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. Re: problem in generating phrase table after tuning (Apurva Joshi)
2. SMT job opening at the JRC: good candidates should apply now
(Ralf Steinberger)
----------------------------------------------------------------------
Message: 1
Date: Wed, 18 Nov 2015 10:47:53 +0530
From: Apurva Joshi <apurvajoshi1992@gmail.com>
Subject: Re: [Moses-support] problem in generating phrase table after
tuning
To: Philipp Koehn <phi@jhu.edu>, moses-support@mit.edu
Message-ID:
<CAMfCXbXfypPMyzgXMg1zrjQoFZzirhy0t7kYCcU_vTDj7r_sfw@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"
thank you philipp koehn for reply, i got solution..thank you once again..
On Wed, Nov 18, 2015 at 3:13 AM, Philipp Koehn <phi@jhu.edu> wrote:
> Hi,
>
> the standard pipeline filters the phrase translation table
> for tuning and evaluation.
>
> The result of tuning, however, is just the feature weights.
> None of the model files should be changed.
>
> Specifically, you should not just use the moses.ini file
> in the tuning directory, since this is uses a phrase table
> filtered for the tuning set.
>
> In the standard experiment.perl pipeline, the script
> $MOSES/scripts/ems/support/substitute-weights.perl
> substitutes weights in the original moses.ini with
> the tuned weights.
>
> -phi
>
>
> On Tue, Nov 17, 2015 at 2:17 AM, Apurva Joshi <apurvajoshi1992@gmail.com>
> wrote:
>
>> hello ,
>>
>> initially i trained with corpora containing 7 lac sentnses...i trained
>> it..and its giving good translation...
>>
>> then for tuning i took new small size corpora containing 2k lines ....
>>
>> now i run tuning with this new processed corpora as input along with my
>> moses.ini (generated after training) as input ....then mert-work folder is
>> generated with run6.moses.ini ....
>>
>> along with this ,phrase table.gz and reordering table is also generated
>> in mert-work folder...
>>
>> now problem is ....
>>
>> 1st problem :
>>
>> before tuning , size of phrase table is 2.5Gb
>>
>> after tuning , size of phrase table is 256.5Mb
>>
>> 2nd problem:
>>
>> now while running moses, should i use moses.ini(in mert-work/filtered
>> folder) OR moses.ini(run6.moses.ini) ...?
>>
>> i ran moses using both above mentioned moses.ini....
>> from both m not getting proper o/p...
>> eg
>>
>> ip = hello
>>
>> op =
>>
>> *BEST TRANSLATION: hello|UNK|UNK|UNK *
>> before tuning , it was giving me correct translation of hello i.e
>>
>> op= ?????? (hindi translation of hello)....
>>
>>
>> so whether is it taking memory problem or what ?? plzz help...
>>
>>
>>
>>
>> _______________________________________________
>> 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/20151118/5b4576fb/attachment-0001.html
------------------------------
Message: 2
Date: Wed, 18 Nov 2015 08:57:11 +0100
From: Ralf Steinberger <ralf.steinberger@jrc.ec.europa.eu>
Subject: [Moses-support] SMT job opening at the JRC: good candidates
should apply now
To: moses-support@mit.edu
Message-ID: <058101d121d6$bb682e60$32388b20$@jrc.ec.europa.eu>
Content-Type: text/plain; charset="us-ascii"
Dear members of this list,
There is an open call for interest to work at the European Commission's
Joint Research Centre (JRC), one of which for a person having experience
with Moses and being able to install, train, maintain and tune the current
SMT system (developed with Moses). Java programming skills are needed for an
integration of the SMT system with the JRC's media monitoring processing
chain.
As the most promising candidates for at least two Language Technology
positions in Ispra (Italy) will be short-listed this or next week, please
submit your application NOW via the following link. Apologies for the late
notification!
