aiaccel.torch.lr_schedulers.SequentialLR

class aiaccel.torch.lr_schedulers.SequentialLR(optimizer: Optimizer, schedulers_fn: list[Callable[[Optimizer], _LRScheduler]], milestones: list[int])

A wrapper of torch.optim.lr_scheduler.SequentialLR to use list of functions to create schedulers.

Parameters:

• optimizer – Optimizer.

• schedulers_fn – List of functions to create schedulers.

• milestones – List of epoch indices. Must be increasing.

… code-block:: yaml

scheduler_generator:

_partial_: True _convert_: “all” _target_: aiaccel.lr_schedulers.SequentialLR schedulers_fn:

• _target_: torch.optim.lr_scheduler.LinearLR _partial_: True start_factor: 1.e-3 end_factor: 1.0 total_iters: 5000

• _target_: torch.optim.lr_scheduler.CosineAnnealingLR

_partial_: True T_max: 95000

milestones: [5000]

__init__(optimizer: Optimizer, schedulers_fn: list[Callable[[Optimizer], _LRScheduler]], milestones: list[int])

Methods

__init__(optimizer, schedulers_fn, 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.