Skip to content

GIZA++ is a statistical machine translation toolkit that is used to train IBM Models 1-5 and an HMM word alignment model. This package also contains the source for the mkcls tool which generates the word classes necessary for training some of the alignment models.

Notifications You must be signed in to change notification settings

moses-smt/giza-pp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

32 Commits
 
 
 
 
 
 
 
 

Repository files navigation

This package contains the GIZA++ toolkit and the mkcls tool, originally
written by F.J. Och and several other authors.

For more information, refer to the README files and the following pages:
  http://www.fjoch.com/mkcls.html
  http://www.fjoch.com/GIZA++.html

UPDATE 4th April 2015, by Hieu Hoang:

Works with most versions of gcc and clang, including really old versions.

Why not try using mgiza? It's multi-threaded and kept up to date to work with most compilers
   https://github.com/moses-smt/mgiza

About

GIZA++ is a statistical machine translation toolkit that is used to train IBM Models 1-5 and an HMM word alignment model. This package also contains the source for the mkcls tool which generates the word classes necessary for training some of the alignment models.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages