Question about the result dimension of torch.gels


I have a question about torch.gels. In the example, why does the result X has a dimension of 5x2, instead of 3x2, which suits the dimension constrains
in Ax=b?


From the docs on the link you mentioned:

Returned tensor X has shape (max(m,n)×k). The first n rows of X contains the solution. If m>=n, the residual sum of squares for the solution in each column is given by the sum of squares of elements in the remaining m−n rows of that column.

So the answer is only in the first 3 rows, but the errors are in the last 2 rows.

Thank you! Now I figure it out.