Re: [Moses-support] integration of efmaral word alignment in Moses pipeline/EMS
> > From the GitHub pages it appears that eflomal supercedes efmaral this is true, I guess > — is there any purpose therefore in using efmaral? it is a bit faster (but consumes more memory). eflomal does not support the fast_align input format at the moment but that can easily be fixed if necessary. > Also, the linked PBML paper has no mention of eflomal — how does it perform > in downstream BLEU tasks? Is it comparable to what you reported in Table 4? eflomal did not exist at the tie of writing that paper. We haven’t done the BLEU evaluations but I expect no big difference. Actually, efmoral can be run with parameter settings that makes it equivalent to efmaral. So, eflomal should be the way to go. It wouldn’t be too hard to support both I guess but I don’t dare to touch the experiment,perl monster Jörg > matt > > >> On Dec 7, 2016, at 2:50 AM, Jorg Tiedemann wrote: >> >> >> efmaral and eflomal are efficient Markov chain word aligners using Gibbs >> sampling that can be used to replace GIZA++/fast_align in the typical Moses >> training pipelines: >> >> https://github.com/robertostling/efmaral >> https://github.com/robertostling/eflomal >> >> Would anyone be interested in adding support in the Moses pipelines and >> experiment.perl? >> Input and output formats are compatible with fast_align and Moses formats. >> >> The tools could also be mentioned at statmt.org/moses >> >> All the best, >> Jörg >> >> — >> Jörg Tiedemann >> Department of Modern Languages >> University of Helsinki >> http://blogs.helsinki.fi/language-technology/ >> — >> >> >> >> >> >> >> ___ >> 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 ___ Moses-support mailing list Moses-support@mit.edu http://mailman.mit.edu/mailman/listinfo/moses-support
Re: [Moses-support] integration of efmaral word alignment in Moses pipeline/EMS
Hi, From the GitHub pages it appears that eflomal supercedes efmaral — is there any purpose therefore in using efmaral? Also, the linked PBML paper has no mention of eflomal — how does it perform in downstream BLEU tasks? Is it comparable to what you reported in Table 4? matt > On Dec 7, 2016, at 2:50 AM, Jorg Tiedemann wrote: > > > efmaral and eflomal are efficient Markov chain word aligners using Gibbs > sampling that can be used to replace GIZA++/fast_align in the typical Moses > training pipelines: > > https://github.com/robertostling/efmaral > https://github.com/robertostling/eflomal > > Would anyone be interested in adding support in the Moses pipelines and > experiment.perl? > Input and output formats are compatible with fast_align and Moses formats. > > The tools could also be mentioned at statmt.org/moses > > All the best, > Jörg > > — > Jörg Tiedemann > Department of Modern Languages > University of Helsinki > http://blogs.helsinki.fi/language-technology/ > — > > > > > > > ___ > 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
[Moses-support] integration of efmaral word alignment in Moses pipeline/EMS
efmaral and eflomal are efficient Markov chain word aligners using Gibbs sampling that can be used to replace GIZA++/fast_align in the typical Moses training pipelines: https://github.com/robertostling/efmaral https://github.com/robertostling/eflomal Would anyone be interested in adding support in the Moses pipelines and experiment.perl? Input and output formats are compatible with fast_align and Moses formats. The tools could also be mentioned at statmt.org/moses All the best, Jörg — Jörg Tiedemann Department of Modern Languages University of Helsinki http://blogs.helsinki.fi/language-technology/ — ___ Moses-support mailing list Moses-support@mit.edu http://mailman.mit.edu/mailman/listinfo/moses-support