Speech:Exps 0282 026

Description
Author: Peter Ferro

Date: 4/23/16

Purpose: To test another new version of makeTest.pl.

Details: This is a test of the softlinking system, as per Mr. Jonas's request to put them back in. I am giving version 7 of Speech:GenFeats.pl a go, as it contains the latest and greatest in handling unequal sub-experiment IDs. I also modified the parameters sent to run_decode.pl to deal with this situation.

Results: I started with this command... /mnt/main/scripts/user/History/makeTest/15/makeTest.pl train/trans/train.trans switchboard/old/first_5hr 0282/005 0282/026 After successfully running the script, I then ran... /mnt/main/scripts/user/History/genFeats/7/genFeats.pl -d 005 ...followed by... nohup run_decode.pl 005 0282/026 1000 & The decode seems to have run without any serious problems, based on a lookup in decode.log. That's exactly what I'm looking for, which means softlinking can properly be done with makeTest.