from torch import rand, randint
from torchmetrics.classification import BinarySpecificity
metric = BinarySpecificity()
metric.update(rand(10), randint(2,(10,)))
fig_, ax_ = metric.plot()