$CFG_WAVFILES_DIR = "$CFG_BASE_DIR/wav";
$CFG_WAVFILE_EXTENSION = 'wav';
$CFG_WAVFILE_TYPE = 'mswav'; # one of nist, mswav, raw
$CFG_FEATFILES_DIR = "$CFG_BASE_DIR/feat";
$CFG_FEATFILE_EXTENSION = 'mfc';
Feature extraction parameters
$CFG_WAVFILE_SRATE = 16000.0;
$CFG_NUM_FILT = 25; # For wideband speech it's 25, for telephone 8khz reasonable value is 15
$CFG_LO_FILT = 130; # For telephone 8kHz speech value is 200
$CFG_HI_FILT = 6800; # For telephone 8kHz speech value is 3500
$CFG_TRANSFORM = "dct"; # Previously legacy transform is used, but dct is more accurate
$CFG_LIFTER = "22"; # Cepstrum lifter is smoothing to improve recognition
$CFG_VECTOR_LENGTH = 13; # 13 is usually enough
$CFG_MIN_ITERATIONS = 1; # BW Iterate at least this many times
$CFG_MAX_ITERATIONS = 10; # BW Don't iterate more than this, somethings likely wrong.
(none/max) Type of AGC to apply to input files
$CFG_AGC = 'none';
(current/none) Type of cepstral mean subtraction/normalization
to apply to input files
$CFG_CMN = 'current';
(yes/no) Normalize variance of input files to 1.0
$CFG_VARNORM = 'no';
(yes/no) Train full covariance matrices
$CFG_FULLVAR = 'no';
(yes/no) Use diagonals only of full covariance matrices for
Forward-Backward evaluation (recommended if CFG_FULLVAR is yes)
$CFG_DIAGFULL = 'no';
(yes/no) Perform vocal tract length normalization in training. This
will result in a "normalized" model which requires VTLN to be done
$CFG_HMM_TYPE = '.ptm.'; # PocketSphinx (larger data sets)
if (($CFG_HMM_TYPE ne ".semi.")
and ($CFG_HMM_TYPE ne ".ptm.")
and ($CFG_HMM_TYPE ne ".cont.")) {
die "Please choose one CFG_HMM_TYPE out of '.cont.', '.ptm.', or '.semi.', " .
"currently $CFG_HMM_TYPE\n";
}
This configuration is fastest and best for most acoustic models in
PocketSphinx and Sphinx-III. See below for Sphinx-II.
$CFG_STATESPERHMM = 3;
$CFG_SKIPSTATE = 'no';
if ($CFG_HMM_TYPE eq '.semi.') {
$CFG_DIRLABEL = 'semi';
Four stream features for PocketSphinx
$CFG_FEATURE = "s2_4x";
$CFG_NUM_STREAMS = 4;
$CFG_INITIAL_NUM_DENSITIES = 256;
$CFG_FINAL_NUM_DENSITIES = 256;
die "For semi continuous models, the initial and final models have the same density"
if ($CFG_INITIAL_NUM_DENSITIES != $CFG_FINAL_NUM_DENSITIES);
} elsif ($CFG_HMM_TYPE eq '.ptm.') {
$CFG_DIRLABEL = 'ptm';
Four stream features for PocketSphinx
$CFG_FEATURE = "s2_4x";
$CFG_NUM_STREAMS = 4;
$CFG_INITIAL_NUM_DENSITIES = 64;
$CFG_FINAL_NUM_DENSITIES = 64;
die "For phonetically tied models, the initial and final models have the same density"
if ($CFG_INITIAL_NUM_DENSITIES != $CFG_FINAL_NUM_DENSITIES);
} elsif ($CFG_HMM_TYPE eq '.cont.') {
$CFG_DIRLABEL = 'cont';
Single stream features - Sphinx 3
$CFG_FEATURE = "1s_c_d_dd";
$CFG_NUM_STREAMS = 1;
$CFG_INITIAL_NUM_DENSITIES = 1;
$CFG_FINAL_NUM_DENSITIES = 8;
die "The initial has to be less than the final number of densities"
if ($CFG_INITIAL_NUM_DENSITIES > $CFG_FINAL_NUM_DENSITIES);
}
Number of top gaussians to score a frame. A little bit less accurate computations
make training significantly faster. Uncomment to apply this during the training
For good accuracy make sure you are using the same setting in decoder
In theory this can be different for various training stages. For example 4 for
It is polite to add files as an archive in attachment.
