Hi, I actually just discovered the cause of the problem. The problem lies with the maximum number of threads mgiza can handle, which is 10 in the current version of it. As long as it doesn't exceed 10, there's no problem with the run-giza/run-giza-inverse stage. When the number is larger than 10, mgiza attaches letters to each part file, which contradicts to the naming rule of those part files, prefix.A3.final.part(n-1) where n is the number of threads, specified on Qin's website.
Cheers, Guchun
_______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
