Hi,

yes, that's right. I fixed the page.

-phi

On Jan 21, 2008 12:42 PM, Loïc Dugast <[EMAIL PROTECTED]> wrote:
> Hello
>
> Am I wrong or the description of the phrase table format below is wrong ?
> http://www.statmt.org/moses/?n=FactoredTraining.ScorePhrase
>
> It seems to me the weights are rather : p(f/e) lex(f/e) p(e/f) lex(e/f)
> phrase_penalty
> ..where "f" is the string in the first column , "e" the one in the second.
>
> See for example:
>
> consistant pour ||| consistent to ||| (0) (1) ||| (0) (1) ||| 0.25 0.0419339
> 1 0.0295274 2.718
> consistant à ||| consistent to ||| (0) (1) ||| (0) (1) ||| 0.5 0.127117
> 0.00699301 0.039091 2.718
> consistante à ||| consistent to ||| (0) (1) ||| (0) (1) ||| 0.25 0.0799019 1
> 0.379741 2.718
>
> Thanks
>
> Loic
>
>
> _______________________________________________
> Moses-support mailing list
> [email protected]
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>

_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to