error at .wav file while training acoustic

[Basit Mahmood]

Hi to all,
Hope you all will be fine. Actually i am training acoustic model. I read the
tutorial Training Acoustic Model For CMUSphinx and then start to follow each
step. When i came to the stage where i supposed to run the command

./scripts_pl/make_feats.pl -ctl etc/testdb_train.fileids

then i got the following stack trace including error in the last

./scripts_pl/make_feats.pl -ctl etc/testdb_train.fileids

-cfg not specified, using the default ./etc/sphinx_train.cfg
-param not specified, using the default ./etc/feat.params
bin/wave2feat \
-verbose yes \
-alpha 0.97 \
-dither yes \
-doublebw no \
-nfilt 40 \
-ncep 13 \
-lowerf 133.33334 \
-upperf 6855.4976 \
-nfft 512 \
-wlen 0.0256 \
-c etc/testdb_train.fileids \
-mswav yes \
-di /usr/basit/sphinx/tutorial/testdb/wav \
-ei wav \
-do /usr/basit/sphinx/tutorial/testdb/feat \
-eo mfc

-help no no
-example no no
-i
-o
-c etc/testdb_train.fileids
-nskip
-runlen
-di /usr/basit/sphinx/tutorial/testdb/wav
-ei wav
-do /usr/basit/sphinx/tutorial/testdb/feat
-eo mfc
-nist no no
-raw no no
-mswav no yes
-input_endian little little
-nchans 1 1
-whichchan 1 1
-logspec no no
-feat sphinx sphinx
-mach_endian little little
-alpha 0.97 9.700000e-01
-srate 16000.0 1.600000e+04
-frate 100 100
-wlen 0.025625 2.560000e-02
-nfft 512 512
-nfilt 40 40
-lowerf 133.33334 1.333333e+02
-upperf 6855.4976 6.855498e+03
-ncep 13 13
-doublebw no no
-warp_type inverse_linear inverse_linear
-warp_params
-blocksize 200000 200000
-dither yes yes
-seed -1 -1
-verbose no yes
INFO: fe_interface.c(100): You are using the internal mechanism to generate
the seed.
INFO: fe_sigproc.c(752): Current FE Parameters:
INFO: fe_sigproc.c(753): Sampling Rate: 16000.000000
INFO: fe_sigproc.c(754): Frame Size: 410
INFO: fe_sigproc.c(755): Frame Shift: 160
INFO: fe_sigproc.c(756): FFT Size: 512
INFO: fe_sigproc.c(757): Lower Frequency: 133.333
INFO: fe_sigproc.c(758): Upper Frequency: 6855.5
INFO: fe_sigproc.c(759): Number of filters: 40
INFO: fe_sigproc.c(760): Number of Overflow Samps: 0
INFO: fe_sigproc.c(761): Start Utt Status: 0
INFO: fe_sigproc.c(763): Will add dither to audio
INFO: fe_sigproc.c(764): Dither seeded with -1
INFO: fe_sigproc.c(771): Will not use double bandwidth in mel filter
INFO: wave2feat.c(139): /usr/basit/sphinx/tutorial/testdb/wav/basit/bandar.wav
LENGTH: 4
INFO: wave2feat.c(786): Reading MS Wav file
/usr/basit/sphinx/tutorial/testdb/wav/basit/bandar.wav:
INFO: wave2feat.c(787): 16 bit PCM data, 2 channels 86240 samples
INFO: wave2feat.c(788): Sampled at 22050
ERROR: "wave2feat.c", line 883: unknown input file format
ERROR: "wave2feat.c", line 201: error reading speech data
FATAL_ERROR: "wave2feat.c", line 90: error converting files...exiting
You have new mail in /var/spool/mail/root

Please tell me why this error is coming.Is this error coming because of wav
file? Actually i just open the recorder in windows xp and then record the
sound. Am i supposed to record my sound with some different tool.

Thanks

[Basit Mahmood]

Hi the problem has solved, it was actually, file was not sampled at 16KH, 16
bit mono. But i have a problem now. In the last step the trainer is not
matching phones in .dictionary and .phone file.
Here what i did.

