TF on Cuda 13.2 -- Possible?

I’m wanting to deploy TF on a newer system which defaults to Cudatoolkit 13.2. Wanted to see the possibility of this?