Hi Jelita
in the new moses, you should use
-weight-overwrite 'TranslationModel0= 1 LM0= 0.8 Distortion0= 0.3
WordPenalty0= -0.2'
this is equivalent to
-weight-t 1 -weight-l 0.8 -weight-d 0.3 -weight-w -0.2
in the old format.
you moses.ini should also be in the new format. Don't mix old and new
format.
On 29/08/2013 10:22, Jelita Asian wrote:
Hi Hieu,
It was my mistake
I just realized that I put my running moses script in 2 lines:
./moses.exe -f bin_id/moses.ini -weight-t 1 -weight-l 0.8 -weight-d 0.3
-weight-w -0.2 -input-file $1.tmp1 > $2.tmp
instead of one line like
./moses.exe -f bin_id/moses.ini -weight-t 1 -weight-l 0.8 -weight-d
0.3 -weight-w -0.2 -input-file $1.tmp1 > $2.tmp
It must be entered by accident.
If we use the new moses, can we still use this command
./moses.exe -f bin_id/moses.ini -weight-t 1 -weight-l 0.8 -weight-d
0.3 -weight-w -0.2 -input-file $1.tmp1 > $2.tmp
but use the new moses.ini with alternate weight setting?
Thanks.
On Thu, Aug 29, 2013 at 4:17 PM, Hieu Hoang <[email protected]
<mailto:[email protected]>> wrote:
what is the EXACT command you entered? Can you also please attach
your ini file.
did you provide the decoder with a file of sentences you wish to
translate? if not, it wait for you to enter a sentence to decode
On 29 August 2013 10:10, Jelita Asian <[email protected]
<mailto:[email protected]>> wrote:
Hello, any pointer what's wrong? I need to press enter for it
to go through to the next line. Is that something wrong with
the phrase table?
Best regards,
Jelita
On Wed, Aug 28, 2013 at 5:52 PM, Jelita Asian
<[email protected] <mailto:[email protected]>> wrote:
When I run the moses command, it suddenly stops with this
error:
Defined parameters (per moses.ini or switch):
config: bin_id/moses.ini
distortion-file: 0-0
wbe-msd-bidirectional-fe-allff 6 bin_id/reordering-
table.Corpus27August2013.for_train.id-en.wbe-msd-bidirectional-fe
distortion-limit: 6
input-factors: 0
lmodel-file: 1 0 4
bin_id/en.LM-Corpus27August2013.for_train-IRSTLM-4-1-
improved-kneser-ney-0-1.blm.mm
<http://improved-kneser-ney-0-1.blm.mm>
mapping: 0 T 0
ttable-file: 1 0 0 5
bin_id/phrase-table.Corpus27August2013.for_train.id
<http://phrase-table.Corpus27August2013.for_train.id>
-en
ttable-limit: 20
weight-d: 0.3 0.3 0.3 0.3 0.3 0.3 0.3
weight-l: 0.8
weight-t: 1 0.2 0.2 0.2 0.2
weight-w: -1
.
ScoreProducer: Distortion start: 0 end: 1
ScoreProducer: WordPenalty start: 1 end: 2
ScoreProducer: !UnknownWordPenalty start: 2 end: 3
Loading lexical distortion models...have 1 models
ScoreProducer:
LexicalReordering_wbe-msd-bidirectional-fe-allff start: 3
end: 9
Creating lexical reordering...
weights: 0.300 0.300 0.300 0.300 0.300 0.300
binary file loaded, default OFF_T: -1
Start loading LanguageModel
bin_id/en.LM-Corpus27August2013.for_train-IRSTLM-4-1
-improved-kneser-ney-0-1.blm.mm
<http://improved-kneser-ney-0-1.blm.mm> : [0.322] seconds
In LanguageModelIRST::Load: nGramOrder = 4
Language Model Type of
bin_id/en.LM-Corpus27August2013.for_train-IRSTLM-4-1-impr
oved-kneser-ney-0-1.blm.mm
<http://oved-kneser-ney-0-1.blm.mm> is 1
Language Model Type is 1
blmt
loadbin()
lmtable::loadbin_dict()
dict->size(): 52515
loadbin_level (level 1)
mapping 52515 1-grams
tableOffs 543479 tableGaps19191-grams
done (level 1)
loadbin_level (level 2)
mapping 758553 2-grams
tableOffs 1331204 tableGaps20484-grams
done (level 2)
loadbin_level (level 3)
mapping 1690187 3-grams
tableOffs 12709499 tableGaps61051-grams
done (level 3)
loadbin_level (level 4)
mapping 2146354 4-grams
tableOffs 38062304 tableGaps51424-grams
done (level 4)
done
OOV code is 52514
IRST: m_unknownId=52514
ScoreProducer: LM start: 9 end: 10
Finished loading LanguageModels : [0.630] seconds
Start loading PhraseTable
bin_id/phrase-table.Corpus27August2013.for_train.id-en
: [0.634] seconds
filePath:
bin_id/phrase-table.Corpus27August2013.for_train.id-en
ScoreProducer: PhraseModel start: 10 end: 15
Finished loading phrase tables : [0.641] seconds
IO from STDOUT/STDIN
Created input-output object : [0.645] seconds
However, sometime when I run the same code without
changing anything, it seems OK. Why is that?
Thanks.
Best regards
Jelita
_______________________________________________
Moses-support mailing list
[email protected] <mailto:[email protected]>
http://mailman.mit.edu/mailman/listinfo/moses-support
--
Hieu Hoang
Research Associate
University of Edinburgh
http://www.hoang.co.uk/hieu
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support
