Moses-support Digest, Vol 99, Issue 63

Send Moses-support mailing list submissions to
moses-support@mit.edu

To subscribe or unsubscribe via the World Wide Web, visit
http://mailman.mit.edu/mailman/listinfo/moses-support
or, via email, send a message with subject or body 'help' to
moses-support-request@mit.edu

You can reach the person managing the list at
moses-support-owner@mit.edu

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Moses-support digest..."


Today's Topics:

1. Arabic public corpora for CASMACAT (Mohamed Z)
2. moses on Cygwin (Hieu Hoang)
3. Problems with different decoding paths (Haydar ?mren)


----------------------------------------------------------------------

Message: 1
Date: Tue, 27 Jan 2015 02:33:12 +0200
From: Mohamed Z <muhamadzeid@gmail.com>
Subject: [Moses-support] Arabic public corpora for CASMACAT
To: moses-support@mit.edu
Message-ID:
<CAFwtJrRrV+HP9qEYq5KkcaEv8XrHqDpmrL9Eeu3O=h5Qiio9Uw@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi all,

I have installed CASMACAT desktop and I would like to add an English>Arabic
engine. I see that some langauages are listed there, but mostly European
languages. Is there a chance to add Arabic to the list? That would be
really awesome.

If this is not possible, how can I add my engine or upload it? I see an
upload button there, but I have no clue?

thanks,
Mohamed
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20150127/7c2cb02e/attachment-0001.htm

------------------------------

Message: 2
Date: Tue, 27 Jan 2015 11:31:46 +0000
From: Hieu Hoang <Hieu.Hoang@ed.ac.uk>
Subject: [Moses-support] moses on Cygwin
To: moses-support <moses-support@mit.edu>
Message-ID:
<CAEKMkbg__p3XLf125eei=BpuC6GPZQtnAtPNQ3L_q8WAbjwafw@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi All

Is anyone using Moses on Cygwin at the moment?

We currently testing the latest Moses code for release soon. If anyone has
experience with moses/cygwin, it would be great if you can help us test it.

--
Hieu Hoang
Research Associate
University of Edinburgh
http://www.hoang.co.uk/hieu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20150127/2444f299/attachment-0001.htm

------------------------------

Message: 3
Date: Tue, 27 Jan 2015 17:18:42 +0100
From: Haydar ?mren <haydarimren@gmail.com>
Subject: [Moses-support] Problems with different decoding paths
To: moses-support@mit.edu
Message-ID:
<CAP3OHV6KOfuZgg=pLttSXrqJ8BQ0MrDdhZZhYmtVqOLRffb6Jw@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi all,

I have a problem with tuning step after training a factored model with
different decoding steps.
Here are the different scenarios I tried:


*First Scenario:*

I have 3 factors on both sides of the parallel data.
My translation factors are: 1-1+2-2+0-0
My generation factors are: 1,2-0+0-1,2
My decoding steps are: t0,t1,g0:t2,g1
Training finished well and moses.ini and all phrase tables and generation
tables are created.
But when I want to run tuning, first of all it says there is no reordering,
how can I add reordering to factored training?


*Second Scenario:*

I tried giving -reordering-factors 0-0. And then I get a segmentation fault
in tuning.


*Third Scenario:*

If I just give one of the translation paths in the decoding steps then
tuning runs ok but it the phrase and generation tables copied under
mert-work/filtered are 0 bytes.

Did anyone encounter something similar?

Kind regards,
Haydar Imren
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20150127/930fc21e/attachment-0001.htm

------------------------------

_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support


End of Moses-support Digest, Vol 99, Issue 63
*********************************************

Tweet