[hts-users:00662] Re: Unseen Model For 1 Mix

["Heiga ZEN (Byung Ha CHUN)" <zen@xxxxxxxxxxxxxxx>]

Hi,

Tamer Fares wrote:

> I am now in the stage of making unseen models(1 Mix)
>
> i checked the log file , i found the the HHEd function load the edit 
> file mku.hed from the folder cmp firstly
>
> after finishing the HHEd load again the file mku.hed from the folder dur
>
> i checked the contents of the file mku.hed stored in the dur , i found 
> that the first line is the loading of " dur.inf.untied.2mix"  , while 
> this file is generated later in the step of tree based context 
> clustering (dur ,2mix) with convstats() function
>
> how the loading of this file come first and it's not generated yet

Did you really read Training.pl?
If you read it, you might found the following lines:

# fix variables
$tre{'dur'}     .= ".2mix";
$cxc{'dur'}     .= ".2mix";
$stats{'dur'}   .= ".2mix";
$tiedlst{'dur'} .= ".2mix";
$untymmf{'dur'} .= ".2mix";
$reclmmf{'dur'} .= ".2mix";
$rclammf{'dur'} .= ".2mix";
$reclmmf{'cmp'} .= ".2mix";
$rclammf{'cmp'} .= ".2mix";

Please understand the meaning of these lines and read the training script again.

# Ask any questions *after checking* the scripts and HTKBook.
# If you read them carefully, you may not ask this kind of boring questions.
# I'm so tired...

Regards,

Heiga ZEN (Byung Ha CHUN)

--
------------------------------------------------
Heiga ZEN     (in Japanese pronunciation)
Byung Ha CHUN (in Korean pronunciation)

Department of Computer Science and Engineering
Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya 466-8555 Japan

http://www.sp.nitech.ac.jp/~zen
------------------------------------------------