Search SRILM-USER Archives

Match: Format: Sort by:
Search:

converting ngram format model to AT&T FSM format

From: Dilek Hakkani <dilekh at ADDRESS HIDDEN>
Date: Thu, 8 Sep 2005 20:29:01 -0700 (PDT)

--0-1568582159-1126236541=:81149
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit

Hi,
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.

Thanks,
Dilek Hakkani-Tür



---------------------------------
Click here to donate to the Hurricane Katrina relief effort.
--0-1568582159-1126236541=:81149
Content-Type: text/html; charset=iso-8859-1
Content-Transfer-Encoding: 8bit

<DIV>Hi,</DIV>
<DIV>I'm trying to convert an n-gram model (e.g., a.lm) into AT&T FSM format.</DIV>
<DIV>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:</DIV>
<DIV>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?</DIV>
<DIV>2. Are the transition probabilities -10000.5*logprobs?</DIV>
<DIV>3. What do the state potentials represent?</DIV>
<DIV> </DIV>
<DIV>Also, is there a better way of doing these?</DIV>
<DIV> </DIV>
<DIV>I appreciate any help about these.</DIV>
<DIV> </DIV>
<DIV>Thanks,</DIV>
<DIV>Dilek Hakkani-Tür</DIV>
<DIV> </DIV><p>
<hr size=1> <a href="http://store.yahoo.com/redcross-donate3/">Click here to donate to the Hurricane Katrina relief effort.</a>
--0-1568582159-1126236541=:81149--

Click here to go to the SRILM home page.