May I have the log info on submission ID 9708560 please? It seems to have an excessively long run time (and always fails on UW data) even though it was intentionally limited to use only 1000 training cases.