Please specify your experience with Moses, your programming skills
(especially in Java) and your foreign language skills.
https://ec.europa.eu/jrc/en/working-with-us/jobs/vacancies/function-group-iv
-researchers .
The new positions as 'contract agents' can only be filled by persons who
have submitted their profile and CV via this central interface. The
procedure is described in detail in the following document:
https://ec.europa.eu/jrc/sites/default/files/jrc-research-fgiv-contract-staf
f-2015-selection_en.pdf .
The maximum duration for contract agents is six years (consisting of several
shorter, renewable contracts). For a list of FAQs (salary, eligibility,
etc.), see:
https://ec.europa.eu/jrc/sites/default/files/jrc-research-fgiv-contract-staf
f-2015-selection-faq-en.pdf .
The following paper provides information on the status of the JRC's SMT
system:
Turchi Marco, Martin Atkinson, Alastair Wilcox, Brett Crawley, Stefano
Bucci, Ralf Steinberger & Erik van der Goot (2012). ONTS: "OPTIMA" News
Translation System. Proceedings of the 13th Conference of the European
Chapter of the Association for Computational Linguistics (EACL), pp. 25-30,
Avignon, France, April 23 - 27 2012. (PDF
<http://aclweb.org/anthology-new/E/E12/E12-2.pdf#page=35> )
General subject areas: The successful candidates to work in the Europe Media
Monitor <http://emm.newsbrief.eu/overview.html> (EMM) team will work in one
or more of the following fields:
- Media monitoring (conventional and social media);
- Multilingual text mining;
- Cross-lingual applications;
- Information extraction;
- Trend monitoring;
- Statistical Machine Translation (SMT).
Useful knowledge and skills to work in the EMM team include:
- Good programming skills, especially in Java;
- Good working knowledge of English;
- Knowledge of at least one (better more) further languages;
- Experience with developing working systems for real-life users.
You will work as part of the <http://emm.newsbrief.eu/overview.html> Europe
Media Monitor (EMM) team of the JRC, which developed the highly multilingual
(20 to 70 languages) news analysis systems <http://emm.newsexplorer.eu/>
NewsExplorer, <http://emm.newsbrief.eu/> NewsBrief and the
<http://medisys.newsbrief.eu/> Medical Information System MedISys (Public
Health monitoring). The EMM applications integrate JRC-developed software
tools for event scenario template filling; named entity recognition and
disambiguation; name variant matching; quotation recognition; summarisation;
machine translation; sentiment analysis (opinion mining); topic detection
and tracking; document clustering and classification; trend detection and
visualisation; cross-lingual information aggregation; early warning systems;
social network analysis; personalised news monitoring; apps for phones and
tablets and more. The team also made available the
<https://ec.europa.eu/jrc/en/language-technologies> freely downloadable
multilingual parallel corpora and other text resources such as JRC-Acquis,
DGT-Translation Memory (and other TMs), JRC-Names, JRC Eurovoc Indexer, and
more. For a list of our
<http://langtech.jrc.ec.europa.eu/JRC_Publications.html> publications, go to
<http://langtech.jrc.ec.europa.eu/JRC_Publications.html>
http://langtech.jrc.ec.europa.eu/JRC_Publications.html.
===========================================
<https://ec.europa.eu/jrc/en/person/ralf-steinberger> Ralf Steinberger
European Commission - Joint Research Centre (JRC)
URL - Applications: <http://emm.newsbrief.eu/overview.html>
http://emm.newsbrief.eu/overview.html
URL - Context:
<https://ec.europa.eu/jrc/en/research-topic/internet-surveillance-systems>
https://ec.europa.eu/jrc/en/research-topic/internet-surveillance-systems
21027 Ispra (VA), Italy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20151118/739586ad/attachment.html
------------------------------
_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support
End of Moses-support Digest, Vol 109, Issue 38
**********************************************
Subscribe to:
Post Comments (Atom)
0 Response to "Moses-support Digest, Vol 109, Issue 38"
Post a Comment