Title: Tiny Train #2
Author: Cedric Woodbury
Date: July 24, 2012
Purpose: A second training run for proof of concept of a working system
For this one experiment, I copied over the data and sphinx_train.cfg file from old/train/train2. That train was successful because it produced the .cont files instead of the .semi files that I got from experiment 0007
Unfortunately this did not work. RunAll.pl was not happy with the data that I had, so a direct copy did not work.
Directly copying the data from the other experiment folder did not work. I need to dig further to find out why they got the correct files for decoding and why mine did not work.