VoxForge
VoxForge
When I execute:
HERest -A -D -T 1 -C config -I phones1.mlf -t 250.0 150.0 3000.0 -S train.scp -H hmm5/macros -H hmm5/hmmdefs -M hmm6 monophones1
It give me this error
ERROR [+7332] CreateInsts: Cannot have successive Tee models
FATAL ERROR - Terminating program HERest
I check all the previous file in hmm5, and they are like the example of the tutorial.
Someone can help me?
Tks
Hi Manuel,
>"Cannot have successive Tee models"
I've seen this error before. This is basically HTK telling you that you cannot have two silence models ("sp" or "sil") in a row. It can be tricky to isolate the source of the problem.
You have to look further back in your steps to see what might be causing the problem.
In Step 2, did you create the global.ded script containing (note that the HDMan command will run even if you don't create it):
| AS sp RS cmu MP sil sil sp |
The third line merges "sil sil" to sp, to avoid successive Tee errors.
Did you add any new prompt words? Did you put a "silence" word or a
SENT-END SENT-START in your prompt file? Does you dictionary (dict) file have any phoneme entries with"sp sp", "sp sil", "sil sil", or "sil sp"?
Ken
I found the problem: in dict file I had two "sp" at the end of each word pronunciation. I modified it with only one short-pause and now it work correctly; but I don't understand how I have this problem in dict if dictionary has only one "sp", anyway thanks a lot for your help.
Manuel
Hi Manuel,
The "AS sp" command in the global.ded script from Step 2:
|
AS sp |
tells HDMan to add a short-pause ("sp") to the end of each pronunciation. So if the pronunciations in your lexicon already contain a short-pause, you will be adding a second sp.
Contrary to what I stated in an earlier post, the MP command merges "sil" and "sp" phones together into a single "sil".
See the HTKBook for details.
Ken