I created a corpus file with sentences

My name is basit
What is your name
This is a cat
Urdu
English
Songs
Dog is an animal
Wo aik Bandar hai (Roman Englis just did it for experiment)
Kabutar aik parinda hai ( (Roman Englis just did it for experiment))

Then i record a file with each sentence, sample it with 16KHz and 16 bit mono.
Then i created a dictionary model by uploading my corpus file to this link

http://www.speech.cs.cmu.edu/tools/lmtool.html

i used the new version. It created the .dic and .lm file. Here is my .dic file
produced by the lmtool

A AH
A(2) EY
AIK EY K
AN AE N
AN(2) AH N
ANIMAL AE N AH M AH L
BANDAR B AE N D AA R
BASIT B AE S AH T
CAT K AE T
DOG D AO G
ENGLISH IH NG G L IH SH
ENGLISH(2) IH NG L IH SH
HAI HH EY
IS IH Z
KABUTAR K AE B Y UW T AH R
MY M AY
NAME N EY M
PARINDA P AH R IH N D AH
SONGS S AO NG Z
THIS DH IH S
URDU ER D UW
WHAT W AH T
WHAT(2) HH W AH T
WO W OW
WO(2) HH W OW
YOUR Y AO R
YOUR(2) Y UH R

Then i created a .phone file and copy each phoneme exactly one time to this
file from .dic file. Here is my .phone file.

AH
EY
K
AE
N
M
L
B
D
AA
R
S
T
K
AO
G
IH
NG
SH
HH
EY
Z
Y
UW
AY
EY
P
DH
ER
W
OW
UH
SIL

here is my .transcript file

Wo aik bandar hai
(/usr/basit/sphinx/tutorial/testdb/wav/basit/bandar)
This is a cat (/usr/basit/sphinx/tutorial/testdb/wav/basit/cat)
Dog is an animal (/usr/basit/sphinx/tutorial/testdb/wav/basit/dog)
English (/usr/basit/sphinx/tutorial/testdb/wav/basit/english)
Kabutar (/usr/basit/sphinx/tutorial/testdb/wav/basit/kabutar)
My name is Basit (/usr/basit/sphinx/tutorial/testdb/wav/basit/myName)
Songs (/usr/basit/sphinx/tutorial/testdb/wav/basit/songs)
Urdu (/usr/basit/sphinx/tutorial/testdb/wav/basit/urdu)
What is your name
(/usr/basit/sphinx/tutorial/testdb/wav/basit/yourName)

But now when i run the command

[root@ivr2 testdb]# ./scripts_pl/RunAll.pl

i get these stack traces

cd /usr/basit/sphinx/tutorial/testdb/

./scripts_pl/RunAll.pl

MODULE: 00 verify training files
O.S. is case sensitive ("A" != "a").
Phones will be treated as case sensitive.
Phase 1: DICT - Checking to see if the dict and filler dict agrees with the
phonelist file.
Found 30 words using 30 phones
WARNING: This phone () occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in the
dictionary (/usr/basit/sphinx/tutorial/testdb/etc/testdb.dic)
Phase 2: DICT - Checking to make sure there are not duplicate entries in the
dictionary
Phase 3: CTL - Check general format; utterance length (must be positive);
files exist
Phase 4: CTL - Checking number of lines in the transcript should match lines
in control file
Phase 5: CTL - Determine amount of training data, see if n_tied_states seems
reasonable.
Total Hours Training: 0.00458525641025641
This is a small amount of data, no comment at this time
Phase 6: TRANSCRIPT - Checking that all the words in the transcript are in the
dictionary
Words in dictionary: 27
Words in filler dictionary: 3
WARNING: This word: Wo was in the transcript file, but is not in the
dictionary ( Wo aik bandar hai ). Do cases match?
WARNING: This word: aik was in the transcript file, but is not in the
dictionary ( Wo aik bandar hai ). Do cases match?
WARNING: This word: bandar was in the transcript file, but is not in the
dictionary ( Wo aik bandar hai ). Do cases match?
WARNING: This word: hai was in the transcript file, but is not in the
dictionary ( Wo aik bandar hai ). Do cases match?
WARNING: This word: This was in the transcript file, but is not in the
dictionary ( This is a cat ). Do cases match?
WARNING: This word: is was in the transcript file, but is not in the
dictionary ( This is a cat ). Do cases match?
WARNING: This word: a was in the transcript file, but is not in the dictionary
( This is a cat ). Do cases match?
WARNING: This word: cat was in the transcript file, but is not in the
dictionary ( This is a cat ). Do cases match?
WARNING: This word: Dog was in the transcript file, but is not in the
dictionary ( Dog is an animal ). Do cases match?
WARNING: This word: is was in the transcript file, but is not in the
dictionary ( Dog is an animal ). Do cases match?
WARNING: This word: an was in the transcript file, but is not in the
dictionary ( Dog is an animal ). Do cases match?
WARNING: This word: animal was in the transcript file, but is not in the
dictionary ( Dog is an animal ). Do cases match?
WARNING: This word: English was in the transcript file, but is not in the
dictionary ( English ). Do cases match?
WARNING: This word: Kabutar was in the transcript file, but is not in the
dictionary ( Kabutar ). Do cases match?
WARNING: This word: My was in the transcript file, but is not in the
dictionary ( My name is Basit ). Do cases match?
WARNING: This word: name was in the transcript file, but is not in the
dictionary ( My name is Basit ). Do cases match?
WARNING: This word: is was in the transcript file, but is not in the
dictionary ( My name is Basit ). Do cases match?
WARNING: This word: Basit was in the transcript file, but is not in the
dictionary ( My name is Basit ). Do cases match?
WARNING: This word: Songs was in the transcript file, but is not in the
dictionary ( Songs ). Do cases match?
WARNING: This word: Urdu was in the transcript file, but is not in the
dictionary ( Urdu ). Do cases match?
WARNING: This word: What was in the transcript file, but is not in the
dictionary ( What is your name ). Do cases match?
WARNING: This word: is was in the transcript file, but is not in the
dictionary ( What is your name ). Do cases match?
WARNING: This word: your was in the transcript file, but is not in the
dictionary ( What is your name ). Do cases match?
WARNING: This word: name was in the transcript file, but is not in the
dictionary ( What is your name ). Do cases match?
Phase 7: TRANSCRIPT - Checking that all the phones in the transcript are in
the phonelist, and all phones in the phonelist appear at least once
WARNING: This phone () occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AA) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AE) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AO) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AY) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (B) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (D) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (DH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (ER) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (EY) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (G) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (HH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (IH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (K) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (L) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (M) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (N) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (NG) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (OW) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (P) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (R) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (S) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (SH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (T) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (UH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (UW) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (W) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (Y) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (Z) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
Something failed:
(/usr/basit/sphinx/tutorial/testdb/scripts_pl/00.verify/verify_all.pl)
You have new mail in /var/spool/mail/root

