====== MGIZA ======

MGIZA++ is a multi-threaded word alignment tool based on GIZA++. It extends GIZA++ in multiple ways:


** Multi-threading **

MGIZA++ can make use of multi-core platforms efficiently. Usually a quad-core machine can have a three-fold speedup over single-thread GIZA++.

** Memory optimization **

By eliminating duplicated tables, MGIZA++ can save a lot of memory comparing to GIZA++. 

** Resume training **

MGIZA++ can resume training from any stage and continue training. For example you may be able to re-use previous available models and continue training directly from IBM Model 4 instead of all the way from Model 1.

** Integrated with Chaski **

MGIZA++ can be integrated into Chaski and run on cluters, which will give you even larger speedup.
===== Download =====

Latest version of MGIZA++ can be download here:
^ Version ^ Data ^ Link ^ Comment ^Release Note  ^
| Version 0.6.3.1 | 2010-01-23 |  [[https://sourceforge.net/projects/mgizapp/files/mgizapp-0.6.3.tar.gz/download|Download]] | Minor code clean and move download to Sf | [[Release Note MGiza#0.6.3 | Release Note]] |
| Version 0.6.3 | 2010-01-11 |  [[http://www.cs.cmu.edu/~qing/release/mgiza-0.6.3-10-01-11.tar.gz|Download]] | Memory optimization and bug fix  | [[Release Note MGiza#0.6.3 | Release Note]] |
| Version 0.6.2 | 2009-12-07 |  [[http://www.cs.cmu.edu/~qing/release/mgiza-0.6.2-09-12-07.tar.gz|Download]] |Minor interface change to keep compatibility with [[chaski:Release Note Chaski#0.2.2|Chaski 0.2.2]]  | [[Release Note MGiza#0.6.2 | Release Note]] |
| Version 0.6.1 | 2009-11-17 |  [[http://www.cs.cmu.edu/~qing/release/mgiza-0.6.1-09-11-17.tar.gz|Download]] |Unnecessary dependencies removed   | [[Release Note MGiza#0.6.1 | Release Note]] |
| Version 0.6 | 2009-11-10 | [[http://www.cs.cmu.edu/~qing/release/qmt-0.6-chc-09-11-10.tar.gz|Download]] | | |

===== Installation =====

<del>To compile MGIZA++ you need the following package installed:</del>

   - <del>Berkeley DB (libdb)</del>
   - <del>Berkeley DB++ (libdb++)</del>
   - Boost library. (regex, string)


<del>After the dependencies are installed</del>. As of version 0.6.1, you do not need the dependencies of berkeley db, but you still need boost library.. Just go to the source directory of the source and 

<code bash>
  ./configure --prefix=${QMT_HOME}
  make
  make install
</code>

If you want to use MGIZA++ with Chaski, you need to add the environment variable ''QMT_HOME'' to your ''.bashrc''.

For boost library you can either download it from [[http://www.boost.org]] or install the header package of your linux distribution. 

===== Usage =====


The basic usage of MGIZA++ is easy, given that you know how to run GIZA++. MGIZA++ is compatible with GIZA++'s parameters, and you can run:

<code bash>
  ${QMT_HOME}/bin/mgiza  -ncpu 5 [ALL-YOUR-GIZA-PARAMETERS]
</code>

to tell mgiza to run five-threads.

The alignment output of MGIZA++ is somehow different from GIZA++, given n-threads, the alignment output will be:

<code bash>
prefix.A3.final.part0
prefix.A3.final.part1
...
prefix.A3.final.part(n-1)
</code>

To combine the alignments you need to run:

<code bash>
${QMT_HOME}/scripts/merge_alignment.py ${prefix}.A3.final.part* > ${prefix}.A3.final
</code>

For advanced usage please refer to the following "HOWTOs"

  * Force alignment and resume training [[mgiza:forcealignment]]