cmu lmtk failed to produce language model

Speech Recognition Toolkit

Brought to you by: air, arthchan2003, awb, bhiksha, and 5 others

This project can now be found here.

cmu lmtk failed to produce language model

Forum: Help

Creator: vijayabharadwaj gsr

Created: 2010-02-26

Updated: 2012-09-22

vijayabharadwaj gsr - 2010-02-26

Dear Sir,

I want to produce 7 gram language model for my experiment. The toolkit
produces wngram and idngram but failes to produce lm.out. I am sharing my
files here. Please give me suggestion.

http://www.4shared.com/file/230161477/5e1065ad/cmulmmodeltar.html

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

vijayabharadwaj gsr - 2010-02-26

Actually I added this topic in wrong forum then I was suggested to create
here. Sorry for multiple posts.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Nickolay V. Shmyrev - 2010-02-26

Since you are using 7-grams, you need to point 7 discount ranges if you want
to point them at all:

-disc_ranges 1 1 1 1 1 1 1 1

Also, for subword LM I'd recommend to add marker to the morpheme than to make
it separate word. It's more reasonable hack I think.

If you would like to refer to this comment somewhere else in this project, copy and paste the following link:

Log in to post a comment.