On Thu, May 31, 2018 at 11:01 AM, tsegay kiros wrote:
>
>
> On Mon, May 28, 2018 at 12:32 PM, tsegay kiros wrote:
>
>> tse...@gmail.com
>>
>> ___
>> Moses-support mailing list
>> Moses-support@mit.edu
>>
On Mon, May 28, 2018 at 12:32 PM, tsegay kiros wrote:
> tse...@gmail.com
>
> ___
> Moses-support mailing list
> Moses-support@mit.edu
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>
build.log.gz
Description: application/gzip
hello there
how is every thing going on I hope it would be nice
I an focused which alignment of IBM is used in the Moses since I want to
write my documentation
On Tue, May 2, 2017 at 10:58 AM, G/her G/libanos wrote:
> hello there
>
>
>
> I am doing well and your
hello there
I am doing well and your contribution is high
my adviser asks me to align the training system using phrase rather than
word based using Giza++
what package can I use to do phrase based alignment
thanks...
On Thu, Apr 13, 2017 at 1:37 AM, G/her G/libanos
hello there
how is every thing
I want to uses the m4loc to split the corpus into train, testing and tuning
randomly
but I didn't get how to apply it.
the idea get from this paper
Phoneme-based English-Amharic Statistical Machine Translation
thanks..
On Mon, Mar 27, 2017 at 3:42
hello there...
I want to ask you some thing
1. in language modeling for the monolingual data using IRSLM what value n
of the N-gram is more value?
2. in the training of the parallel corpus how I can isolate the testing and
tuning data randomly from the total corpus, since choosing myself is not
hello there I would like thanks your contribution in my work...
now I have one problem to create web-based translation system
when I check using the Netcat by following the manual in the
https://github.com/moses-smt/mosesdecoder/tree/master/
contrib/iSenWeb/Introduction
it display these result and
Hi,
Did you perhaps forget the leading "/" in the file path to an existing
installation of XML-RPC?
Regards
Mathias
On Sat, Feb 25, 2017 at 10:46 AM, G/her G/libanos
wrote:
> hello there I am doing on the Moses server but on installing the xmlrpc
> create an error
> what
hello there I am doing on the Moses server but on installing the xmlrpc
create an error
what I can do
thanks for every thing you doing all
On Wed, Feb 8, 2017 at 3:07 AM, Marwa Refaie wrote:
> For web server dig little on those links I think could help
>
>
For web server dig little on those links I think could help
http://www.statmt.org/moses/?n=Moses.WebTranslation
https://github.com/moses-smt/mosesdecoder/tree/master/contrib/iSenWeb/Introduction
Marwa N Refaie
On 7 Feb 2017, at 21:05, G/her G/libanos
You should use the moses server. More details here
http://www.statmt.org/moses/?n=Advanced.Moses#ntoc1
On 07/02/2017 19:03, G/her G/libanos wrote:
the Moses decoder works for our system on Ubuntu using terminal
but we want to make user interactive whether in the form of web page
or
the Moses decoder works for our system on Ubuntu using terminal
but we want to make user interactive whether in the form of web page or
application based
could you help me any information to done that
10qu...
On Mon, Dec 12, 2016 at 8:43 AM, G/her G/libanos wrote:
> we
Hi G/her,
Look for the XMLRPC plugin and add it to your installation directory. It
allows moses-server - web client integration.
I think the default moses installation already contains it. Otherwise, add
it using --with-xmlrpc extension. Don't look for a Windows application..
Windows is shit
we create the translation model and we need to uses the window application
using
http://www.statmt.org/moses/?n=Moses.Packages
but when we import our model into the mainWindow application it create an
error
what we can do.
since it works for EN into Fr
10qu for ...
On Mon, Dec 12, 2016 at 7:17
Hi G/her,
It means you're not running the command in the right dir. Or the file does
not have the correct permission. Please be sure of the location of
train-model.perl and that it has the correct executable permission. I'll
give you details once I am back to my desk.
