Hi, this should not happen with the binarized translation table. It should load failry quickly (seconds, if that) and not use very little RAM. Maybe something went wrong with the binarization? How about the reordering table (if you use one)?
-phi On Sat, May 24, 2008 at 2:37 PM, sushil ronghe <[EMAIL PROTECTED]> wrote: > > Hello all, > > I have trained the translation model using english-german parallel corpus > (europarl) > Size of phrase-table = 4.2 GB and language model = 249.1MB > > I am getting this error: > Loading table into memory..terminate called after throwing an instance of > 'std:bad_alloc' > what(): std:bad_alloc > Aborted (core dumped) > > > I tried with binary tables : collection of binary file produce is about 2 GB > But I am still getting the same error. (I have also configured the moses.ini > so that > it will find the binary tables). > I am using moses compiled with irstlm on ubuntu > (machine configuration: core 2 Duo 2gz 2.512 GB of ram). > Should i need to upgrade my ram or is there any other way. > > > thanks > > > > > > -- > > > _______________________________________________ > 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
