Bases: object
Hyper-parameters for the optional per-element coefficient refinement
(Adam with annealed gradient noise).
Methods
-
learning_rate: float = 0.003
-
beta1: float = 0.9
-
beta2: float = 0.999
-
epsilon: float = 1e-08
-
noise_eta: float = 0.001
-
noise_gamma: float = 0.55
-
seed: int = 0
-
__init__(learning_rate=0.003, beta1=0.9, beta2=0.999, epsilon=1e-08, noise_eta=0.001, noise_gamma=0.55, seed=0)