> > > Hi , > > Thank you for your reply and suggestions. > > The problem get solved after commenting > [//CHECK(factors.size()==m_output.size());] line no. 273 in file > "moses/src/PDTAimp.h. > I personally wouldn't so that.
> I Have some observations to share- > > 1- > Earlier I was not using any factor at source side(English) and single > factor(POS) at target side(Hindi). The sample of training corpus - > ------- > Source - fresh breath and shining teeth your personality enhance . > Target - ताजा|JJ साँसें|NN और|CCD > चमचमाते|JJ दाँत|NN आपके|PRP > व्यक्तित्व|NN को|PSP निखारते|VM > हैं|VAUX ।|PUNC > ------- > And the mapping was only for translation as- > -translation-factors 0-0,1 > After converting phrase table in binary format configuration in moses.ini > for [ttable-file] was- > ([-ttable-file] 1 0 0,1 5 > /home/raj/anuvad/engSource/tgtPosEng-hindi/model/phrase-table.0-0,1 ) > > As I had told in earlier conversation that I stuck with some decoding > error after binarizing the phrase table. > ERROR-(Check factors.size()==m_output.size() failed in > moses/src/PDTAimp.h:273). > > But when I recompiled the moses commenting > [//CHECK(factors.size()==m_output.size());] line no. 273 in file > "moses/src/PDTAimp.h", Everything started working fine. > > 2- > Again I had trained the system with factors at source as well as at > target. sample of training corpus- > ------- > Source - Fresh|JJ|fresh breath|NN|breath and|CC|and shining|VBG|shine > teeth|NNS|tooth enhance|VBP|enhance your|PRP$|you > personality|NN|personality .|.|. > Target - ताजा|JJ|ताज साँसें|NN|साँस > और|CCD|और चमचमाते|JJ|चमचम > दाँत|NN|दाँत आपके|PRP|आपक > व्यक्तित्व|NN|व्यक्तित्व > को|PSP|को निखारते|VM|निखार > हैं|VAUX|हैं ।|PUNC|। > ------- > There was no problem at any stage. Everything get done. > It should be noted that moses was recompiled without commenting line no > 273 in file "moses/src/PDTAimp.h". > > > I think condition "CHECK(factors.size()==m_output.size());" is unable to > handle the case if we have no factor at source side and some factors at > target side or vice versa. Please correct if I am wrong. You have factors in your source training data but no factors in your source input. That is more likely the problem. I have used no factors source + factors target many times without problems > > Thanks. > > > Regards: > Raj Nath Patel > > > > > > On Thu, Mar 14, 2013, Philipp Koehn <[email protected]> said: > >> Hi, >> >> this should work and it is not clear why you are experiencing >> an error here. Are you sure that after binarizing the model >> and changing the config file that you still have the right number >> of factors specified in the [ttable-file] section? >> >> -phi >> >> On Sat, Mar 9, 2013 at 10:37 AM, Rajnath Patel <[email protected]> wrote: >>> Hi, >>> >>> I have trained the factored model using moses (release 0.91) and >>> GIZA++ (version 1.0.7). Everything was working fine as after training >>> as well as after tuning . But after converting the phrase table and >>> reordering model >>> >>> in binary format , I stuck with the following decoding error - >>> >>> ------------------------- >>> Finished loading phrase tables : [0.693] seconds >>> IO from STDOUT/STDIN >>> Created input-output object : [0.693] seconds >>> how are you ? >>> Translating line 0 in thread id 2811820864 >>> Translating: how are you ? >>> >>> reading bin ttable >>> size of OFF_T 8 >>> binary phrasefile loaded, default OFF_T: -1 >>> binary file loaded, default OFF_T: -1 >>> Check factors.size()==m_output.size() failed in moses/src/PDTAimp.h:273 >>> Aborted (core dumped) >>> -------------------------------- >>> >>> What should I do ? Any suggestion ? >>> >>> Thanks. >>> >>> -- >>> Regards: >>> Raj Nath Patel >>> KBCS dept. >>> CDAC Mumbai, India. >>> _______________________________________________ >>> Moses-support mailing list >>> [email protected] >>> http://mailman.mit.edu/mailman/listinfo/moses-support >> >> :: Disclaimer :: >> The information contained in this e-mail message is intended only for the use >> of the individual or entity to which it is addressed. If you are not the >> intended recipient, you should return it to the sender immediately. Please >> note that while we scan all e-mails for viruses we cannot guarantee that any >> e-mail is virus-free and accept no liability for any damage caused by any >> virus transmitted by this email. >> ~ > > -- > > > > > ------------------------------------------------------------------------------------------------------------------------------- > > This e-mail is for the sole use of the intended recipient(s) and may > contain confidential and privileged information. If you are not the > intended recipient, please contact the sender by reply e-mail and destroy > all copies and the original message. Any unauthorized review, use, > disclosure, dissemination, forwarding, printing or copying of this email > is strictly prohibited and appropriate legal action will be taken. > ------------------------------------------------------------------------------------------------------------------------------- > _______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
