hi thomas,
I've written something on the website about syntax models. Also figured
out why 'inputtype=3' is need even for hiero models.
The new docs is at the bottom of the page
http://www.statmt.org/moses/?n=Moses.SyntaxTutorial
Section: Different Kinds of Syntax Models
It doc is still pretty rough at the moment, if there's anything that's
confusing, please PM me and i'll try & sort it out
On 30/04/2012 11:19, Hieu Hoang wrote:
As far as I know, the inputtype '0' should work with pb and hiero
models. I'll take a look at why it doesn't in the next few weeks.
The inputtype '3' is a tree input, ie. the sentence you give to the
decoder is a parse tree, formatted using XML
http://www.statmt.org/moses/?n=Moses.SyntaxTutorial
This is used for tree-to-string/tree-to-tree models.
However, if you don't mark the input wit any XML, it is still a valid
input, it just means the input sentence just consist of terminals and
no constituent span labels.
i'll try and make this a bit clearer in the next few months.
On 30/04/2012 10:52, Thomas Schoenemann wrote:
Hi all,
I would greatly appreciate an explanation regarding this input type
stuff. The latest mert-moses.pl only explains input types 0-2, which
are plain text, confusion networks and lattices. Where is the
difference between confusion networks and lattices? What does type 3
mean? Why should the input for hierarchical models be different from
the input for phrase-based models? Clearly for hierarchical models no
parse trees are input?
Thanks much!
Thomas
P.S: I think MOSES would be easier to use if we switch from numbers
to text, e.g. "plain" instead of "0".
------------------------------------------------------------------------
*Von:* John Morgan <johnjosephmor...@gmail.com>
*An:* Moses-support <Moses-support@mit.edu>
*Gesendet:* 14:20 Donnerstag, 26.April 2012
*Betreff:* [Moses-support] default inputtype changed?
Hello,
My system builds with ems that worked last week are failing now at the
tuning step.
If I set
decoder-settings = "-inputtype 3"
under the TUNING stanza in my experiment.perl configuration file
tuning runs again.
Did something change?
Thanks,
--
Regards,
John J Morgan
_______________________________________________
Moses-support mailing list
Moses-support@mit.edu <mailto:Moses-support@mit.edu>
http://mailman.mit.edu/mailman/listinfo/moses-support
_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support
_______________________________________________
Moses-support mailing list
Moses-support@mit.edu
http://mailman.mit.edu/mailman/listinfo/moses-support
