optim_ignite_sgd function

LibTorch implementation of SGD

Implements stochastic gradient descent (optionally with momentum). Nesterov momentum is based on the formula from On the importance of initialization and momentum in deep learning.

optim_ignite_sgd(
  params,
  lr = optim_required(),
  momentum = 0,
  dampening = 0,
  weight_decay = 0,
  nesterov = FALSE
)

Arguments

• params: (iterable): iterable of parameters to optimize or dicts defining parameter groups
• lr: (float): learning rate
• momentum: (float, optional): momentum factor (default: 0)
• dampening: (float, optional): dampening for momentum (default: 0)
• weight_decay: (float, optional): weight decay (L2 penalty) (default: 0)
• nesterov: (bool, optional): enables Nesterov momentum (default: FALSE)

Fields and Methods

See OptimizerIgnite.

Examples

if (torch_is_installed()) {
## Not run:

optimizer <- optim_ignite_sgd(model$parameters(), lr = 0.1)
optimizer$zero_grad()
loss_fn(model(input), target)$backward()
optimizer$step()
## End(Not run)

}