from torch import randint
from torchmetrics.classification import Dice
metric = Dice()
values = [ ]
for _ in range(10):
    values.append(metric(randint(2,(10,)), randint(2,(10,))))
fig_, ax_ = metric.plot(values)