Skip to main content

Installing SRILM 156 in Ubuntu 8.04

0. untar the source distribution in SRILM_DIR

1. set the environment variable
bash> export SRILM=SRILM_DIR

2. check the gcc version and the location of tcl library
e.g.
bash> gcc --version
bash> which gcc
bash> whereis tcl

3. edit the common/Makefile.machine.i686-gcc4

bash> chmod u+rwx Makefile.machine.i686-gcc4
bash> emacs -nw Makefile.machine.i686-gcc4
# GCC_FLAGS = -mtune=pentium3 -Wreturn-type -Wimplicit
# CC = /usr/local/lang/gcc-4.2.2/bin/gcc $(GCC_FLAGS)
# CXX = /usr/local/lang/gcc-4.2.2/bin/g++ $(GCC_FLAGS) -DINSTANTIATE_TEMPLATES
GCC_FLAGS = -mtune=pentium3 -Wreturn-type -Wimplicit
CC = /usr/bin/gcc $(GCC_FLAGS)
CXX = /usr/bin/g++ $(GCC_FLAGS) -DINSTANTIATE_TEMPLATES
...
#ADDITIONAL_INCLUDES =
ADDITIONAL_INCLUDES = -I/usr/include/tcl

4. Make
bash> make 1> logFile 2> errFile

5. That is it

Comments

Ferosh jacob said…
Check this link, they have the debian packages for almost all static LMs
http://cl.aist-nara.ac.jp/~eric-n/ubuntu-nlp/dists/dapper/nlp/
For srilm for jaundy,i86 it is
http://cl.aist-nara.ac.jp/~eric-n/ubuntu-nlp/pool/dapper/nlp/srilm_1.5.9-1nlp1~0dapper1_i386.deb

Popular posts from this blog

Semi-supervised Learning

Ways of Doing Semi-supervised algorithm Transductive SVM ( minimize the risk directly ) Co-training and Tri-training ( using two/three different view of the same datasets) Genetic Algorithm Non-parametric approach ( modified KDE convergence algorithm ) Parametric approach ( modified EM algorithm ) Graph-based approach