why it is not matching phonemes. What i am doing wrong. Please tell me.

Thanks

[Nasir Hussain]

Hello Basit Bhai,

First of all you need minimum of 2 hours of data to create a good acoustic
model.Your data is insufficient.

Secondly, I would like to recommend you to Create your acoustic model by
referring to this http://cmusphinx.sourceforge.net/wiki/tutorialam

Thirdly,do not leave any blank space in any of the files.

Fourthly,Since java is a case sensitive language,So make your corpus file full
of upper case character.Like you used

Wo aik Bandar hai
Kabutar aik parinda hai

Just make them into uppercase and then use lmtools on it.like

WO EIK BANDAR HAI

(lol nice line used )

-Nasir

[Basit Mahmood]

Hi Nasir :)
First of all thanks for the comment :)

"WO EIK BANDAR HAI" (lol nice line used ) :D

Anyways Now some questions

First of all you need minimum of 2 hours of data to create a good acoustic
model.Your data is insufficient.

So if i haven't 2 hours of data it means i can't train my model? I am not
interested in good model right now i just want to see the progress of this bad
acoustic model.

Secondly, I would like to recommend you to Create your acoustic model by
referring to this http://cmusphinx.sourceforge.net/wiki/tutorialam

I am following the same mentioned tutorial:)

Thirdly,do not leave any blank space in any of the files.

What do you mean by blank space. Do you mean any extra line?

Fourthly,Since java is a case sensitive language,So make your corpus file
full of upper case character.

Hmm tanks for telling me this

One more question ca you tell me any site where phonemes for words are given
l. I want to know it for Roman English. Means if i break the word

WO
AIK
BANDAR
HAI

I want to produce specific sound using phonemes

Thanks

[Basit Mahmood]

sorry one more thing. In .phone file is it necessary to use each phone exactly
ones ? can i use something like this in my .phone file

AH
AH
AA
N
AH

Thanks

[Nasir Hussain]

Hello Basit,

So if i haven't 2 hours of data it means i can't train my model? I am not
interested in good model right now i just want to see the progress of this bad
acoustic model.

No,You can train your model even with a single word like "KABUTER".. But your
recognition will not be good.. But if you doing it for Testing purpose,Than go
on make it :).No worries.

I am following the same mentioned tutorial:)

If you following the same tutorial than why havent you made the .fileids file
:P.You have instead given the location of your audio files in the
transcription file...Its ok to use it.But When you got Standards thn why not
follow it :P

What do you mean by blank space. Do you mean any extra line?

