[Subject Prev][Subject Next][Thread Prev][Thread Next][Date Index][Thread Index]

[hts-users:02568] Error in REGTR step of ARCTIC-ADAPT script


Hi,

I have a problem in training HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT for hts
2.1.1 script

From beggining to the Synthesizing waveform of Speaker Independent model
is working properly.

But REGTR (Building regression-class trees for adaptation) isn't go well.

Here is some log about REGTR procedures


=========================================================================================
Start building regression-class trees for adaptation at 2010. 07. 19.
(월) 13:26:55 KST
=========================================================================================

/home/kmchun/Speech_Synthesis/Tools/HTK3.4_HTS2.1.1/bin/HHEd -A -B -C
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/configs/trn.cnf
-D -T 1 -p -i -H
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/models/qst001/ver1/cmp/re_clustered.mmf
-M
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/models/qst001/ver1/cmp/regTrees
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/edfiles/qst001/ver1/cmp/reg.hed
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/data/lists/full.list


HTK Configuration Parameters[10]
Module/Tool Parameter Value
# MINDUR 5
# MAXSTDDEVCOEF 10
# APPLYDURVARFLOOR TRUE
# DURVARFLOORPERCENTILE 1.000000
# SHRINKOCCTHRESH Vector 5 500.0 100.0 100.0 100.0 100.0
# VFLOORSCALESTR Vector 5 0.01 0.01 0.01 0.01 0.01
# MINLEAFOCC 0
# NATURALWRITEORDER TRUE
# NATURALREADORDER TRUE
# APPLYVFLOOR TRUE

HHEd
27259/27259 Models Loaded [7 states max, 2 mixes max]

// load stats file
LS
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/stats/qst001/ver1/cmp.stats.re-clustered
Loading state occupation stats
Stats loaded for 27259 models
Mean Occupation Count = 14.622191

// construct regression class tree
RC 32 reg
Building regression tree with 32 terminals (5 streams)
Creating regression class tree with ident reg.tree and baseclass reg.base
Splitting Node 1, score 1.000000e+10
(Stream splitting)
Splitting Node 3, score 1.000000e+10
(Stream splitting)
Splitting Node 5, score 1.000000e+10
(Stream splitting)
Splitting Node 7, score 1.000000e+10
(Stream splitting)
Splitting Node 8, score 1.000000e+10
(MSD splitting)
Splitting Node 6, score 1.000000e+10

.
.
.


Splitting Node -887, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -885, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -883, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -881, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -879, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -877, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -875, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -873, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -871, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -869, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -867, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -865, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -863, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -861, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -859, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -857, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -855, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -853, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -851, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -849, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -847, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -845, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -843, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -841, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -839, score 1.178150e+03
(Stream=3, vSize=1)
Splitting Node -837, score 1.178150e+03
(Stream=3, vSize=1)

Error in /home/kmchun/Speech_Synthesis/Tools/HTK3.4_HTS2.1.1/bin/HHEd -A
-B -C
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/configs/trn.cnf
-D -T 1 -p -i -H
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/models/qst001/ver1/cmp/re_clustered.mmf
-M
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/models/qst001/ver1/cmp/regTrees
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/edfiles/qst001/ver1/cmp/reg.hed
/home/kmchun/Speech_Synthesis/hts-demo_2.1.1/HTS-demo_CMU-ARCTIC-ADAPT_STRAIGHT/data/lists/full.list



Most errors in HTK and HTS have Error code. But this time, even the
error code doesn't show in the log,

So I can't figure what is solution for that problem.


Anyone who solved or just encountered similar problem, please share some
clue of the problem.

Thank you for reading.


Regards
KwangMyung Jeon

------------------------------------------------------------------
KwangMyung Jeon
Human Computing Laboratory
School of Information and Communications
Gwangju Institute of Science and Technology

1 Oryong-dong, Buk-gu
Gwangju 500-712, SOUTH KOREA
Tel) +82-62-970-3121  Fax) +82-62-970-2204
Email) kmjeon@xxxxxxxxxx
----------------------------------------------------------------------