Thank you, my loop refinement calculation has ended successfully now, however...
You're only building one loop model. I'd recommend building several, so
that you can sample the conformational space more effectively. This also
allows for the possibility that some of the models are not produced due
to optimizer failure, as you have discovered.