/home/bsnayak/tutorial8/an4

Training an4 on ubuntu

MODULE: 000 Computing feature from audio files (2014-02-02 16:02)

Extracting features from segments starting at (part 1 of 1)

sphinx_fe Log File

Extracting features from segments starting at (part 1 of 1)

sphinx_fe Log File

Feature extraction is done

MODULE: 00 verify training files (2014-02-02 16:02)

Phase 1: Checking to see if the dict and filler dict agrees with the phonelist file.

Found 40 words using 38 phones

Phase 2: Checking to make sure there are not duplicate entries in the dictionary

Phase 3: Check general format for the fileids file; utterance length (must be positive); files exist

Phase 4: Checking number of lines in the transcript file should match lines in fileids file

Phase 5: Determine amount of training data, see if n_tied_states seems reasonable.

Estimated Total Hours Training: 0.047625

This is a small amount of data, no comment at this time

Phase 6: Checking that all the words in the transcript are in the dictionary

Words in dictionary: 37

Words in filler dictionary: 3

Phase 7: Checking that all the phones in the transcript are in the phonelist, and all phones in the phonelist appear at least once

MODULE: 0000 train grapheme-to-phoneme model (2014-02-02 16:02)

Skipped (set $CFG_G2P_MODEL = 'yes' to enable)

Feature type is s2_4x which is 4 streams

LDA/MLLT only has sense for single stream features, for example 1s_c_d_dd

Skipping LDA training

Feature type is s2_4x which is 4 streams

LDA/MLLT only has sense for single stream features, for example 1s_c_d_dd

Skipping MLLT training

MODULE: 05 Vector Quantization (2014-02-02 16:02)

agg_seg Log File

kmeans_init Log File

MODULE: 10 Training Context Independent models for forced alignment and VTLN (2014-02-02 16:02)

Skipped: $ST::CFG_FORCEDALIGN set to 'no' in sphinx_train.cfg

Skipped: $ST::CFG_VTLN set to 'no' in sphinx_train.cfg

MODULE: 11 Force-aligning transcripts (2014-02-02 16:02)

Skipped: $ST::CFG_FORCEDALIGN set to 'no' in sphinx_train.cfg

MODULE: 12 Force-aligning data for VTLN (2014-02-02 16:02)

Skipped: $ST::CFG_VTLN set to 'no' in sphinx_train.cfg

MODULE: 20 Training Context Independent models (2014-02-02 16:02)

Phase 1: Cleaning up directories:

Phase 2: Flat initialize

mk_mdef_gen Log File

mk_flat Log File

init_mixw Log File

Phase 3: Forward-Backward

Baum welch starting for 256 Gaussian(s), iteration: 1 (1 of 1)

bw Log File

Normalization for iteration: 1

norm Log File

Current Overall Likelihood Per Frame = -207.190667833188

Baum welch starting for 256 Gaussian(s), iteration: 2 (1 of 1)

bw Log File

Normalization for iteration: 2

norm Log File

Current Overall Likelihood Per Frame = -199.102653834937

Convergence Ratio = 8.08801399825072

Baum welch starting for 256 Gaussian(s), iteration: 3 (1 of 1)

bw Log File

Normalization for iteration: 3

norm Log File

Current Overall Likelihood Per Frame = -196.116652085156

Convergence Ratio = 2.98600174978097

Baum welch starting for 256 Gaussian(s), iteration: 4 (1 of 1)

bw Log File

Normalization for iteration: 4

norm Log File

Current Overall Likelihood Per Frame = -194.565529308836

Convergence Ratio = 1.5511227763196

Baum welch starting for 256 Gaussian(s), iteration: 5 (1 of 1)

bw Log File

Normalization for iteration: 5

norm Log File

Current Overall Likelihood Per Frame = -193.827879848352

Convergence Ratio = 0.737649460483709

Baum welch starting for 256 Gaussian(s), iteration: 6 (1 of 1)

bw Log File

Normalization for iteration: 6

norm Log File

Current Overall Likelihood Per Frame = -193.391834354039

Convergence Ratio = 0.436045494312935

Baum welch starting for 256 Gaussian(s), iteration: 7 (1 of 1)

bw Log File

Normalization for iteration: 7

norm Log File

Current Overall Likelihood Per Frame = -193.098512685914

Convergence Ratio = 0.293321668124719

Baum welch starting for 256 Gaussian(s), iteration: 8 (1 of 1)

