Moses-support Digest, Vol 95, Issue 13

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. Native MS Window for Moses Release 3.x, MGIZA++ and IRSTLM
(Tom Hoar)
2. Re: Incremental retraining (Sandipan Dandapat)
3. AMTA 2014: Call for Participation (Marciano, Jay)


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

Message: 1
Date: Mon, 08 Sep 2014 16:07:40 +0700
From: Tom Hoar <tahoar@precisiontranslationtools.com>
Subject: [Moses-support] Native MS Window for Moses Release 3.x,
MGIZA++ and IRSTLM
To: moses-support@mit.edu, Hieu Hoang <Hieu.Hoang@ed.ac.uk>
Message-ID: <540D71DC.5080600@precisiontranslationtools.com>
Content-Type: text/plain; charset="utf-8"

Hi all,

PTTools is starting an effort to update/add native 64-bit MS Windows
(i.e. Without Cygwin) support to these three Moses components. Our goal
is to complete the updates in time to be included in the MosesCore
Release 3.x. We believe the Moses trunk is at a point where there should
be minimal updates/changes before the completion of Release 3.

* Moses Release 3.x - we will start with the current dev trunk. Our
goal is to update train-model.perl and mert-moses.pl and all Perl
and binary sub-components. We'll start with the existing Visual
Studio projects which are reported to compile on MS Windows.
* MGIZA++ - this is reported to compile/run on MS Windows. However, we
have identified one bug that will cause the runtime to crash on
hosts with more than 10 cores. We will fix this and submit the
updates to merge into the trunk.
* IRSTLM - unknown status of MS Windows support. We suspect this
package will require more work. Again, we will submit the updates to
the IRST team if they would like to merge into the trunk.

With this message, I'm soliciting community support for the project.
Specifically:

1. What features do you believe are "must-have" for a native Windows
version? I'm targeting the moses, moseschart and mosesserver
binaries and the core training/tuning scripts and binaries. Many
scripts and project code in the scripts and contrib trees will not
be touched.
2. What Windows platforms are desirable? We're planning 64-bit Windows
XP through Windows 8.1, Windows Server 2008 R2. Are there others?
3. What about compact phrase tables? Is this important?
4. Binarize utilities for phrase/reordering and hierarchical models?
5. Within the decoder, we're planning to support only KenLM.
--with-irstlm, --with-srilm, and others will not be added/supported
unless its already there.

I'm happy to hear from anyone who has an interest in the project at any
level.
--

Best regards,
Tom Hoar
Managing Director
*Precision Translation Tools Co., Ltd.*
Bangkok, Thailand
Web: www.precisiontranslationtools.com
<http://www.precisiontranslationtools.com>
Mobile: +66 87 345-1875
Skype: tahoar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20140908/f1dd3a85/attachment-0001.htm

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

Message: 2
Date: Mon, 8 Sep 2014 10:11:44 +0100
From: Sandipan Dandapat <sandipandandapat@gmail.com>
Subject: Re: [Moses-support] Incremental retraining
To: Hieu Hoang <Hieu.Hoang@ed.ac.uk>
Cc: moses-support <moses-support@mit.edu>, Ulrich Germann
<ulrich.germann@gmail.com>
Message-ID:
<CAGr2oZSC6eu4ny+GRa-t9THV3CJ-OJz-3uG5vtkVq2obuqwFOg@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi,
This worked.
Thanks and regards,
sandipan

On 7 September 2014 16:09, Hieu Hoang <Hieu.Hoang@ed.ac.uk> wrote:

