Learn practical skills, build real-world projects, and advance your career
Updated 4 years ago
PyTorch
These are the few tensor functions I've choosen.
For more functions visit:
https://pytorch.org/docs/stable/torch.html
- torch.cat
- torch.unsqueeze
- torch.view
- torch.zeros_like
- torch.arange
# Import torch and other required modules
import torch
Function 1 - torch.cat
torch.cat
is a built-in function which can concatenate sequences of given tensor in the given dimension
# here are two 2 x 2 tensors (x and y)
# manual tensors.
tensor_x = torch.tensor([[5, 6],
[8, 7]])
tensor_y = torch.tensor([[10, 11],
[12, 13]])
# concatenate tensor in the given dim (0 - rows, 1 - column)
# concat on dim = 0 (rows)
vec_row = torch.cat((tensor_x, tensor_y), dim=0)
# concat on dim = 1 (columns)
vec_col = torch.cat((tensor_x, tensor_y), dim=1)
print(vec_row, vec_col, vec_row.shape, vec_col.shape, sep='\n\n',)
tensor([[ 5, 6],
[ 8, 7],
[10, 11],
[12, 13]])
tensor([[ 5, 6, 10, 11],
[ 8, 7, 12, 13]])
torch.Size([4, 2])
torch.Size([2, 4])
- Not only two tensors,
torch.cat
can concatenate n-number of tensor in the given dimension.