GitHub - Futrell/lmpy: Custom n-gram language model in python

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 34 Commits
.gitignore		.gitignore
README		README
bible.txt		bible.txt
context_cols		context_cols
context_mat.gz		context_mat.gz
context_rows		context_rows
contextparsers.py		contextparsers.py
contexttypes.py		contexttypes.py
defaultordereddict.py		defaultordereddict.py
distribMat.py		distribMat.py
distribmat.py		distribmat.py
kjv.lm.gz		kjv.lm.gz
lm.py		lm.py
lunyu.txt		lunyu.txt
makeDistribMat.py		makeDistribMat.py
out_cols		out_cols
out_mat		out_mat
out_rows		out_rows
possible OOV solution.txt		possible OOV solution.txt
probest.py		probest.py
quran.txt		quran.txt
similarity.py		similarity.py
test.txt		test.txt
tng-bigrams		tng-bigrams
tok.py		tok.py
typicality.py		typicality.py
weighting.py		weighting.py
words		words

Repository files navigation

These are some python things I've made so I can experiment with various smoothing methods and have an easy-to-customize language model handy for future other experiments.
The main thing I'm trying out is a new kind of similarity-based smoothing.