Yup i Mean no extra Line as well as no unnecessary blank spaces

I want to know it for Roman English. Means if i break the word WO AIK BANDAR
HAI I want to produce specific sound using phonemes

Yes you can create phone from the words by using Sequiter G2P given here
http://www-i6.informatik.rwth-
aachen.de/web/Software/g2p.html
(keep in mind that this converter is for normal english and not for roman
english)
For Roman english you need to specify phones by yourself as there isnt any
other solution as per my knowledge ;)

In .phone file is it necessary to use each phone exactly ones ? can i use
something like this in my .phone file

Just Use each and every phones that was used in dictonary ONCE.Sirk ek
baar :)

[Basit Mahmood]

Hi thanks:)
it's been a long long long time that i ever saw a word in Urdu while
communicating on forums. But you wrote a whole sentence

Sirk ek baar. :D

If i wrote this sentence, i wrote it like this :)

Sirf aik baar ( :) just kidding)

Anyways :) now answer to your question

If you following the same tutorial than why havent you made the .fileids
file :P.You have instead given the location of your audio files in the
transcription file...Its ok to use it.But When you got Standards thn why not
follow it :P

I did but i didn't mention it here because i was getting error regarding .dic
and .phone :)

Now i am acting upon your suggestions and do the things that you suggested.
This time hope for good result.
Thanks:);)

[Nasir Hussain]

Hello Basit,

Koi nhi yaara...Any Problem Just Buzz me :)

-Nasir

[Basit Mahmood]

Hahaha :) thanks again. Cha gaye tum to yaar :D

;)

[Basit Mahmood]

Hi,
I am still facing problems. phonemes not matching. This time here my all files
.dic

A AH
A(2) EY
AIK EY K
AN AE N
AN(2) AH N
ANIMAL AE N AH M AH L
BANDAR B AE N D AA R
BASIT B AE S AH T
CAT K AE T
DOG D AO G
ENGLISH IH NG G L IH SH
ENGLISH(2) IH NG L IH SH
HAI HH EY
IS IH Z
KABUTAR K AE B Y UW T AH R
MY M AY
NAME N EY M
PARINDA P AH R IH N D AH
SONGS S AO NG Z
THIS DH IH S
URDU ER D UW
WHAT W AH T
WHAT(2) HH W AH T
WO W OW
WO(2) HH W OW
YOUR Y AO R
YOUR(2) Y UH R

.filler

SIL
<sil> SIL
</sil> SIL

.phone

AA
AE
AH
AO
AY
B
D
DH
ER
EY
G
HH
IH
K
L
M
N
NG
OW
P
R
S
SH
SIL
T
UH
UW
W
Y
Z

.filieds

basit/bandar
basit/cat
basit/dog
basit/english
basit/kabutar
basit/myName
basit/songs
basit/urdu
basit/yourName

.transcription

MY NAME IS BASIT (basit/myName)
WHAT IS YOUR NAME (basit/yourName)
THIS IS A CAT (basit/cat)
URDU (basit/urdu)
ENGLISH (basit/english)
SONGS (basit/songs)
DOG IS AN ANIMAL (basit/dog)
WO AIK BANDAR HAI (basit/bandar)
KABUTAR AIK PARINDA HAI (basit/kabutar)

and this time the stack trace are

./scripts_pl/RunAll.pl

MODULE: 00 verify training files
O.S. is case sensitive ("A" != "a").
Phones will be treated as case sensitive.
Phase 1: DICT - Checking to see if the dict and filler dict agrees with the
phonelist file.
Found 30 words using 30 phones
Phase 2: DICT - Checking to make sure there are not duplicate entries in the
dictionary
Phase 3: CTL - Check general format; utterance length (must be positive);
files exist
Phase 4: CTL - Checking number of lines in the transcript should match lines
in control file
Phase 5: CTL - Determine amount of training data, see if n_tied_states seems
reasonable.
Total Hours Training: 0.00458525641025641
This is a small amount of data, no comment at this time
Phase 6: TRANSCRIPT - Checking that all the words in the transcript are in the
dictionary
Words in dictionary: 27
Words in filler dictionary: 3
Phase 7: TRANSCRIPT - Checking that all the phones in the transcript are in
the phonelist, and all phones in the phonelist appear at least once
WARNING: This phone (AA) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AO) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (AY) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (DH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (ER) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (G) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (L) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (M) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (NG) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (OW) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (S) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (SH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (UH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (W) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (Z) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
MODULE: 01 Vector Quantization
Skipped for continuous models
MODULE: 02 Training Context Independent models for forced alignment and VTLN
Skipped: $ST::CFG_FORCEDALIGN set to 'no' in sphinx_train.cfg
Skipped: $ST::CFG_VTLN set to 'no' in sphinx_train.cfg
MODULE: 03 Force-aligning transcripts
Skipped: $ST::CFG_FORCEDALIGN set to 'no' in sphinx_train.cfg
MODULE: 04 Force-aligning data for VTLN
Skipped: $ST::CFG_VTLN set to 'no' in sphinx_train.cfg
MODULE: 05 Train LDA transformation
Skipped (set $CFG_LDA_MLLT = 'yes' to enable)
MODULE: 06 Train MLLT transformation
Skipped (set $CFG_LDA_MLLT = 'yes' to enable)
MODULE: 20 Training Context Independent models
Phase 1: Cleaning up directories:
accumulator...logs...qmanager...models...
Phase 2: Flat initialize
Phase 3: Forward-Backward
Baum welch starting for 1 Gaussian(s), iteration: 1 (1 of 1)
0%
WARNING: This step had 0 ERROR messages and 2 WARNING messages. Please check
the log file for details.
Only 0 parts of 1 of Baum Welch were successfully completed
Parts 1 failed to run!
Training failed in iteration 1
Something failed:
(/usr/basit/sphinx/tutorial/testdb/scripts_pl/20.ci_hmm/slave_convg.pl)

