VoxForge
Hi
I receive the following error message when attempting to create hmm8 files using HERest command
Here is the output from HERest:
Processing Data: sample1.mfc; Label sample1.lab
Error [+7332] CreateInsts: Cannot have Tee models at start or end of transcription
Can anyone tell me what is going wrong here an what I need to change in order to proceed?
Thanks
Figured it out. For some reason in my aligned.mlf file every instance of "sil" was followed by an erroneous "sp". I figured it out by comparing my aligned.mlf file with the example file so thanks for giving example output in the tutorial :) . After removing the extra "sp" lines everything proceeded nicely.
HVite -A -D -T 1 -l '*' -o SWT -b SENT-END -C config -H hmm7/macros -H hmm7/hmmdefs -i aligned.mlf -m -t 250.0 150.0 1000.0 -y lab -a -I words.mlf -S train.scp dict monophones1> HVite_log
My problem with dict file
My dict file does not contain sp at the en of each line??i do not know why? Any way if I used it in command above I get aligned.mlf but without sp???
If I add sp at each line in my dict
I got empty aligned file??
What I can do???
Another questions :why we need exactly SENT-END not silence as there in toturials?????
>My dict file does not contain sp at the en of each line??
Did you create your global.ded filein Step 2 (with the "AS sp" command)?
>why we need exactly SENT-END not silence as there in
>toturials?????
See Step 2.
You could probably use any key-word as long as you add it to your pronunciation dictionary, with a pronunciation of 'sil'.