bw Log File

Normalization for iteration: 8

norm Log File

Current Overall Likelihood Per Frame = -192.884456109653

Convergence Ratio = 0.214056576261015

Baum welch starting for 256 Gaussian(s), iteration: 9 (1 of 1)

bw Log File

Normalization for iteration: 9

norm Log File

Current Overall Likelihood Per Frame = -192.721084864392

Convergence Ratio = 0.163371245261061

Baum welch starting for 256 Gaussian(s), iteration: 10 (1 of 1)

bw Log File

Normalization for iteration: 10

norm Log File

Maximum desired iterations 10 performed. Terminating CI training

Training completed after 10 iterations

MODULE: 30 Training Context Dependent models (2014-02-02 16:03)

Phase 1: Cleaning up directories:

Phase 2: Initialization

mk_mdef_gen Log File

init_mixw Log File

Phase 3: Forward-Backward

Baum welch starting for iteration: 1 (1 of 1)

bw Log File

Normalization for iteration: 1

norm Log File

Current Overall Likelihood Per Frame = -197.948556430446

Baum welch starting for iteration: 2 (1 of 1)

bw Log File

Normalization for iteration: 2

norm Log File

Current Overall Likelihood Per Frame = -194.863925342666

Convergence Ratio = 3.08463108778048

Baum welch starting for iteration: 3 (1 of 1)

bw Log File

Normalization for iteration: 3

norm Log File

Current Overall Likelihood Per Frame = -192.28393117527

Convergence Ratio = 2.57999416739622

Baum welch starting for iteration: 4 (1 of 1)

bw Log File

Normalization for iteration: 4

norm Log File

Current Overall Likelihood Per Frame = -191.700379119277

Convergence Ratio = 0.583552055993238

Baum welch starting for iteration: 5 (1 of 1)

bw Log File

Normalization for iteration: 5

norm Log File

Current Overall Likelihood Per Frame = -191.359463400408

Convergence Ratio = 0.3409157188687

Baum welch starting for iteration: 6 (1 of 1)

bw Log File

Normalization for iteration: 6

norm Log File

Current Overall Likelihood Per Frame = -191.104403616215

Convergence Ratio = 0.25505978419335

Baum welch starting for iteration: 7 (1 of 1)

bw Log File

Normalization for iteration: 7

norm Log File

Current Overall Likelihood Per Frame = -191.007757363663

Training completed after 7 iterations

MODULE: 40 Build Trees (2014-02-02 16:03)

Phase 1: Cleaning up old log files...