Now what is going wrong:( Why i am getting errors this time

Parts 1 failed to run!
Training failed in iteration 1
Something failed:
(/usr/basit/sphinx/tutorial/testdb/scripts_pl/20.ci_hmm/slave_convg.pl)

Some phonemes are matching and some aren't. As regard to Roman English
Phonemes. It's ok but standard English phonemes should match.
what i am doing wrong this time:(

Thanks

[Nasir Hussain]

Hello Basit Bhai,

Tension nhi lene ka...^_^

See the error is infront of you.You need to look it thorougly.:)

WARNING: This phone (AA) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)

Since you have a small DIctonary and you dont use AA phone.Than why you put
that in your .phone???
Remove the following phones from .phone

AA
 AO
 AY
 DH
 ER
 G 
 L
 M
 NG
 OW 
S 
SH 
UH 
W 
Z

After that run the scripts :)

-Nasir

[Basit Mahmood]

But these phones are present in my dictionary

BANDAR B AE N D AA R AA phone
DOG D AO G AO phone
MY M AY AY
THIS DH IH S DH
URDU ER D UW ER
....

All phones are in the dic file?

[Nasir Hussain]

hmmm,

Accha tell me what version of sphinxtrain you using.
And also tell me the names of files you kept in your etc folder???

-Nasir

[Basit Mahmood]

This time i removed the Roman English words and then repeat the same
procedure. Even this time i am also getting errors . Had hai yaar. This time
the stack trace are

