PyTorch torch.ravel function
Pytorch torch reference manual
torch.ravelIt is a function in PyTorch used to flatten a tensor into one dimension. It returns a view of the input tensor (if possible), not a copy.
Function definition
torch.ravel(input)
Parameter description:
input: input tensor
Usage example
Example
import torch
# Create a 2x3 matrix
x = torch.tensor([[1, 2, 3], [4, 5, 6]])
# Flatten to one dimension
y = torch.ravel(x)
print(y)
# Create a 2x3 matrix
x = torch.tensor([[1, 2, 3], [4, 5, 6]])
# Flatten to one dimension
y = torch.ravel(x)
print(y)
The output result is:
tensor([1, 2, 3, 4, 5, 6])
Example
import torch
# Create a three-dimensional tensor
x = torch.arange(12).reshape(2, 3, 2)
# Flatten to one dimension
y = torch.ravel(x)
print(y)
# Create a three-dimensional tensor
x = torch.arange(12).reshape(2, 3, 2)
# Flatten to one dimension
y = torch.ravel(x)
print(y)
The output result is:
tensor([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11])
Other extensions