umm.... if i were to make a trigram language model, should input an idngram file with trigrams only or can i also put unigrams and bigrams along with the trigrams to that idngram file? if so, is there a format or you just arrange them by unigrams then bigrams then trigrams?
thx in advance...
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
From my experience idngram file just contains trigrams either in ascii format or in binary. Each line has n ids and a number of occurences. Most programs working with idngram have -n argument to point n.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
umm.... if i were to make a trigram language model, should input an idngram file with trigrams only or can i also put unigrams and bigrams along with the trigrams to that idngram file? if so, is there a format or you just arrange them by unigrams then bigrams then trigrams?
thx in advance...
From my experience idngram file just contains trigrams either in ascii format or in binary. Each line has n ids and a number of occurences. Most programs working with idngram have -n argument to point n.