Moses-support Digest, Vol 126, Issue 22

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: probingPT with lexical reordering (Hieu Hoang)
2. Re: Tuning for PT2? (Hieu Hoang)
3. Re: "-ttable-limit" (Hieu Hoang)


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

Message: 1
Date: Sun, 16 Apr 2017 19:25:17 +0100
From: Hieu Hoang <hieuhoang@gmail.com>
Subject: Re: [Moses-support] probingPT with lexical reordering
To: Jorg Tiedemann <tiedeman@gmail.com>
Cc: moses-support <moses-support@mit.edu>
Message-ID:
<CAEKMkbgj+=f_-_JJemenFovFhiSf3S+8G__py3Bqra4Mn4GkjQ@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

This tells you a bit more about it:
http://www.statmt.org/moses/?n=Site.Moses2#ntoc10
Or look inside the script
scripts/generic/binarize4moses2.perl


* Looking for MT/NLP opportunities *
Hieu Hoang
http://moses-smt.org/


On 16 April 2017 at 10:28, Jorg Tiedemann <tiedeman@gmail.com> wrote:

> Hi,
>
> I was wondering how I should use probingPT with integrated lexical
> reordering tables (how to create such a table and how to use it in
> decoding).
> I saw that it only works with moses2 - is that so?
>
> Thanks!
> J?rg
>
>
>
>
>
>
>
> _______________________________________________
> 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/20170416/3e5d7661/attachment-0001.html

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

Message: 2
Date: Sun, 16 Apr 2017 19:35:44 +0100
From: Hieu Hoang <hieuhoang@gmail.com>
Subject: Re: [Moses-support] Tuning for PT2?
To: Mike Ladwig <mdladwig@gmail.com>
Cc: moses-support <moses-support@mit.edu>
Message-ID:
<CAEKMkbg4OSfdyKY=c3etk=kda=Knyp1B_vFKAuKoUkVfhEFDWg@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

For our paper
https://arxiv.org/abs/1610.01108
I also ran comparison between Moses, Moses2 and NMT:


zh-en en-zh ru-en en-ru fr-en en-fr es-en en-es
pb Moses 43.21 37.51 52.95 43.22 52.11 49.73 59.47 60.26

Moses2 43.15 37.56 52.92 43.14 51.86 49.74 59.74 60.24










hiero Moses 47.26 41.65 53.56 43.24 52.97 49.33 60.65 60.4

Moses2 47.5 41.92 53.3 43.38 52.78 49.74 60.58 60.36










NMT
50.11 44.94 51.25 43.37 50.8 49.78 59.9 60.34
Basically, there's no consistent difference between Moses and Moses2.

I didn't use the integrate probing pt/re-ordering model, rather I used
separate reordering model. However, I don't think it would make any
difference


* Looking for MT/NLP opportunities *
Hieu Hoang
http://moses-smt.org/


On 15 April 2017 at 18:08, Mike Ladwig <mdladwig@gmail.com> wrote:

> The moses2 phrase/reordering datastructure created with mosesdecoder/bin/
> CreateProbingPT2
>
> On Sat, Apr 15, 2017 at 12:57 PM, Hieu Hoang <hieuhoang@gmail.com> wrote:
>
>> Whats pt2?
>>
>> On 15 Apr 2017 1:22 p.m., "Mike Ladwig" <mdladwig@gmail.com> wrote:
>>
>>> I've been looking, without success, for any experiences tuning PT2 with
>>> the various controls (num-lex-scores, prune, num-scores, etc.)
>>>
>>> Can anyone share size-speed-quality experiences or suggest good starting
>>> points?
>>>
>>> Thanks
>>>
>>> _______________________________________________
>>> 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/20170416/dc265501/attachment-0001.html

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

Message: 3
Date: Sun, 16 Apr 2017 20:09:05 +0100
From: Hieu Hoang <hieuhoang@gmail.com>
Subject: Re: [Moses-support] "-ttable-limit"
To: Gary <garyhess999@hotmail.com>
Cc: "moses-support@mit.edu" <moses-support@mit.edu>
Message-ID:
<CAEKMkbhC_=7Cy_bQW3zJMv4mo4DESafUWHs2boE5dM6kpqUB+Q@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

ah, didn't know that parameter had been deleted. It's now a more
long-winded arg:
-feature-overwrite "pt table-limit=999"
Also, table-limit=0 doesn't mean no limit, it prune all translation rules!
So set it to a large number to get the same behaviour as before

* Looking for MT/NLP opportunities *
Hieu Hoang
http://moses-smt.org/


On 15 April 2017 at 11:04, Gary <garyhess999@hotmail.com> wrote:

> Hi --
>
>
> I am getting started with Moses and working through the Moses phrased
> based tutorial (http://www.statmt.org/moses/?n=Moses.Tutorial). There is
> the following command which doesn't work because "-ttable-limit" is now
> deprecated:
>
>
> echo 'das ist ein kleines haus' | moses -f phrase-model/moses.ini
> -ttable-limit 0 -v 2
>
>
> Is there a command-line equivalent for "-ttable-limit 0", or would I have
> to do that in moses.ini?
>
>
> Thanks,
>
> Gary Hess
>
>
>
> _______________________________________________
> 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/20170416/013bf369/attachment.html

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

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


End of Moses-support Digest, Vol 126, Issue 22
**********************************************

Tweet