Hi Hieu,
it is here in the repository:
https://github.com/moses-smt/mosesdecoder/blob/master/scripts/server/moses.py
Regards
Oskar Solich
IT and Organisation
tsd Technik-Sprachendienst GmbH
Übersetzung & Dokumentation
Mittelstraße 12 - 14
50672 Köln
Tel. +49-221-92 59 86-34
Fax +49-221-2 57 73
since you have 2 factors on the target side, the binarization command
should be
CreateOnDiskPt 1 *2* 4 100 2
If you also have 2 factors on the source side, then it should be
CreateOnDiskPt *2* *2* 4 100 2
Hieu Hoang
http://www.hoang.co.uk/hieu
On 12 October 2015 at 08:20, Tomasz Gawryl
I'm not sure if anyone else encounters it but the compact lexical
reordering table crashes for me on OSX/clang during loading.
The stack trace i have for this is
LexicalReorderingTableCompact::LexicalReorderingTableCompact
LexicalReorderingTableCompact::Load line 180
Hi,
in t2, you do generate an output lemma factor - which may be the cause of
this problem (even though you do not seem to use the output lemma anywhere
else).
Does it still core dump, if you change translation factors to:
translation-factors = "lemma -> lemma, pos -> pos, word -> word + lemma
Hi Hieu,
Command below (I hope this is it ;).
Regards,
Tomek
moses@SKR-moses:~/working/experiments/FACTORED/steps/1$ more
TRAINING_build-ttable.1
#!/bin/bash
[...]
/home/moses/src/mosesdecoder/scripts/training/train-model.perl -mgiza
-mgiza-cpus 8 -cores 8 -parallel -sort-buffer-size 10G
Hi,
Chris Dyer's fast align also allows for model dump & force alignment:
https://github.com/clab/fast_align
I found this more reliable than mgiza, which crashes quite often during
fast-alignment.
-phi
On Wed, Oct 7, 2015 at 8:38 AM, DUGAST, LOIC wrote:
> Hello
>
>
>
Hello,
Pretty sure there is no academic importance to this, but :
For the tokenizer we have the -x option to skip XML/HTML tags
For the detokenizer it WILL SKIP whatever.
cf :
while() {
if (/^<.+>$/ || /^\s*$/) {
#don't try to detokenize XML/HTML tag lines
Hi,
to translate a set of English sentences, you have to put them into a
separate file, say, "english-input.txt", and then run the decoder as
follows:
~/mosesdecoder-RELEASE-3.0/bin/moses -f ~/working/train/model/moses.ini <
english-input.txt
-phi
On Thu, Oct 8, 2015 at 10:59 PM, Apurva Joshi
Hello all,
I am trying to train a factored model but I don't remember how to set right
definition of factors. I am changing toy config.factored file.
I have data of type "Word1|NN Word2|VB Word3|JJ " on the target side
with no factors on the source side. It will be great if someone can tell
Hello,
I need to skip factors from an external pos-tagger like in the
config.factored. Instead I need it work it for my data.
Sample training sentence :
Source : Pressure of excretion remains .
Target : kupoRaNa|r6 ke|na kaI|na rUpa|k1s hEM|main .|na
I am attaching my config file along with
10 matches
Mail list logo