On Saturday, December 3,
hi...
this is G/her
when I try to train it creates this error for the
train-model.perl -external-bin-dir $HOME/mosesdecoder/tools
On Fri, Nov 25, 2016 at 4:25 PM, G/her G/libanos wrote:
> hello there
>
> I am doing my research on local Ethiopian language Amharic-Tigrigna
Hi Gebregziabher,
You'll need to create a script of your own to "tokenize" ge'ez languages
since ge'ez punctuation are different from that of English.
for eg.
sed 's/።/ ።/g' < corpus.tg > corpus.tok.tg
On Fri, Nov 25, 2016 at 4:25 PM, G/her G/libanos wrote:
> hello there
Why do i get " Use of uninitialized value in string eq at
/home/mosesdecoder/scripts/Transliteration/clean.pl line 139, <$IN> line
1." while training transliteration model ... what is wrong
On Fri, May 6, 2016 at 4:20 PM, Sanjanashree Palanivel <
sanjanash...@gmail.com> wrote:
> I installed
Hi,
it's probably a good idea to use the full path instead of "-root-dir
train". But what is in "training.out"? That should give some clues. Are any
files created?
-phi
On Wed, Mar 9, 2016 at 11:57 PM, BIRENDRA CHAUHAN SINGH <
birendrachauh...@gmail.com> wrote:
> on running this command:
>
>
Hello Sanjanasri,
Basically, you can forget all results that you obtained without tuning.
They are not a meaningful indicator of the quality of NPLM. If you add a
new language model, the weight of the other language models, translation
models etc. needs to be balanced accordingly, and that is
Hi Sanjanasri,
1) your corpus is very small, and you may have to use more iterations of
NPLM training and smaller vocabulary sizes. Just to double-check, are
you tuning your systems? MERT (or PRO or MIRA) should normally ensure
that adding a model doesn't make BLEU go down.
2) I'm not sure
Hi Sanjanasri,
we have seen improvements in BLEU from having both KENLM and NPLM in our
system. Things can go wrong during training though (e.g. a bad choice of
hyperparameters (vocabulary size, number of training epochs)). I
recommend using a development set during NPLM training, and
Hi Sanjanasri,
if you first compiled moses without the option '--with-nplm', and then
add the option later, the build system isn't smart enough to know which
files it needs to recompile. if you change one of the compile options,
use the option '-a' to force recompilation from scratch.
best
Hi Sanjanasri,
this error occurs when Moses was compiled without the option '--with-nplm'.
best wishes,
Rico
On 15.09.2015 15:08, Sanjanashree Palanivel wrote:
Dear Rico,
I updated moses and NPLM has been compiled succesfully
with moses. However, when I perform decoding I am
Hello Sanjanasri,
this looks like a version mismatch between Moses and NPLM. Specifically,
you're using an older Moses commit that is only compatible with nplm 0.2
(or specifically, Kenneth's fork at https://github.com/kpu/nplm ).
If you use the latest Moses version from
I think that Hieu has an Eclipse project.
On Thu, Feb 26, 2015 at 7:00 PM, Tom Hoar
tah...@precisiontranslationtools.com wrote:
Not to my knowledge. Your contribution of one would be welcome.
On 02/27/2015 07:27 AM, Winston Yuen wrote:
Is there Xcode or Eclipse project for Moses? I try to
are you sure your version of IRSTLM is 5.80.06? Is there a reason you
need to use IRSTLM, as opposed to KENLM, which is built into Moses?
On 01/01/15 14:11, Dawit Mulugeta wrote:
: ‘irstlm’ is not a
___
Moses-support mailing list
I'm getting a different error when compiling irstlm5.80.06 with the
latest moses from github.
moses/LM/IRST.cpp:60:21: error: invalid use of incomplete type
‘class lmContainer’
if (m_lmtb) m_lmtb-reset_mmap();
Using irstlm5.80.03 works fine
Hi,
it is true that metrics like BLEU that only count words as correct when
they are identically
and do not give partial credit to morphological variants are more
appropriate for English
and other less inflected languages. Scores for English are typically
higher, which does
reflect to some degree
Hi,
one recently proposed successful strategy is to first translate the
source sentences in the parallel corpus with the rule-based system (in
this case Apertium) and then build a statistical system from the
rule-based output to the correct translation in the parallel corpus.
