Hi Cuong You need to set the orientation parameter according to which probabilities you estimated in your new model. The different configurations for lexicalised reordering are described here:
http://www.statmt.org/moses/?n=Moses.AdvancedFeatures#ntoc1 For example, if your model just allows "monotonic" and "non-monotonic" for the orientation, then you should set the orientation to "monotonicity", and make sure you have the correct number of weights. The number of weights depends on the model type, direction and condition, where the posibilities are given by the enums in LexicalReorderingState.h. cheers - Barry On 31/01/13 08:22, Cuong Hoang wrote: > Hi Bary Haddow, > I don't make any modifications to LexicalReordering.cpp, and use the > lastest version of MOSES on September 2012. > I copy my moses.ini file bellow, which I usually use (before I convert > the translation tables to the binary format). > In summarization, my question is: How can I configure the > *Orientation* parameter in moses.ini > to run MOSES decoder to decode on my new models? That new distortion > model is completely different > to the default distortion cost model! > Do you have any suggestion? > Thanks. > C. Hoang > > ######################### > ### MOSES CONFIG FILE ### > ######################### > > # input factors > [input-factors] > 0 > > # mapping steps > [mapping] > 0 T 0 > > # translation tables: table type (hierarchical(0), textual (0), binary > (1)), source-factors, target-factors, number of scores, file > # OLD FORMAT is still handled for back-compatibility > # OLD FORMAT translation tables: source-factors, target-factors, > number of scores, file > # OLD FORMAT a binary table type (1) is assumed > [ttable-file] > 0 0 0 5 /home/hoangcuong2011/Desktop/mosesdecoder/model/phrase-table.gz > > # no generation models, no generation-file section > > # language models: type(srilm/irstlm), factors, order, file > [lmodel-file] > 8 0 5 /home/hoangcuong2011/Desktop/mosesdecoder/corpus.lowercased.en.lm > > > # limit on how many phrase translations e for each phrase f are loaded > # 0 = all elements loaded > [ttable-limit] > 20 > > # distortion (reordering) files > [distortion-file] > *0-0 wbe-msd-bidirectional-fe-allff 6 > /home/hoangcuong2011/Desktop/mosesdecoder/model/reordering-table.wbe-msd-bidirectional-fe.gz* > > # distortion (reordering) weight > [weight-d] > 0.3 > 0.3 > 0.3 > 0.3 > 0.3 > 0.3 > 0.3 > > # language model weights > [weight-l] > 0.5000 > > > # translation model weights > [weight-t] > 0.20 > 0.20 > 0.20 > 0.20 > 0.20 > > # no generation models, no weight-generation section > > # word penalty > [weight-w] > -1 > > [distortion-limit] > 6 > > > > On Thu, Jan 31, 2013 at 4:01 PM, Barry Haddow > <[email protected] <mailto:[email protected]>> wrote: > > On 30/01/13 16:54, Cuong Hoang wrote: > > /ERROR:Lexical reordering model (type > wbe-msd-bidirectional-fe- allff): expected 0 weights, got 4/ > > Hi Cuong > > I'm a bit puzzled by this error message, and from inspecting the > code, I can't really figure out how it happens. > > Which git revision are you using? Did you make any modifications > to LexicalReordering.cpp ? > > cheers - Barry > > -- > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > > > > -- > / > Best Regards, > C. Hoang > > {Mimosa, SMT}@Addict > / -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
