from torch import rand, randint
from torchmetrics.classification import BinaryCalibrationError
metric = BinaryCalibrationError(n_bins=2, norm='l1')
metric.update(rand(10), randint(2,(10,)))
fig_, ax_ = metric.plot()