Any way through which MT System developed through MOSES Tool Kit can be
integrated with some existing system. Giving u an example, We are going to
release one Indic Word Processor having various language tools using
Microsoft Office Framework. In this framework, we have integrated various
language
Hello Moses team,
I am running a phrase based En-Ar SMT system, in the tuning step it stops
after run1 with this line in the end of mert.out:
Line empty in nbestlist 'run1.best100.out' at
/home/sara/mymoses/scripts/training/mert-moses.pl line 1481.
Any help?
the command I used:
nohup nice /hom
Hi Moses developers,
Can we generally assume that parameters provided by a command line switch overwrite the corresponding parameters in the moses.ini?
For the stack parameter, the output (when initializing the decoder) is a bit confusing. Let's say we set
[s]
1500
in the moses.in
Thank you, Philipp,
That’s a very useful answer. My audience (the users I have in mind) consists of
companies involved with localization who are looking at implementing Moses into
their workflow. At the level I am writing for, they need to understand what
changes are likely to matter to them in
hi miriam
Hieu Hoang
http://www.hoang.co.uk/hieu
On 31 August 2016 at 12:50, "Miriam Käshammer"
wrote:
> Hi Moses developers,
>
> Can we generally assume that parameters provided by a command line switch
> overwrite the corresponding parameters in the moses.ini?
>
yes, if you see it acting any
Hi,
on that level:
Video documentation / guides to Moses
http://www.statmt.org/moses/?n=Moses.Video
Windows integration:
http://www.slate.rocks/
CASMACAT Home Edition (training infrastructure / CAT tool)
http://casmacat.eu/index.php?n=UserGuide.HomeEdition
Features:
- incremental updating of p
Hi,
you should run the Moses command by hand
(the exact way it was run can be found in the
log files) and go from there.
-phi
On Wed, Aug 31, 2016 at 7:43 AM, Sara Ebrahim
wrote:
> Hello Moses team,
>
> I am running a phrase based En-Ar SMT system, in the tuning step it stops
> after run1 with
Hi,
the easiest to do this is to run Moses as a server process
and talk to Moses via the XMLPRC interface
http://www.statmt.org/moses/?n=Advanced.Moses
You can also compile Moses as library into your code,
but that gets much more complicated and also requires
you to deal with things like tokeniza
Hi,
there are no guarantees that more data will lead to better performance.
It typically does, but the type of data also matters, and there is a
significant amount of variance so that any single alignment decision
may be best handled under any data condition / setting.
-phi
On Thu, Aug 25, 2016
Hi,
can you be a bit more specific what you are trying to do?
The current tuning setup (which can also be used for
reranking) does include an implementation of the MIRA
algorithm which is a variant of perceptron training.
-phi
On Tue, Aug 23, 2016 at 1:57 AM, Selva Nalladurai
wrote:
> Hello g
10 matches
Mail list logo