They only say that they performed nearly 200 sessions per monkey, where each session involves multiple movement task trials. You would have to contact the author for details: mschiebe@ur.rochester.edu
They don't report testing whether the learned associations persist, although that would certainly be interesting.