If I have a tensor at size n-by-m. For each of the 1-by-m row, I want to shuffle the m elements in it. Each row is shuffled independently.
Is there a native way in pytorch to do it?
If I have a tensor at size n-by-m. For each of the 1-by-m row, I want to shuffle the m elements in it. Each row is shuffled independently.
Is there a native way in pytorch to do it?