Re: [Moses-support] Compiling moses with SRILM, checking for trigram_init in -loom... no

Try KenLM. Run ./configure (no argument), change your moses.ini so the first digit is 8, and get the same results with less time, memory, and compilation headache. If you still want to use SRI with moses: Is your machine actually 64-bit but SRI annoyingly decided to compile 32-bit? If so, modif

Re: [Moses-support] Lower scores with Word Lattice

Thanks Chris, Hieu. These should be identical. However, it is possible that the beam > search used by the decoder might run slightly differently in the two > cases, leading to different results. With no pruning and unlimited > stack sizes, the results should not be different (unless you have > fea

[Moses-support] Compiling moses with SRILM, checking for trigram_init in -loom... no

Hi, I am trying to compile Moses. After: ./configure --with-srilm=/home/jon/tools/srilm I get an error: checking for trigram_init in -loom... no configure: error: Cannot find SRILM’s library in /home/jon/tools/srilm/lib/i686 If I then look at the directory in question, I see: libdstruct.a libf

Re: [Moses-support] Inverse giza crash

Hi all, Thanks for your help I installed the new gizapp(1.0.5 version) i was running on the 1.0.4 version I also checked the option -DBINARY_SEARCH_FOR_TTABLE as suggested by Qin Gao and I got the whole training process done. 1. I also had a question that how many iterations does the tuning step t

Re: [Moses-support] Sentence mismatch error

Hi, the error clearly indicates that there is a sentence length ratio problem, so I am afraid that you will have to track down what went wrong there. Maybe a sentence got lost somewhere, maybe parsing shortened the sentence, etc. Make sure that you have the latest version of the code first. -ph

Re: [Moses-support] Proposal to replace vertical bar as factor delimeter

In the scripts directory, several places have | hardcoded. If someone wants to replace it, egrep -R "['\"\\]\\|" * can help spot them (it checks | after a quote or a \ ; there is also a lot of unrelated things, and probably some places are not spotted). I think having the possibility to choose the

[Moses-support] Sentence mismatch error

Hi, I'm training the syntactic model, but I have some problems when I run inverse giza . I clean the data before and after running the Collins parser , but inverse-giza has fertility problems: WARNING: The following sentence pair has source/target sentence length ration more than the maximum allo