pytorch_pfn_extras.training.extensions.accumulate.min_accumulate.SummaryBase#

class pytorch_pfn_extras.training.extensions.accumulate.min_accumulate.SummaryBase#

Bases: ABC

Methods

__init__()

add(value[, weight])

compute_accumulate()

load_state_dict(to_load)

state_dict()

__init__()#
Return type:

None

abstract add(value, weight=1)#
Parameters:

  • value (Union[Tensor, ndarray, floating, float, Callable[[], float]]) –

  • weight (Union[Tensor, ndarray, floating, float]) –

Return type:

None

abstract compute_accumulate()#
Return type:

Union[Tensor, ndarray, floating, float]

abstract load_state_dict(to_load)#
Parameters:

to_load (Dict[str, Any]) –

Return type:

None

abstract state_dict()#
Return type:

Dict[str, Any]