Requirements
To follow this setup, it is required to have following …
•
No Requirements
Basic Types [torch]
# Scalar
scalar = torch.tensor(1)
scalar
# Vector
vector = torch.tensor([1,2,3])
vector
# Matrix
matrix = torch.tensor([[1,2,3],[4,5,6],[7,8,9]])
matrix
# Tensor
tensor = torch.tensor([[[1,2,3],[4,5,6],[7,8,9]],[[11,12,13],[14,15,16],[17,18,19]],[[21,22,23],[24,25,26],[27,28,29]]])
tensor
Python
복사
Synopsis [torch.functions]
a = torch.tensor(4., requires_grad = True)
# Get torch shape/size
a.shape
>>> torch.Size([3, 3, 3])
>>> torch.Size([z, y, x])
# if Numpy
a.dtype
>>> dtype(’float64’)
# if PyTorch
a.dtype
>>> torch.foat64
# Compute derivatives
# Resquires to have requires_grad = True
a.backward()
Python
복사
