Hi everyone,
I have a doubt about how rule scoring for GHKM string-to-tree rules is
implemented in the current version of Moses. Are the scores normalized
over the source side of the rules only (so rules 1) and 2) would be
competing) or source side and aligned target non-terminals (so 1) and
2) ar
Hi everyone,
I built a toy hiero system for German-English (trained on 50'000 sentences)
and used sigtest filtering (code in contrib/sigtest-filter) to prune the
rule-table. I get a BLEU of 16.21 on the wmt 2009 test set. For comparison,
I built the same system using a script that prunes the rule
Hi to all,
I have trouble building IRSTLM. It looks like it does not work with gcc
4.8.1 (Red Hat 4.8.1-1) . I get the error message below :
lmclass.cpp: In member function 'void lmclass::checkMap()':
lmclass.cpp:190:72: error: 'reallocf' was not declared in this scope
MapScore = (double*)
Hi moses support,
I would like to have the alignments between non-terminals available in
moses-chart after rule-table loading and also during decoding.
Is there an option for this? If not, how shall I best implement it, so that
it does not conflict in any way with the alignments between non-termin
Hi all,
I tried to use MERT with the --pairwise-ranked option but the length
parameter seems to be strangely tuned : If I take the weights obtained with
PRO and use them on the test set I get a bad BLEU score and bad length ratio
(0.96...).
I would like to combine PRO with line MERT :
- I tried t
Hello,
ghkm seems to be implemented for moses. Is there any documentation or
tutorial about this implementation?
Bests,
Fabienne
___
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support
Hello everyone,
I have a problem with adding feature functions in moses_chart : It seems to
work at decoding : it appears in the score breakdown of the ChartHypothesis
and it affects the output.
The problem I have is with mert : my new feature seems to be read but it
does not appear in the n-best
that you've implemented all the abstract methods.
>
> best regards
> Barry
>
> On Friday 26 November 2010 17:58, Fabienne Braune wrote:
> > Hi,
> >
> > I have implemented a new type of phrase dictionnary in order to write my
> > own Get
Hi,
I have implemented a new type of phrase dictionnary in order to write my own
GetChartRuleCollection(...) method. I get the error-message
"/.../mosesdecoder/moses/src/PhraseDictionaryMyImpl.h:62: undefined
reference to `vtable for Moses::PhraseDictionaryMyImpl". When I try to
simply rename an e
Hi,
For aligning europarl and also corpora that are not aligned at paragraph
level, you can try out the following :
http://sourceforge.net/projects/gargantua/ (latest version is
Gargantua1.0b.tar.gz)
It should perform better than Gale and Church although slower. You find a
comparison of speed and
Hi,
I am trying to use moses with 3 different rule tables. It looks like
something goes wrong with finding the glue rules but I cannot find out what
it is.
Could somebody help me with this issue?
I have copied the output of the decoder and my moses.ini file below :
Thanks in advance
Fabienne
-
11 matches
Mail list logo