PyTorch torch.logsumexp Function
PyTorch torch Reference Manual
torch.logsumexpIt is a function in PyTorch used to compute the log-sum-exp of tensor elements.
Function Definition
torch.logsumexp(input, dim, keepdim=False, out=None)
Usage Example
Example
import torch
# Create tensor
x = torch.tensor([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]])
# Compute logsumexp by dimension
y = torch.logsumexp(x, dim=1)
print("logsumexp:", y)
# Create tensor
x = torch.tensor([[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]])
# Compute logsumexp by dimension
y = torch.logsumexp(x, dim=1)
print("logsumexp:", y)
The output result is:
logsumexp: tensor([ 3.4076, 6.4076])
Other Extensions