Hi everybody,

Thank you so much for your replies to my question. Unfortunately, I didn't
have time to test the modified script. I hope to test it in the next few
days and will come back with my answer asap.

Regards
Konstantinos


2013/12/5 Thomas Meyer <[email protected]>

>  Hi all,
>
> Merged and tested. All works for me. I attach the latest github
> moses-parallel.pl wherein my changes are commented out and to be found
> under #SGE-make-it-work:
>
> Hope it helps,
> Thomas
>
> On 05/12/13 12:05, Hieu Hoang wrote:
>
>  The script seems to have been forked just before the last changes
>
> https://github.com/moses-smt/mosesdecoder/commit/93433cf0157ecadd19162929a077aa6179dc0d8b
>
> If someone who uses SGE can do a merge with the latest version in github
> and test that it works, I can check it in
>
>
>
> On 2 December 2013 09:03, Thomas Meyer <[email protected]> wrote:
>
>>  Hi all,
>>
>> Attached is moses-parallel.pl with some modifications (marked #thm:) to
>> make it work with the SGE architecture we have here. I just ran a test
>> whether it outputs the word alignment and all works normally.
>>
>> Here is the command I used:
>>
>> ./moses-parallel.pl -jobs 2 -alignment-output-file alignment.txt
>> -input-file /temp/tmeyer/test.en -decoder
>> /temp/tmeyer/mosesdecoder/bin/moses -config
>> /temp/tmeyer/en_de/test/moses.tuned.ini > test.txt
>>
>> which outputs, in alignment.txt:
>>
>> 0-0 1-1 2-2 3-3 4-4 5-5 6-6 7-7 8-8 9-9 10-10 11-11 12-12 13-13 14-14
>> 15-15 16-16 17-17 18-18 19-19 20-20 21-21 22-22 23-23 24-24
>> 0-0 1-1 2-2 3-3 4-4 5-4 6-5 7-6 8-7 9-8 10-9 11-10 12-11 13-12 13-13
>> 14-14 16-15 17-16 18-17 20-18 21-19 22-20 23-21 24-22 25-23 26-24 27-25
>> 28-26 29-27 30-28 31-28 32-29 33-30 34-31 36-32 37-33 38-34 39-35
>>
>> etc.
>>
>> Hope it will work for you, too, Konstantinos.
>>
>> Best,
>> Thomas
>>
>>
>>
>> On 29/11/13 21:59, Philipp Koehn wrote:
>>
>> Hi Thomas,
>>
>> in that case - can you share your modifications if they are of general
>> interest and check if there is a problem with the word alignment reporting?
>>
>> -phi
>>
>> On Fri, Nov 29, 2013 at 6:52 PM, Thomas Meyer <[email protected]> 
>> <[email protected]> wrote:
>>
>>  Hi all,
>>
>> I 'still' do and find it quite convenient. At least with release 1.0 and
>> some modifications to moses-parallel.pl
>>
>> Thomas
>>
>>
>> From: CHATZITHEODOROU Konstantinos
>> Sent: Friday, 29 November 2013 7:48 PM
>> To: Hieu Hoang
>> Cc: [email protected]
>> Subject: Re: [Moses-support] Word-to-word alignment when moses-parallel.pl
>> is used
>>
>> Hi Hieu,
>>
>> Thanks for your reply. I have a question. Why not many people still use SGE
>> to run Moses?
>>
>> Thank you in advance.
>>
>> Best,
>> Konstantinos
>>
>>
>> 2013/11/29 Hieu Hoang <[email protected]> <[email protected]>
>>
>>  I'm not sure if many people still uses Sun Grid Engine to run Moses. If
>> not, scripts like moses-parallel.pl may become outdated.
>>
>> I know at Edinburgh, we no longer use SGE. Instead we run our experiment
>> on 1 large multi-core machine. Most of moses has been re-engineered to be
>> multithreaded to take advantage of multicores
>>
>>
>> On 28 November 2013 10:25, CHATZITHEODOROU Konstantinos<[email protected]> 
>> <[email protected]> wrote:
>>
>>  Hi Philipp,
>>
>> Thank you your your prompt reply. I have tried the option
>> "alignment-output-file" for moses-parallel.pl but didn't use it as part of
>> the decoder setting. But it didn't work. Each individual sub-job finished
>> with an error status stating something the sub-job is untranslated.
>>
>> Thank you for your time and consideration.
>>
>> Regards,
>> Konstantinos
>>
>>
>>
>> 2013/11/27 Philipp Koehn <[email protected]> <[email protected]>
>>
>>  Hi,
>>
>> the script has an option "alignment-output-file" - did you try this?
>>
>> -phi
>>
>> On Tue, Nov 26, 2013 at 8:13 PM, CHATZITHEODOROU 
>> Konstantinos<[email protected]> <[email protected]> wrote:
>>
>>  Hi all,
>>
>> Could you please give an example on how to output the word-to-word
>> alignment
>> during decoding when moses-parallel.pl is used?
>>
>> Thank you in advance for your help.
>>
>> Regards,
>> Konstantinos
>>
>>
>> _______________________________________________
>> Moses-support mailing 
>> [email protected]http://mailman.mit.edu/mailman/listinfo/moses-support
>>
>>
>> --
>> CHATZITHEODOROU Konstantinos
>>
>>
>> _______________________________________________
>> Moses-support mailing 
>> [email protected]http://mailman.mit.edu/mailman/listinfo/moses-support
>>
>>  --
>> Hieu Hoang
>> Research Associate
>> University of Edinburghhttp://www.hoang.co.uk/hieu
>>
>>  --
>> CHATZITHEODOROU Konstantinos
>>
>>
>>
>> _______________________________________________
>> Moses-support mailing 
>> [email protected]http://mailman.mit.edu/mailman/listinfo/moses-support
>>
>> _______________________________________________
>> Moses-support mailing 
>> [email protected]http://mailman.mit.edu/mailman/listinfo/moses-support
>>
>>
>>
>
>
> --
> Hieu Hoang
> Research Associate
> University of Edinburgh
> http://www.hoang.co.uk/hieu
>
>


-- 
CHATZITHEODOROU Konstantinos
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to