Re: [Moses-support] Using Input Features (Lattices) Breaks mert-moses

2011-08-29 Thread Graham Neubig
Thanks everyone, and sorry I missed the previous post. Apparently just commenting out the sanity check worked, but it would be nice to have it fixed in a less hacky way. Graham On Mon, Aug 29, 2011 at 9:28 AM, Alexander Fraser wrote: > Hi -- > > If you look back through this list, you will find

Re: [Moses-support] Using Input Features (Lattices) Breaks mert-moses

2011-08-29 Thread Alexander Fraser
Hi -- If you look back through this list, you will find a thread from Anita Gojun describing a hack to fix this (google for lattices mert). Nicola -- are you maintaining this code these days and if so would there be any chance we could get a cleaner fix for this problem that could be integrated i

Re: [Moses-support] Using Input Features (Lattices) Breaks mert-moses

2011-08-29 Thread Barry Haddow
Hi Graham A quick inspection of the code suggests that all should be well if you roll back to a version prior to the clean-up of mert-moses.perl in revision 4149. I think the fix would be to bring the ShowWeights() function in line with OutputNBest() (both are in moses-cmd/src). The latter ch

Re: [Moses-support] Using Input Features (Lattices) Breaks mert-moses

2011-08-29 Thread Nicola Bertoldi
Dear Grahan, have you specified the parameter weight-i in the configuration file? Nicola On Aug 29, 2011, at 5:35 PM, Graham Neubig wrote: > Hi, > > I am trying to run mert-moses.pl on lattice input (using the latest > version from SVN) and am getting the following error after the first > itera

[Moses-support] Using Input Features (Lattices) Breaks mert-moses

2011-08-29 Thread Graham Neubig
Hi, I am trying to run mert-moses.pl on lattice input (using the latest version from SVN) and am getting the following error after the first iteration finishes: --- The decoder returns the scores in this order: d d d d d d d lm w I tm tm tm tm tm Mismatched lambdas. Decoder returned d d d d d d