converting ngram format model to AT&T FSM format

Dilek Hakkani dilekh at
Thu Sep 8 20:29:01 PDT 2005

I'm trying to convert an n-gram model (e.g., a.lm) into AT&T FSM format.
I have first used make-ngram-pfsg (e.g., make-ngram-pfsg a.lm > a.pfsg), then I used pfsg-to-fsm (e.g., pfsg-fsm a.pfsg > a.fsm). I have some questions regarding the interpretation of the transition probabilities and labels:
1. words are represented as themselves in the n-gram format, but in the FSM format model, the transitions seem to have an index. Which word is represented with which index? Can it be extracted from the order of the unigrams in the ngram format file? Is 0 representing an epsilon?
2. Are the transition probabilities -10000.5*logprobs?
3. What do the state potentials represent?
Also, is there a better way of doing these?
I appreciate any help about these.
Dilek Hakkani-Tür

 Click here to donate to the Hurricane Katrina relief effort.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the SRILM-User mailing list