Moses-support Digest, Vol 121, Issue 35

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. 3-5 year postdoc/fellowship in fast neural machine
translation (Kenneth Heafield)
2. Output Loglinear score with n-best outputs (Nat Gillin)
3. Re: Output Loglinear score with n-best outputs (Nat Gillin)
4. Re: Output Loglinear score with n-best outputs (Vito Mandorino)


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

Message: 1
Date: Tue, 22 Nov 2016 23:04:24 +0000
From: Kenneth Heafield <moses@kheafield.com>
Subject: [Moses-support] 3-5 year postdoc/fellowship in fast neural
machine translation
To: "moses-support@mit.edu" <moses-support@mit.edu>
Message-ID: <c9b625d0-5df2-9eb7-d1a7-161fc1ee2485@kheafield.com>
Content-Type: text/plain; charset=utf-8

Dear Moses,

The Alan Turing Institute, a joint venture of five universities,
including the University of Edinburgh, is recruiting research fellows
(~postdocs): https://www.turing.ac.uk/opportunities/ . These last 3-5
years. The position is in London or possibly Edinburgh depending on
your preference.

Intel is a partner of the Alan Turing Institute and particularly
interested in accelerating neural networks, including neural machine
translation.

The key requirements are:

- a PhD or equivalent qualification in a relevant subject at the point
of taking up the Fellowship.

- an excellent publication record and/or demonstrated excellent research
potential such as via preprints.

If you work on neural machine translation or language modeling and want
to make it train or translate faster, please contact me before applying.

Kenneth


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

Message: 2
Date: Wed, 23 Nov 2016 16:41:28 +0800
From: Nat Gillin <nat.gillin@gmail.com>
Subject: [Moses-support] Output Loglinear score with n-best outputs
To: moses-support@mit.edu
Message-ID:
<CAD2EOZhVkXXr-h+LeGkk2SwTn-j8fHxnPGs+N4hK75fCaGy5Ow@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Dear Moses community,

When decoding, is there a way to output the loglinear score of each
hypothesis together with the n-best output?

Regards,
Nat
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20161123/aa566176/attachment-0001.html

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

Message: 3
Date: Wed, 23 Nov 2016 17:06:56 +0800
From: Nat Gillin <nat.gillin@gmail.com>
Subject: Re: [Moses-support] Output Loglinear score with n-best
outputs
To: moses-support@mit.edu
Message-ID:
<CAD2EOZhcZ8eWJwFpbpFzxC=W0SNS7ojOj+F0MBo9XpavFFRJfg@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Dear Moses community,

I found it! It's the last column in the -n-best-list output:

0 ||| one member for kara will be selected . ||| LexicalReordering0=
-2.12026 0 0 -1.60944 0 0 Distortion0= 0 LM0= -20.2445 WordPenalty0= -8
PhrasePenalty0= 2 TranslationModel0= -2.42529 -36.1231 -1.039 -16.2225 |||
-3.80866

I realized that the hypothesis from the n-best list isn't unique. Is there
a way to get 3 unique n-best hypothesis, instead of the first 3 hypothesis
(non-unique)?

Regards,
Nat


On Wed, Nov 23, 2016 at 4:41 PM, Nat Gillin <nat.gillin@gmail.com> wrote:

> Dear Moses community,
>
> When decoding, is there a way to output the loglinear score of each
> hypothesis together with the n-best output?
>
> Regards,
> Nat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20161123/ceb9bcf0/attachment-0001.html

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

Message: 4
Date: Wed, 23 Nov 2016 11:05:45 +0100
From: Vito Mandorino <vito.mandorino@linguacustodia.com>
Subject: Re: [Moses-support] Output Loglinear score with n-best
outputs
To: Nat Gillin <nat.gillin@gmail.com>
Cc: moses-support <moses-support@mit.edu>
Message-ID:
<CA+8mSmFxSMuih6rvS-Jp6GwViuZ7m1h4Uj9qEazRKi1dJRKTgQ@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi Nat,

this is obtained by adding 'distinct' in the command i.e.

./moses -f moses.ini -n-best-list filename 10 distinct

Vito

2016-11-23 10:06 GMT+01:00 Nat Gillin <nat.gillin@gmail.com>:

> Dear Moses community,
>
> I found it! It's the last column in the -n-best-list output:
>
> 0 ||| one member for kara will be selected . ||| LexicalReordering0=
> -2.12026 0 0 -1.60944 0 0 Distortion0= 0 LM0= -20.2445 WordPenalty0= -8
> PhrasePenalty0= 2 TranslationModel0= -2.42529 -36.1231 -1.039 -16.2225 |||
> -3.80866
>
> I realized that the hypothesis from the n-best list isn't unique. Is there
> a way to get 3 unique n-best hypothesis, instead of the first 3 hypothesis
> (non-unique)?
>
> Regards,
> Nat
>
>
> On Wed, Nov 23, 2016 at 4:41 PM, Nat Gillin <nat.gillin@gmail.com> wrote:
>
>> Dear Moses community,
>>
>> When decoding, is there a way to output the loglinear score of each
>> hypothesis together with the n-best output?
>>
>> Regards,
>> Nat
>>
>
>
> _______________________________________________
> Moses-support mailing list
> Moses-support@mit.edu
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>


--
*M**. Vito MANDORINO -- Chief Scientist*


[image: Description : Description : lingua_custodia_final full logo]

*The Translation Trustee*

*1, Place Charles de Gaulle, **78180 Montigny-le-Bretonneux*

*Tel : +33 1 30 44 04 23 Mobile : +33 6 84 65 68 89*

*Email :* *vito.mandorino@linguacustodia.com
<massinissa.ahmim@linguacustodia.com>*

*Website :*
*www.linguacustodia.finance <http://www.linguacustodia.com/>*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/mailman/private/moses-support/attachments/20161123/9cd6f990/attachment.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.jpg
Type: image/jpeg
Size: 4421 bytes
Desc: not available
Url : http://mailman.mit.edu/mailman/private/moses-support/attachments/20161123/9cd6f990/attachment.jpg

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

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


End of Moses-support Digest, Vol 121, Issue 35
**********************************************

Tweet