> sorry, i meant
>
>
> On 7 September 2014 16:08, Hieu Hoang <Hieu.Hoang@ed.ac.uk> wrote:
>
>> you HAVE to change both
>> num-features=7
>> AND
>> [weight]
>> PT0= 0.1 0.2 0.3 0.4 0.5 0.6 0.7
>>
>>
>> On 7 September 2014 16:06, Sandipan Dandapat <sandipandandapat@gmail.com>
>> wrote:
>>
>>> Hi Hieu,
>>> Even I tried with '7' and it fails with the error message
>>>
>>> Exception: moses/ScoreComponentCollection.cpp:248 in void
>>> Moses::ScoreComponentCollection::Assign(const Moses::FeatureFunction*,
>>> const std::vector<float>&) threw util::Exception'.
>>> Feature function PT0 specified 7 dense scores or weights. Actually has 4
>>>
>>> In contrast, when I am using binarised pharse table, I use
>>> num-features=4 and this works fine. I am attaching the Moses.ini file in
>>> case I am doing anything wrong there.
>>>
>>> Thanks and regards,
>>> sandipan
>>>
>>>
>>> On 7 September 2014 15:46, Hieu Hoang <Hieu.Hoang@ed.ac.uk> wrote:
>>>
>>>> maybe's it's 7 scores
>>>>
>>>>
>>>> On 7 September 2014 14:59, Sandipan Dandapat <
>>>> sandipandandapat@gmail.com> wrote:
>>>>
>>>>> Hi Hieu,
>>>>> I also tried the same but generates the error below:
>>>>>
>>>>> Exception: moses/TranslationModel/UG/mmsapt.cpp:381 in virtual void
>>>>> Moses::Mmsapt::Load() threw util::Exception because
>>>>> `this->m_feature_names.size() != this->m_numScoreComponents'.
>>>>> At moses/TranslationModel/UG/mmsapt.cpp:381: number of feature values
>>>>> provided by Phrase table (7) does not match number specified in Moses
>>>>> config file (4)!
>>>>>
>>>>> Thanks and regards,
>>>>> sandipan
>>>>>
>>>>>
>>>>> On 6 September 2014 09:50, Hieu Hoang <hieuhoang@gmail.com> wrote:
>>>>>
>>>>>> I'm not sure how many scores there are in the phrase table
>>>>>> PhraseDictionaryBitextSampling
>>>>>> It may be 4. In which case you must specify
>>>>>>
>>>>>> [feature]
>>>>>> PhraseDictionaryBitextSampling name=PT0 num-features=4 ...
>>>>>>
>>>>>> [weight]
>>>>>> PT0= 0.1 0.2 0.3 0.4
>>>>>>
>>>>>>
>>>>>> On 05/09/14 14:12, Sandipan Dandapat wrote:
>>>>>>
>>>>>> Hi,
>>>>>> During incremental retraining I specified the following line in
>>>>>> moses .ini
>>>>>> PhraseDictionaryBitextSampling name=PT0 output-factor=0
>>>>>> num-features=9
>>>>>> path=/home/sandipan/inc_retrain/MT_sys/EnPl/mtdata_pro/train. L1=en L2=pl
>>>>>> pfwd=g pbwd=g smooth=0 sample=1000 workers=1
>>>>>>
>>>>>> this generates the error:
>>>>>> Feature function PT0 specified 9 dense scores or weights. Actually
>>>>>> has 0.
>>>>>>
>>>>>> which is solved when num-features is changed to '0'
>>>>>> but generates the error below:
>>>>>>
>>>>>> Exception: moses/TranslationModel/UG/mmsapt.cpp:381 in virtual void
>>>>>> Moses::Mmsapt::Load() threw util::Exception because
>>>>>> `this->m_feature_names.size() != this->m_numScoreComponents'.
>>>>>> At moses/TranslationModel/UG/mmsapt.cpp:381: number of feature values
>>>>>> provided by Phrase table (7) does not match number specified in Moses
>>>>>> config file (0)!
>>>>>> Changing it to 7 also does not help.
>>>>>>
>>>>>> I have tried with
>>>>>> Mmsapt name=PT0 output-factor=0 num-features=0
>>>>>> base=/home/sandipan/inc_retrain/MT_sys/EnPl/mtdata_pro/train. L1=en L2=pl
>>>>>>
>>>>>> but does not work.
>>>>>> What I need to do at this stage of retraining using moses?
>>>>>>
>>>>>> Thanks and regards,
>>>>>> sandipan
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Moses-support mailing listMoses-support@mit.eduhttp://mailman.mit.edu/mailman/listinfo/moses-support
>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Moses-support mailing list
>>>>>> Moses-support@mit.edu
>>>>>> 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
>>>> Moses-support@mit.edu
>>>> http://mailman.mit.edu/mailman/listinfo/moses-support
>>>>
>>>>
>>>
>>
>>
>> --
>> Hieu Hoang
>> Research Associate
>> University of Edinburgh
>> http://www.hoang.co.uk/hieu
>>
>>
>
>
> --
> Hieu Hoang
> Research Associate
> University of Edinburgh
> http://www.hoang.co.uk/hieu
>
>
> _______________________________________________
> Moses-support mailing list
> Moses-support@mit.edu
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20140908/e6d48f72/attachment-0001.htm

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

Message: 3
Date: Mon, 8 Sep 2014 12:16:41 +0000
From: "Marciano, Jay" <Jay.Marciano@lionbridge.com>
Subject: [Moses-support] AMTA 2014: Call for Participation
To: "moses-support@mit.edu" <moses-support@mit.edu>
Message-ID:
<8BAE8990EA732742A89E8405052E55310129A83B8C@BIL-EXC11-06.corpnet.liox.org>

Content-Type: text/plain; charset="us-ascii"

AMTA 2014: Call for participation

Join us in Vancouver for the latest on MT research, development, and use!
October 22-26, 2014, Renaissance Vancouver Harbourside Hotel, Vancouver, BC, Canada.

Things are moving fast in the world of machine translation. Academic research continues to push the boundaries of what is possible with the technology, and its widespread adoption in large-scale translation projects and real-time use cases is driving language service providers, government agencies, and many large corporations to develop innovative approaches to applying MT, improving output quality, and measuring the results.

The AMTA Conference is your opportunity to hear about the most current developments and trends in Machine Translation and to meet and share your ideas with people who are deeply involved in and passionate about MT.
The 11th biennial conference of the Association for Machine Translation in the Americas will be held from October 22-26 at the Renaissance Vancouver Harbourside Hotel in Vancouver, BC. The main conference, October 23-25, is organized into three tracks (Research, Commercial, and Government), each of which has a full program of presentations that will highlight the state of the art and practice of MT. You can concentrate on one track or pick and choose among the sessions of greatest interest to you. In addition to the main conference, a series of three-hour tutorials and full-day workshops will be presented on Wednesday (October 22) and Sunday (October 26).

For more information on the conference and venue, please visit the conference website at http://amta2014.amtaweb.org/

Registration for AMTA 2014 is now open. Register by September 28 to take advantage of substantial early bird savings!

http://amta2014.amtaweb.org/Registration.aspx

See you in Vancouver!

******

Jay Marciano
Publicity Director, AMTA
Director, Real-Time Translation Development
Lionbridge Technologies, Inc




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

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


End of Moses-support Digest, Vol 95, Issue 13
*********************************************

Tweet