xlnstorch.optim.lr_scheduler.LNSSequentialLR

xlnstorch.optim.lr_scheduler.LNSSequentialLR#

class xlnstorch.optim.lr_scheduler.LNSSequentialLR(optimizer, schedulers, milestones, last_epoch=-1)#

A scheduler that applies a sequence of schedulers in order.

Note that this scheduler is a subclass of torch’s SequentialLR, and is implemented for completeness. You can use the torch version directly with LNS optimizers.

See also: torch.optim.lr_scheduler.SequentialLR

Parameters:

  • optimizer (LNSOptimizer) – Wrapped optimizer.

  • schedulers (List[torch.optim.lr_scheduler.LRScheduler]) – List of schedulers to apply sequentially.

  • milestones (List[int]) – List of epochs at which to switch to the next scheduler.

  • last_epoch (int, optional) – The index of last epoch. Default: -1.

__init__(optimizer, schedulers, milestones, last_epoch=-1)#
Parameters:

Methods

__init__(optimizer, schedulers, milestones)

get_last_lr()

Return last computed learning rate by current scheduler.

get_lr()

Compute learning rate using chainable form of the scheduler.

load_state_dict(state_dict)

Load the scheduler's state.

recursive_undo([sched])

Recursively undo any step performed by the initialisation of schedulers.

state_dict()

Return the state of the scheduler as a dict.

step()

Perform a step.
Contents