We haven't actually run the thing to completion. We ran it for a couple days and then stopped after finding tens of thousands of solutions. Might explore again if we have time to improve the solver efficiency.
I did write a program to generate an exact cover. Thr output is a little over 5 megabyte when using a notation that list the indices of the positions that have a 1. I have not yet verified that it is correct. My solver found one solution so far and edtimates that the number of solutions is more than 10^50.