Re: [Moses-support] HOWTO add a new feature function to Moses

function. As it seemed to work for me, I'm fairly certain that my list is (was) exhaustive :-) Thanks, Jason 2008/4/21 Jason Katz-Brown jaso...@mit.edu Hi all, After snuffing out what I think is all of the steps required to add a new feature function to the Moses decoder and training system

[Moses-support] How to cite Background page on moses website

Hi, I would like to cite in my thesis the excellent stat MT Background page on the Moses website: http://www.statmt.org/moses/?n=Moses.Background Instead of listing the URL, is there a paper that text appeared in? Thanks, Jason ___ Moses-support

Re: [Moses-support] Seemingly buggy behavior when specifying

look like it. I unzipped the table and grep 0\.00 reordering-table.msd-bidirectional-fe.0.5.0-0 doesn't print anything :-/ Thanks, Jason PS. I'm sorry about the half missing email title! 2008/8/7 Jason Katz-Brown [EMAIL PROTECTED]: Hi again all, I want to compare translation results

[Moses-support] What is lexical weighting lex(f|e) in the phrase tables?

Hi, I am trying to take an inventory of all the feature functions used by a baseline Moses system, but one has tripped me up. http://www.statmt.org/moses/?n=FactoredTraining.ScorePhrases mentions phrase translation score lexical weighting lex(f|e) (or lex(e|f)). Could somebody explain what this

Re: [Moses-support] What is lexical weighting lex(f|e) in the phrase tables?

2008/8/6 Jason Katz-Brown [EMAIL PROTECTED] Hi, I am trying to take an inventory of all the feature functions used by a baseline Moses system, but one has tripped me up. http://www.statmt.org/moses/?n=FactoredTraining.ScorePhrases mentions phrase translation score lexical weighting lex(f

[Moses-support] Seemingly buggy behavior when specifying

Hi again all, I want to compare translation results with my tuned Japanese-English system, and the same system with the lexical reordering feature weights set to zero. But in doing so, I am seeing confusing behavior and -inf lexical reordering feature values. To perform my experiment, I passed

[Moses-support] mert-moses with NIST score optimization

(I sent this mail once before, but it bounced because one of MIT's mailman servers filled a disk. So I hope this is not a duplicate.) Hi all, Has anybody enhanced mert-moses and cmert to allow one to tune for maximum NIST score, or to maximize some other metric? If not, I am interested in adding

[Moses-support] HOWTO add a new feature function to Moses

Hi all, After snuffing out what I think is all of the steps required to add a new feature function to the Moses decoder and training system, I made a list of them. Perhaps it is helpful for others. Moses could certainly use more score producers to increase the count from anemic 5 :-) --Jason

[Moses-support] Learning weight for a new feature in moses MERT training

Hi, I added a new score producer that produces one score, and a corresponding parameter called weight-r, to my moses binary. I made a unit test, and I get correct scoring and translation behavior if I run my test with these two excerpted versions of the ini file: [weight-r] 0 or [weight-r] -99

[Moses-support] GIZA++ nan errors

Hi all, Sometimes during the GIZA++ step of the moses training script, I see many nan pop up with big ERROR: labels in front of them. I have appended an excerpt of some typical error-ful output. Has anybody seen errors like ERROR: nan nan nan 22 7 during the GIZA++ step? It might be that they

Re: [Moses-support] Difficulty setting up a factored reordering model

Breen. With this much data and the other problems inherent in translating Japanese, the translation results are merely okay :-) Thanks, Jason On Jan 21, 2008 2:53 AM, Jason Katz-Brown [EMAIL PROTECTED] wrote: Hi all, I am working on improving a baseline Japanese-English moses system I

[Moses-support] Difficulty setting up a factored reordering model

Hi all, I am working on improving a baseline Japanese-English moses system I developed. It achieves okay translation results already. I am experimenting with adding part-of-speech factors, so I prepared training data that looks like: word0|pos0 word1|pos1 word2|pos2 ... I tried two moses