Total Hours Training: 0.00349594017094017
This is a small amount of data, no comment at this time
Phase 6: TRANSCRIPT - Checking that all the words in the transcript are in the
dictionary
Words in dictionary: 20
Words in filler dictionary: 3
Phase 7: TRANSCRIPT - Checking that all the phones in the transcript are in
the phonelist, and all phones in the phonelist appear at least once
WARNING: This phone (HH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
WARNING: This phone (UH) occurs in the phonelist
(/usr/basit/sphinx/tutorial/testdb/etc/testdb.phone), but not in any word in
the transcription
(/usr/basit/sphinx/tutorial/testdb/etc/testdb_train.transcription)
MODULE: 01 Vector Quantization
Skipped for continuous models
MODULE: 02 Training Context Independent models for forced alignment and VTLN
Skipped: $ST::CFG_FORCEDALIGN set to 'no' in sphinx_train.cfg
Skipped: $ST::CFG_VTLN set to 'no' in sphinx_train.cfg
MODULE: 03 Force-aligning transcripts
Skipped: $ST::CFG_FORCEDALIGN set to 'no' in sphinx_train.cfg
MODULE: 04 Force-aligning data for VTLN
Skipped: $ST::CFG_VTLN set to 'no' in sphinx_train.cfg
MODULE: 05 Train LDA transformation
Skipped (set $CFG_LDA_MLLT = 'yes' to enable)
MODULE: 06 Train MLLT transformation
Skipped (set $CFG_LDA_MLLT = 'yes' to enable)
MODULE: 20 Training Context Independent models
Phase 1: Cleaning up directories:
accumulator...logs...qmanager...models...
Phase 2: Flat initialize
Phase 3: Forward-Backward
Baum welch starting for 1 Gaussian(s), iteration: 1 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 1
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 1.74042686804451
Baum welch starting for 1 Gaussian(s), iteration: 2 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 2
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 6.88838950715421
Convergence Ratio = 2.95787357321931
Baum welch starting for 1 Gaussian(s), iteration: 3 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 3
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 11.1469554848967
Convergence Ratio = 0.618223747846946
Baum welch starting for 1 Gaussian(s), iteration: 4 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 4
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 12.8393084260731
Convergence Ratio = 0.151821987938271
Baum welch starting for 1 Gaussian(s), iteration: 5 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 5
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 13.0796343402226
Training completed after 5 iterations
MODULE: 30 Training Context Dependent models
Phase 1: Cleaning up directories:
accumulator...logs...qmanager...
Phase 2: Initialization
WARNING: This step had 0 ERROR messages and 1 WARNING messages. Please check
the log file for details.
Phase 3: Forward-Backward
Baum welch starting for iteration: 1 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 1
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 13.1007392686804
Baum welch starting for iteration: 2 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 2
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 17.6407790143084
Convergence Ratio = 0.346548362845582
Baum welch starting for iteration: 3 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 3
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 24.830492845787
Convergence Ratio = 0.407562150494998
Baum welch starting for iteration: 4 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
the log file for details.
Normalization for iteration: 4
WARNING: This step had 0 ERROR messages and 6 WARNING messages. Please check
the log file for details.
Current Overall Likelihood Per Frame = 24.8373608903021
Training completed after 4 iterations
MODULE: 40 Build Trees
Phase 1: Cleaning up old log files...
Phase 2: Make Questions
Phase 3: Tree building
Processing each phone with each state
AE 0
AE 1
AE 2
AH 0
AH 1
AH 2
AO 0
AO 1
AO 2
AY 0
AY 1
AY 2
B 0
B 1
B 2
D 0
D 1
D 2
DH 0
DH 1
DH 2
ER 0
ER 1
ER 2
EY 0
EY 1
EY 2
G 0
G 1
G 2
HH 0
FATAL_ERROR: "main.c", line 771: Initialization failed
WARNING: This step had 0 ERROR messages and 1 WARNING messages. Please check
the log file for details.
IH 0
IH 1
IH 2
K 0
K 1
K 2
L 0
L 1
L 2
M 0
M 1
M 2
N 0
N 1
N 2
NG 0
NG 1
NG 2
R 0
R 1
R 2
S 0
S 1
S 2
SH 0
SH 1
SH 2
Skipping SIL
T 0
T 1
T 2
UH 0
FATAL_ERROR: "main.c", line 771: Initialization failed
WARNING: This step had 0 ERROR messages and 1 WARNING messages. Please check
the log file for details.
UW 0
UW 1
UW 2
W 0
W 1
W 2
Y 0
Y 1
Y 2
Z 0
Z 1
Z 2
MODULE: 45 Prune Trees
Phase 1: Tree Pruning
FATAL: "main.c", line 167: Unable to open
/usr/basit/sphinx/tutorial/testdb/trees/testdb.unpruned/HH-0.dtree for
reading; No such file or directory
MODULE: 50 Training Context dependent models
Phase 1: Cleaning up directories:
accumulator...logs...qmanager...
Phase 2: Copy CI to CD initialize
Phase 3: Forward-Backward
Baum welch starting for 1 Gaussian(s), iteration: 1 (1 of 1)
0% FATAL_ERROR: "main.c", line 1054: initialization failed

Failed to start bw
Only 0 parts of 1 of Baum Welch were successfully completed
Parts 1 failed to run!
Training failed in iteration 1
Something failed:
(/usr/basit/sphinx/tutorial/testdb/scripts_pl/50.cd_hmm_tied/slave_convg.pl)
You have new mail in /var/spool/mail/root

Why it is failing every time:(

I am using the version mentioned on the tutorial site

here is the list of the files

i have a directory named tutorial(/usr/basit/sphinx/tutorial), in which there
is a folder named testdb(/usr/basit/sphinx/tutorial/testdb) in which two
folders are present etc(/usr/basit/sphinx/tutorial/testdb/etc) and
wav(/usr/basit/sphinx/tutorial/testdb/wav) .
My etc folder right now contains these files

feat.params
sphinx_decode.cfg
sphinx_train.cfg
testdb.dic
testdb.filler
testdb.lm
testdb.lm.DMP
testdb.phone
testdb_test.fileids
testdb_test.transcription
testdb_train.fileids
testdb_train.transcription

Here are my modified files

A AH
A(2) EY
AN AE N
AN(2) AH N
ANIMAL AE N AH M AH L
BASIT B AE S AH T
CAT K AE T
DOG D AO G
ENGLISH IH NG G L IH SH
ENGLISH(2) IH NG L IH SH
IS IH Z
MY M AY
NAME N EY M
SONGS S AO NG Z
THIS DH IH S
URDU ER D UW
WHAT W AH T
WHAT(2) HH W AH T
YOUR Y AO R
YOUR(2) Y UH R

AE
AH
AO
AY
B
D
DH
ER
EY
G
HH
IH
K
L
M
N
NG
R
S
SH
SIL
T
UH
UW
W
Y
Z

MY NAME IS BASIT (basit/myname)
WHAT IS YOUR NAME (basit/yourname)
THIS IS A CAT (basit/cat)
URDU (basit/urdu)
ENGLISH (basit/english)
SONGS (basit/songs)
DOG IS AN ANIMAL (basit/dog)

Now what?