Phase 2: Make Questions

make_quests Log File

Phase 3: Tree building

Processing each phone with each state

A 0

bldtree Log File

A 1

bldtree Log File

A 2

bldtree Log File

AA: 0

bldtree Log File

AA: 1

bldtree Log File

AA: 2

bldtree Log File

AE 0

bldtree Log File

AE 1

bldtree Log File

AE 2

bldtree Log File

AI 0

bldtree Log File

AI 1

bldtree Log File

AI 2

bldtree Log File

AU 0

bldtree Log File

AU 1

bldtree Log File

AU 2

bldtree Log File

B 0

bldtree Log File

B 1

bldtree Log File

B 2

bldtree Log File

CC 0

bldtree Log File

CC 1

bldtree Log File

CC 2

bldtree Log File

D 0

bldtree Log File

D 1

bldtree Log File

D 2

bldtree Log File

E 0

bldtree Log File

E 1

bldtree Log File

E 2

bldtree Log File

EE 0

bldtree Log File

EE 1

bldtree Log File

EE 2

bldtree Log File

EE: 0

bldtree Log File

EE: 1

bldtree Log File

EE: 2

bldtree Log File

EI 0

bldtree Log File

EI 1

bldtree Log File

EI 2

bldtree Log File

F 0

bldtree Log File

F 1

bldtree Log File

F 2

bldtree Log File

G 0

bldtree Log File

G 1

bldtree Log File

G 2

bldtree Log File

H 0

bldtree Log File

H 1

bldtree Log File

H 2

bldtree Log File

I: 0

bldtree Log File

I: 1

bldtree Log File

I: 2

bldtree Log File

II 0

bldtree Log File

II 1

bldtree Log File

II 2

bldtree Log File

K 0

bldtree Log File

K 1

bldtree Log File

K 2

bldtree Log File

L 0

bldtree Log File

L 1

bldtree Log File

L 2

bldtree Log File

M 0

bldtree Log File

M 1

bldtree Log File

M 2

bldtree Log File

N 0

bldtree Log File

N 1

bldtree Log File

N 2

bldtree Log File

NN 0

bldtree Log File

NN 1

bldtree Log File

NN 2

bldtree Log File

OO 0

bldtree Log File

OO 1

bldtree Log File

OO 2

bldtree Log File

OY 0

bldtree Log File

OY 1

bldtree Log File

OY 2

bldtree Log File

P 0

bldtree Log File

P 1

bldtree Log File

P 2

bldtree Log File

QQ 0

bldtree Log File

QQ 1

bldtree Log File

QQ 2

bldtree Log File

R 0

bldtree Log File

R 1

bldtree Log File

R 2

bldtree Log File

S 0

bldtree Log File

S 1

bldtree Log File

S 2

bldtree Log File

SE 0

bldtree Log File

SE 1

bldtree Log File

SE 2

bldtree Log File

Skipping SIL

SS 0

bldtree Log File

SS 1

bldtree Log File

SS 2

bldtree Log File

T 0

bldtree Log File

T 1

bldtree Log File

T 2

bldtree Log File

U 0

bldtree Log File

U 1

bldtree Log File

U 2

bldtree Log File

UU 0

bldtree Log File

UU 1

bldtree Log File

UU 2

bldtree Log File

V 0

bldtree Log File

V 1

bldtree Log File

V 2

bldtree Log File

X 0

bldtree Log File

X 1

bldtree Log File

X 2

bldtree Log File

YY 0

bldtree Log File

YY 1

bldtree Log File

YY 2

bldtree Log File

Z 0

bldtree Log File

Z 1

bldtree Log File

Z 2

bldtree Log File

MODULE: 45 Prune Trees (2014-02-02 16:03)

mk_mdef_gen Log File

Phase 1: Tree Pruning

prunetree Log File

Phase 2: State Tying

tiestate Log File

MODULE: 50 Training Context dependent models (2014-02-02 16:03)

Phase 1: Cleaning up directories:

Phase 2: Copy CI to CD initialize

init_mixw Log File

Phase 3: Forward-Backward

Baum welch starting for 256 Gaussian(s), iteration: 1 (1 of 2)

bw Log File

Baum welch starting for 256 Gaussian(s), iteration: 1 (2 of 2)

bw Log File

Normalization for iteration: 1

norm Log File

Current Overall Likelihood Per Frame = -192.49752114319

Baum welch starting for 256 Gaussian(s), iteration: 2 (1 of 2)

bw Log File

Baum welch starting for 256 Gaussian(s), iteration: 2 (2 of 2)

bw Log File

Normalization for iteration: 2

norm Log File

Current Overall Likelihood Per Frame = -192.414756488772

mk_s2sendump Log File

Training for 256 Gaussian(s) completed after 2 iterations

MODULE: 60 Lattice Generation (2014-02-02 16:04)

Skipped: $ST::CFG_MMIE set to 'no' in sphinx_train.cfg

MODULE: 61 Lattice Pruning (2014-02-02 16:04)

Skipped: $ST::CFG_MMIE set to 'no' in sphinx_train.cfg

MODULE: 62 Lattice Format Conversion (2014-02-02 16:04)

Skipped: $ST::CFG_MMIE set to 'no' in sphinx_train.cfg

MODULE: 65 MMIE Training (2014-02-02 16:04)

Skipped: $ST::CFG_MMIE set to 'no' in sphinx_train.cfg

MODULE: 90 deleted interpolation (2014-02-02 16:04)

Phase 1: Cleaning up directories: logs...

Phase 2: Doing interpolation...

delint Log File

Phase 3: Dumping senones for PocketSphinx...

mk_s2sendump Log File

MODULE: DECODE Decoding using models previously trained (2014-02-02 16:04)

Decoding 7 segments starting at 0 (part 1 of 1)

pocketsphinx_batch Log File

Aligning results to find error rate

SENTENCE ERROR: 14.3% (1/7) WORD ERROR RATE: 4.5% (1/22)