pytorch_pfn_extras.training.triggers.TimeTrigger#

class pytorch_pfn_extras.training.triggers.TimeTrigger(period)#

Bases: Trigger

Trigger based on a fixed time interval.

This trigger accepts iterations with a given interval time.

Parameters:

period (float) – Interval time. It is given in seconds.

Methods

__init__(period)

load_state_dict(to_load)

may_fire(iteration, epoch_len)

Flags if the trigger may fire at the current iteration

state_dict()

__call__(manager)#

Call self as a function.

Parameters:

manager (ExtensionsManagerProtocol) –

Return type:

bool

__init__(period)#
Parameters:

period (float) –

Return type:

None

load_state_dict(to_load)#
Parameters:

to_load (Dict[str, Any]) –

Return type:

None

state_dict()#
Return type:

Dict[str, Any]