[Moses-support] A question with train-factored-phrase-model.perl

[yifeng_yu]

Hi,
nbsp;
Following the first example innbsp;nbsp;Factored Tutorial,nbsp;I enter:
nbsp;
train-factored-phrase-model.perl \
--corpus factored-corpus/proj-syndicate.1000 \
--root-dir unfactorednbsp; \
--f de \
--e en \
--lm 0:3:/cygdrive/c/usr/phrase/factored-corpus/surface.lm:0
nbsp;
But I get the error message back:
you have specify which languages to condition on (f or fe); it is not done in 
distance

I found the related source code in train-factored-phrase-model.perl:
nbsp;
nbsp;(1)
nbsp;nbsp;nbsp;nbsp; elsif ($reoconf =~ /^((fe)|(f))/) {
nbsp;nbsp;nbsp;nbsp;nbsp;nbsp;nbsp; 
$REORDERING_MODELS[$model_num]{lang} = $reoconf;
nbsp;nbsp;nbsp;nbsp;nbsp; }
nbsp;
(2)nbsp;
nbsp; #check that the required attributes are given
nbsp; if (!defined($REORDERING_MODELS[$model_num]{lang})) {
nbsp;nbsp;nbsp;nbsp; print STDERR you have specify which languages to 
condition on (f or fe); it is not done in $r\n;
nbsp;nbsp;nbsp;nbsp; exit(1);
nbsp; }
nbsp;
I think you have specify ...nbsp;may be you have to specify...
nbsp;
My question is: should I really need to add a parameter to specify which 
languages to condition on?
nbsp;
nbsp;
Regards,
Yifeng Yu___
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support

[Moses-support] What is the use of the lm parameter in the model training stage?

[yifeng_yu]

nbsp;In Factored Tutorial, the first example is:
nbsp;
nbsp;% train-model.perl \
nbsp;nbsp;nbsp; --corpus factored-corpus/proj-syndicate \
nbsp;nbsp;nbsp; --root-dir unfactored \
nbsp;nbsp;nbsp; --f de --e en \
nbsp;nbsp;nbsp; --lm 0:3:factored-corpus/surface.lm:0

I think the language model is usually used in the decoding stage in SMT. What 
is the use of the lm parameter which lists a language modelnbsp;in the model 
training stage?___
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support