No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore
README.md
__init__.py
berkeley_pbs_setup.py
config.py
dict_utils.py
feature.py
feature_set.py
fridge.py
hypergraph.py
lambda_calc.py
lambda_calc2.py
list_utils.py
logging_utils.py
math_utils.py
misc_utils.py
optimize.py
pbs_utils.py
py_utils.py
scientist.py
semiring.py
string_utils.py
test.py
training.py
tree_utils.py
vocabulary.py

README.md

natural language processing tools

While optimized versions of all these things exist, it's nice to have a clean implementation of various NLP workhorses (hypergraph inside/outside, perceptron & gradient training, etc.) lying around. Those will go here.

currently implemented:

  • Packed forest I-O with arbitrary semirings
  • The expected counts semiring for log-linear models.

to do

  • A generic CKY parser
  • Training algorithms: perceptron, gradient descent, EM