tidigits.lm

来自「WinCE平台上的语音识别程序」· LM 代码 · 共 64 行

#############################################################################
## Copyright (c) 1996, Carnegie Mellon University, Cambridge University,
## Ronald Rosenfeld and Philip Clarkson
#############################################################################
=============================================================================
===============  This file was produced by the CMU-Cambridge  ===============
===============     Statistical Language Modeling Toolkit     ===============
=============================================================================
This is a 2-gram language model, based on a vocabulary of 13 words,
  which begins "<s>", "</s>", "oh"...
This is an OPEN-vocabulary model (type 1)
  (OOVs were mapped to UNK, which is treated as any other vocabulary word)
Absolute discounting was applied.
1-gram discounting constant : 0.04
2-gram discounting constant : 0.04
This file is in the ARPA-standard format introduced by Doug Paul.

p(wd3|wd1,wd2)= if(trigram exists)           p_3(wd1,wd2,wd3)
                else if(bigram w1,w2 exists) bo_wt_2(w1,w2)*p(wd3|wd2)
                else                         p(wd3|w2)

p(wd2|wd1)= if(bigram exists) p_2(wd1,wd2)
            else              bo_wt_1(wd1)*p_1(wd2)

All probs and back-off weights (bo_wt) are given in log10 form.

Data formats:

Beginning of data mark: \data\
ngram 1=nr            # number of 1-grams
ngram 2=nr            # number of 2-grams

\1-grams:
p_1     wd_1 bo_wt_1
\2-grams:
p_2     wd_1 wd_2 

end of data mark: \end\

\data\
ngram 1=14
ngram 2=1

\1-grams:
-1.6805 <UNK>	0.0000
-99.0000 <s>	0.0000
-1.3795 </s>    0.0000
-1.0695 oh	0.0000
-1.0695 zero	0.0000
-1.0695 one	0.0000
-1.0695 two	0.0000
-1.0695 three	0.0000
-1.0695 four	0.0000
-1.0695 five	0.0000
-1.0695 six	0.0000
-1.0695 seven	0.0000
-1.0695 eight	0.0000
-1.0695 nine	0.0000

\2-grams:
-99.0177 </s> <s> 

\end\