[Nasir Hussain]

Hello Basit,

Now do 1 thing, Pack your testdb into a zip file and upload it to some site
like mediafire or rapidshare...After uploading. paste the link here... I will
check out your files by myself. Ab paani sir ke upar chahla gaya hai :P...And
paste link fast :)

-Nasir

[Basit Mahmood]

what's your email id. Give me your id and i send you zip file immediately

[Nasir Hussain]

nasir_hussain20@rediffmail.com

Jaldi bhej m fired up :)

[Basit Mahmood]

I think it is better to upload because of large size:)

[Nasir Hussain]

Accha you can just zip two folders (etc and wav folder ) and send it to
me...Its less in size

-Nasir

[Basit Mahmood]

I have send you etc and wav folder. Please check

Thanks

[Nasir Hussain]

Hello Basit,

I successfully creted the acoustic model from your data.:)

The problem was that,As you had a very very small data.The system was unable
to take phones from the secondary similar words like for example

WHAT    W AH T
WHAT(2)    HH W AH T
YOUR    Y AO R
YOUR(2)    Y UH R

The system was not able to recognise the secondary same pronunciation and
hence it gave error that phones didnt occur...:)
What you need to do is just do it vice versa in the dict file like this

WHAT(2)    W AH T
WHAT    HH W AH T
YOUR(2)    Y AO R
YOUR    Y UH R

And then successfully create your model...:)

Tell me when u successfully create it :)

-Nasir

[Basit Mahmood]

HI,
Thanks, I think this time the model is successfully created. This time it has
no error

the stack trace are

Y 2
Z 0
Z 1
Z 2
MODULE: 45 Prune Trees
Phase 1: Tree Pruning
WARNING: This step had 0 ERROR messages and 1 WARNING messages. Please check
te log file for details.
Phase 2: State Tying
MODULE: 50 Training Context dependent models
Phase 1: Cleaning up directories:
accumulator...logs...qmanager...
Phase 2: Copy CI to CD initialize
Phase 3: Forward-Backward
Baum welch starting for 1 Gaussian(s), iteration: 1 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
te log file for details.
Normalization for iteration: 1
This step had 958 ERROR messages and 0 WARNING messages. Please check the log
ile for details.
Current Overall Likelihood Per Frame = 13.6681558028617
Baum welch starting for 1 Gaussian(s), iteration: 2 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 2
Current Overall Likelihood Per Frame = 24.8309300476948
Convergence Ratio = 0.81669937084679
Baum welch starting for 1 Gaussian(s), iteration: 3 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 3
Current Overall Likelihood Per Frame = 24.8991573926868
Split Gaussians, increase by 1
Current Overall Likelihood Per Frame = 24.8991573926868
Convergence Ratio = 0.00274767577617727
Baum welch starting for 2 Gaussian(s), iteration: 1 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
te log file for details.
Normalization for iteration: 1
This step had 2704 ERROR messages and 0 WARNING messages. Please check the
logfile for details.
Current Overall Likelihood Per Frame = 24.5129014308426
Baum welch starting for 2 Gaussian(s), iteration: 2 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 2
Current Overall Likelihood Per Frame = 27.9089904610493
Convergence Ratio = 0.138542923602412
Baum welch starting for 2 Gaussian(s), iteration: 3 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 3
Current Overall Likelihood Per Frame = 31.3251112877583
Convergence Ratio = 0.122402163972097
Baum welch starting for 2 Gaussian(s), iteration: 4 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 4
Current Overall Likelihood Per Frame = 34.4044356120827
Convergence Ratio = 0.0983021032563021
Baum welch starting for 2 Gaussian(s), iteration: 5 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 5
Current Overall Likelihood Per Frame = 34.733799682035
Split Gaussians, increase by 2
Current Overall Likelihood Per Frame = 34.733799682035
Convergence Ratio = 0.00957330251441781
Baum welch starting for 4 Gaussian(s), iteration: 1 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 8 WARNING messages. Please check
te log file for details.
Normalization for iteration: 1
This step had 8286 ERROR messages and 0 WARNING messages. Please check the
logfile for details.
Current Overall Likelihood Per Frame = 34.3978378378378
Baum welch starting for 4 Gaussian(s), iteration: 2 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 2
Current Overall Likelihood Per Frame = 37.8346422893482
Convergence Ratio = 0.0999133860596864
Baum welch starting for 4 Gaussian(s), iteration: 3 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 3
Current Overall Likelihood Per Frame = 42.6366693163752
Convergence Ratio = 0.126921433280709
Baum welch starting for 4 Gaussian(s), iteration: 4 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 4
Current Overall Likelihood Per Frame = 49.5984181240064
Convergence Ratio = 0.163280784340192
Baum welch starting for 4 Gaussian(s), iteration: 5 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 5
Current Overall Likelihood Per Frame = 50.3082988871224
Split Gaussians, increase by 4
Current Overall Likelihood Per Frame = 50.3082988871224
Convergence Ratio = 0.0143125686254985
Baum welch starting for 8 Gaussian(s), iteration: 1 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 1
This step had 25040 ERROR messages and 0 WARNING messages. Please check the lo
file for details.
Current Overall Likelihood Per Frame = 49.9876788553259
Baum welch starting for 8 Gaussian(s), iteration: 2 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 2
Current Overall Likelihood Per Frame = 54.1723290937997
Convergence Ratio = 0.0837136337253219
Baum welch starting for 8 Gaussian(s), iteration: 3 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 3
Current Overall Likelihood Per Frame = 59.8880445151033
Convergence Ratio = 0.105509870388014
Baum welch starting for 8 Gaussian(s), iteration: 4 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 4
Current Overall Likelihood Per Frame = 68.4884419713832
Convergence Ratio = 0.143607919175102
Baum welch starting for 8 Gaussian(s), iteration: 5 (1 of 1)
0% 100%
WARNING: This step had 0 ERROR messages and 7 WARNING messages. Please check
te log file for details.
Normalization for iteration: 5
Current Overall Likelihood Per Frame = 69.3101589825119
Split Gaussians, increase by 0
Training for 8 Gaussian(s) completed after 5 iterations
MODULE: 90 deleted interpolation
Skipped for continuous models
MODULE: 99 Convert to Sphinx2 format models
Can not create models used by Sphinx-II.
If you intend to create models to use with Sphinx-II models, please rerun wth:
$ST::CFG_HMM_TYPE = '.semi.' or
$ST::CFG_HMM_TYPE = '.cont' and $ST::CFG_FEATURE = '1s_12c_12d_3p_12dd' and
$ST:CFG_STATESPERHMM = '5'

