Yes, the output log is reporting many zero length sentences. I must have something misconfigured up stream.
On Sun, Sep 13, 2009 at 11:03 PM, Chris Dyer <[email protected]> wrote: > Is it possible that you have a sentence of length zero? size2 is, I > believe, one of the dimensions of the trellis, which in one direction > is the source sentence length and in the other is the target sentence > length. > > On Mon, Sep 14, 2009 at 1:28 AM, John Kolen <[email protected]> wrote: > > I'm trying to run the example from the moses step-by-step guide. I have a > > problem with GIZA++, where it generates a segv int > ForwardBackwardTraining. > > The problems is that the HMMNetwork it is passed has a size2 of zero. The > > code clearly assumes that size2() must be positive, as it is a dimension > of > > a rectangular array. Unfortunately, I can't figure where the zero > dimension > > is creeping in. > > > > Here's what it is executing: > > > > GIZA++ -CoocurrenceFile demo-en-fr/giza.fr-en/fr-en.cooc -c > > demo-en-fr/corpus/fr-en-int-train.snt -m1 5 -m2 0 -m3 3 -m4 3 > > -model1dumpfrequency 1 -model4smoothfactor 0.4 -nodumps 1 -nsmooth 4 -o > > demo-en-fr/giza.fr-en/fr-en -onlyaldumps 1 -p0 0.999 -s > > demo-en-fr/corpus/en.vcb -t demo-en-fr/corpus/fr.vcb > > > > Any hints would be greatly appreciated. > > > > Thanks, > > John > > > > _______________________________________________ > > 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