This is referred to
Hi,
Ah, in that case it can actually cause problems: your training data should
always be formatted in the same way as your dev/test data.
2 possibilities:
- re-tokenize training data with the actual tokenizer script to have the
same mark-up (then retrain your system)
- re-tokenize your dev/test
Ok,
Thank you
Bests
Cyrine
2014-02-21 14:55 GMT+01:00 Thomas Meyer ithurts...@gmail.com:
Hi,
Ah, in that case it can actually cause problems: your training data should
always be formatted in the same way as your dev/test data.
2 possibilities:
- re-tokenize training data with the
These are some things you can try:
1. collect more data
2. collect more data, of the same genre as your test set
3. use hierachical model
4. create a better tokenizer for your language
5. replace names, numbers and dates with placeholders.
If your data has lots of names, numbers and
Hi,
you can try many different model configurations, but yes, having more
training data will certainly help.
It will be helpful for you you explore a bit more how each translations
was formed and why the translation decisions were made by digging
deeper into the models and the decoding process.
-support-boun...@mit.edu per conto di
Hieu Hoang hieu.ho...@ed.ac.uk
Inviato: domenica 26 gennaio 2014 10:48
A: weihua wang
Cc: moses-support
Oggetto: Re: [Moses-support] (no subject)
Did IRSTLM compile ok? Also, the latest IRSTLM is 5.80.03, which you can
download from here:
http
:* domenica 26 gennaio 2014 10:48
*A:* weihua wang
*Cc:* moses-support
*Oggetto:* Re: [Moses-support] (no subject)
Did IRSTLM compile ok? Also, the latest IRSTLM is 5.80.03, which you
can download from here:
http://sourceforge.net/projects/irstlm/
Here's some instructions on compiling IRSTLM
Hi,
TAUS put together a basic slide presentation:
https://www.taus.net/press-releases/free-open-source-machine-translation-tutorial-is-made-available-by-taus
-phi
On Tue, Dec 10, 2013 at 11:27 AM, Kalyani Baruah kajubarua...@gmail.com wrote:
hii
Can you provide me with a ppt(power point
Hi,
the purpose of the language model is to encourage fluent output and
the purpose of the translation model is to encourage similarity between
input and output.
-phi
On Mon, Dec 9, 2013 at 8:37 AM, Kalyani Baruah kajubarua...@gmail.com wrote:
hii,
What is the work of language
thanks
Regards,
*Kalyanee Kanchan Baruah*
Department of Information Technology,
Institute of Science and Technology,
Gauhati University,Guwahati,India
Phone- +91-9706242124
On Mon, Dec 9, 2013 at 4:21 AM, Philipp Koehn pko...@inf.ed.ac.uk wrote:
Hi,
the purpose of the language model is to
that's strange. I'm using ubuntu and it's working for me. You may hace to
look around on the internet for a fix.
the EMS will still run even when these errors occur, but it would be nice
to fix them and see the images
On 3 December 2013 03:44, amir haghighi amir.haghighi...@gmail.com wrote:
ok, i've added it back
https://github.com/moses-smt/mosesdecoder/commit/21c51194fab91f8e79409545db98642cbac9b505
let me know if it works for you
On 12 March 2013 19:05, Greg Hanneman ghanne...@gmail.com wrote:
Hi Hieu,
Sorry about the delay in writing back -- I've had a busy last few days.
Hi Hieu,
Sorry about the delay in writing back -- I've had a busy last few days.
Yes, I think it would be great to have the -print-alignment-info flag
re-implemented. We could in theory get the alignment information from a
separate output file, but it would be much more convenient to have it as
hi greg
printing word alignment is still a feature. In fact it's been cleaned
up, optimized, debugged, and now on by default. It works for both pb,
and hiero/syntax models.