I want to ask one thing it means

WHAT(2) W AH T

has higher precedence than

WHAT W AH T

Means if secondary phone is present then trainer look for the secondary
phone first (WHAT(2))

By the way i am not getting what are the secondary similar words and why the
system not able to recognise the secondary same pronunciation . Can you please
explain it to me.

What is the difference between WHAT and WHAT(2)

One thing more that the all error story was happening because of the very very
small data?

Thanks

[Nasir Hussain]

Hello Basit,

See WHAT and WHAT(2) are same words but with different pronunciation.It
totally depends upon you that you want to give as many pronunciation for each
word... Here you got two pronunciation as per the US english is concerned(as
you created it with lmtools)

And there is no precedence and stuff... The Main problem was that your Data
was very very small.... Else it would have been created successfully with out
giving any error.Just try to make test models with about minimum of 500 words
atleast :)
Its Just a suggestion :)

One thing more that the all error story was happening because of the very
very small data?

Somewhat Yes...:)

For starters i would recommend you to have only 1 pronunciation for each
word....until you get huge database :)
Now you Can Add BANDAR and KABUTER too :D
Try experimenting it :)

-Nasir

[Basit Mahmood]

Tutorial says after creating model go to scripts_pl directory and run all .pl
file. But when i execute command. i get this error

cd scripts_pl/

You have new mail in /var/spool/mail/root

cd 00.verify/

ls

verify_all.pl

verify_all.pl

-bash: verify_all.pl: command not found

perl verify_all.pl

Configuration (e.g. etc/sphinx_train.cfg) not defined
Compilation failed in require at verify_all.pl line 47.
BEGIN failed--compilation aborted at verify_all.pl line 47.
You have new mail in /var/spool/mail/root

Now what is the problem ?

Thanks

[Nasir Hussain]

Hello Basit,

You dont need to go inside script.pl
What you do is
Go to the Terminal and change your directory to the folder that you created
through sphinxtrian("testdb" in your case)
And type
[quoteperl $SPHINXTRAINDIR/scripts_pl/Runall.pl

Than all the scripts will run successfully and you will get the desired Model
:)

-Nasir