How can I make the length of tensor doubled when filter size N using Transposed 1dconv?

my question is the title.
How can I make the length of tensor doubled when filter size N using Transposed 1dconv?
thank you for giving your attention to my question.