The 2 arguments are
-print-alignment-info-in-n-best [0 or 1]
and
-alignment-output-file [file]
The arg (
ah, i didn't misread your email.
i deleted both
-use-alignment-info
-print-alignment-info
but -print-alignment-info wasn't redundant.
I can reimplement
-print-alignment-info
if you can't get the info from using
-alignment-output-file [file]
On 07/03/2013 00:23, Hieu Hoang wrote:
hi
It is possible to design an interface, right? is there an application for
designing the interface to be used in ubuntu 11.04...please help...i would
appreciate any kind of response...thank you!!
On Sun, Dec 23, 2012 at 9:14 AM, Eleni Teshome tesh.el...@gmail.com wrote:
Dear all,
i've been
if you want to create a server so that you can write client gui programs
to connect to it, look in
contrib/server
there is also sample java clients and a web server client in there too
On 24/12/2012 12:29, Eleni Teshome wrote:
It is possible to design an interface, right? is there an
Hi Herry
I would strongly recommend that you use the standard boost packages from
your distribution - I think ubuntu 12.04 is on version 1.46 of boost.
Probably boost is installed in a non-standard location, or in a
non-standard layout. What happens if you run the following?
g++ -dM -x
More to the point, what happens without the 2/dev/null ?
When installing Boost on your own, the only supported method is in
BUILD_INSTRUCTIONS.txt .
Kenneth
On 09/24/12 10:54, Barry Haddow wrote:
Hi Herry
I would strongly recommend that you use the standard boost packages from
your
Simenesh,
You might try DoMY CE. You'll find the link here:
http://www.statmt.org/moses/?n=Moses.Packages [1]
On 2012-09-21 21:14,
Simenesh Tolossa wrote:
Dear All,
I have been trying to install
moses and i followed all the instructions but i cant seem to get those
files in the bin
hi eleni
a few prblems i see:
1. don't use relative path, eg
~/mosesdecoder/dist
use absolute path, eg.
/home/eleni/mosesdecoder/dist
2. the latest code doesn't have dist directory anymore. All the
documentation should have been updated. If you see any
Hi,
the comment at the beginning on the section on specifying the external
bin dir containing GIZA++, mkcls, etc., is meant to inform that
whenever you run any of the invocations of train-model.perl described
later in the instructions, you must also include this specification.
Running
Hi Eleni
Did the tokenisation step succeed? What files did it produce?
cheers - Barry
On 31/07/12 15:51, Eleni Teshome wrote:
Hi,
I started usin the site http://www.statmt.org/moses/?n =moses.baseline and
while doing the corpus preparation, in the second step i.e. The true casing
step,
Hi Eleni
You didn't say which files the tokeniser produced. Check the tokenisation
command carefully, you may have mis-typed
Cheers - Barry
Sent from my ZX81
- Reply message -
From: Eleni Teshome tesh.el...@gmail.com
Date: Tue, Jul 31, 2012 16:52
Subject: [Moses-support] (no subject)
Hi Abbasov
It's explained in this paper
http://homepages.inf.ed.ac.uk/pkoehn/publications/phrase2003.pdf
best regards - Barry
On Thursday 17 February 2011 11:18, Abbasov Samir wrote:
Hi everyone, in phrase-table has five scores, i would know how is
computed lexical weighting ( lex(e|f) ) ?
better if you ask your advisor.
M.
On Nov 21, 2008, at 7:06 AM, Mekh jaishi wrote:
can anybody say how can i run giza++ for Natural language procesing.
Invite your mail contacts to join your friends list with Windows
Live Spaces. It's easy! Try it!ATT1.txt
Hi,
GIZA++ is run by the train-factored-model.perl script, so you do
not need it in isolation. But if you want get some idea how it is
run, please check how the training script does it.
-phi
On Fri, Nov 21, 2008 at 6:06 AM, Mekh jaishi [EMAIL PROTECTED] wrote:
can anybody say how can i run
56 matches
Mail list logo