What steps will reproduce the problem?
Run estimate-lm -v vocab.file -t training.txt -wl lm.arpa. Where vocab.file
contains a subset of the vocab in the LM training text.
What is the expected output? What do you see instead?
Program output:
0.000 Loading vocab vocab...
0.010 Loading corpus training.txt...
Segmentation fault
What version of the product are you using? On what operating system?
Latest SVN, ubuntu 10.04 64bit
Original issue reported on code.google.com by
edohashira@gmail.comon 14 Jul 2011 at 4:34