一、item()

input:

x = torch.tensor([1.0])
x.item()

output:

1.0

二、tolist()

input:

a = torch.randn(2, 2)
a.tolist()
a[0,0].tolist()

output:

[[0.012766935862600803, 0.5415473580360413],[-0.08909505605697632, 0.7729271650314331]]0.012766935862600803