defaults match upstream.

Parameters:

• lr (float)

• beta1 (float)

• beta2 (float)

• eps (float)

• max_grad_norm (float | None)

__init__(lr=0.001, beta1=0.9, beta2=0.999, eps=1e-08, max_grad_norm=None)

Parameters:

• lr (float)

• beta1 (float)

• beta2 (float)

• eps (float)

• max_grad_norm (float | None)

Return type:

None

Methods

__init__([lr, beta1, beta2, eps, max_grad_norm])

Attributes

beta1
beta2
eps
lr
max_grad_norm

lr: float = 0.001

beta1: float = 0.9

beta2: float = 0.999

eps: float = 1e-08

max_grad_norm: float | None = None