kawu
commented
11 years ago I think you are right, there should be a possiblity to run Maca on a different machine and then feed Concraft with the Maca output. It should be also possible to use a custom preprocessing pipeline, although that's a more dangerous path (user might want to use a preprocessing pipeline totally different from the one used when training the model). It's also simpler to train the model with Maca, you don't have to remember to reanalyse the training corpus, for example, Concraft will do it for you.
New version doesn't allow for input already preprocessed with MACA.
It would be reasonable to have this option - as the user may have no MACA or Morfeusz installed, or uses different version of Morfeusz.