torch. unbind()

effect ： The length of a dimension is 1 The section of , And return all slicing results .

Take an example ：

import torch

x = torch.tensor([[1, 2],[ 3,4]])
torch.unbind(x,0)# The first 0 On dimensions, the length is 1 The section of .

result ：

(tensor([1, 2]), tensor([3, 4]))

Application reduces the amount of code ：

m,n=torch.unbind(x,0)
print(m)
print(n)
m,n=x[0],x[1]# Although this is OK , But if x The first dimension of is very big , This is very complicated .
print(m)
print(n)

result ：

tensor([1, 2])
tensor([3, 4])
tensor([1, 2])
tensor([3, 4])

Add another torch.unbind() The equivalent form of ：

x.unbind(0)

result ：

(tensor([1, 2]), tensor([3, 4]))