FedGenGeneratorOptimizer

Enums · AiDotNet.Enums

Update rule used to train the FedGen server generator.

For Beginners

After each training step the generator’s weights move against the gradient; this chooses how far and in which way they move.

Fields

Field Summary
FedGenGeneratorOptimizer Adam Adam with L2 weight decay added to the gradient (torch.optim.Adam), as in the official FedGen code.
FedGenGeneratorOptimizer Sgd Plain stochastic gradient descent with L2 weight decay added to the gradient.
int value__