I would like to understand why the RFT value is always 0.02
Your model is very small so decoding is quite fast. "Always" is not applicable here since you didn't change any decoding parameters. If you increase the BEAMWIDTH in config file decoding will be slower.
and wherein it is measured RFT. It is in seconds?
real-time factor is a ratio between lenght of audio (in seconds) and decoding time (in seconds also) so it is unitless (dimensionless).
Last edit: Nickolay V. Shmyrev 2015-11-22
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello, I'm decoding my model trained with SphinxTrain and got these results:
INFO: batch.c (777): TOTAL 4648.31 seconds speech, 83.07 seconds CPU, 83.12 seconds wall
INFO: batch.c (779): AVERAGE 0:02 XRT (CPU), 0:02 XRT (elapsed)
INFO: ngram_search_fwdtree.c (432): TOTAL 76.58 fwdtree CPU 0016 XRT
INFO: ngram_search_fwdtree.c (435): TOTAL 76.55 fwdtree wall XRT 0016
INFO: ngram_search_fwdflat.c (176): fwdflat TOTAL 6:37 CPU 0001 XRT
INFO: ngram_search_fwdflat.c (179): fwdflat TOTAL 6:43 wall XRT 0001
INFO: ngram_search.c (303): TOTAL bestpath 0:08 CPU 0000 XRT
INFO: ngram_search.c (306): TOTAL bestpath 00:10 wall XRT 0000
Thu Nov 19 14:59:26 2015
How could I calculate the Real Time Factor (RTF)?
If you correctly copy-paste this line
That 0.02xRT is the real-time factor.
Thanks Nickolay!
I could not differentiate CPU and elapsed.
http://stackoverflow.com/questions/7335920/what-specifically-are-wall-clock-time-user-cpu-time-and-system-cpu-time-in-uni
What is the xRT?
RFT is measured where? Seconds?
Nikolay, I did several treatments, changing the parameters and always gave RFT = 0.02 you tell me why?
Sure, as soon as you provide details on what exactly did you try, share the command line options, data files you are using and so on.
Well..
Here is my file db-1-1.log
Configuration script for sphinx trainer --mode:Perl--
$CFG_VERBOSE = 1; # Determines how much goes to the screen.
These are filled in at configuration time
$CFG_DB_NAME = "db";
Experiment name, will be used to name model files and log files
$CFG_EXPTNAME = "$CFG_DB_NAME";
Directory containing SphinxTrain binaries
$CFG_BASE_DIR = "/home/lorena/astout";
$CFG_SPHINXTRAIN_DIR = "/usr/local/lib/sphinxtrain";
$CFG_BIN_DIR = "/usr/local/libexec/sphinxtrain";
$CFG_SCRIPT_DIR = "/usr/local/lib/sphinxtrain/scripts";
Audio waveform and feature file information
$CFG_WAVFILES_DIR = "$CFG_BASE_DIR/wav";
$CFG_WAVFILE_EXTENSION = 'wav';
$CFG_WAVFILE_TYPE = 'mswav'; # one of nist, mswav, raw
$CFG_FEATFILES_DIR = "$CFG_BASE_DIR/feat";
$CFG_FEATFILE_EXTENSION = 'mfc';
Feature extraction parameters
$CFG_WAVFILE_SRATE = 16000.0;
$CFG_NUM_FILT = 25; # For wideband speech it's 25, for telephone 8khz reasonable value is 15
$CFG_LO_FILT = 130; # For telephone 8kHz speech value is 200
$CFG_HI_FILT = 6800; # For telephone 8kHz speech value is 3500
$CFG_TRANSFORM = "dct"; # Previously legacy transform is used, but dct is more accurate
$CFG_LIFTER = "22"; # Cepstrum lifter is smoothing to improve recognition
$CFG_VECTOR_LENGTH = 13; # 13 is usually enough
$CFG_MIN_ITERATIONS = 1; # BW Iterate at least this many times
$CFG_MAX_ITERATIONS = 10; # BW Don't iterate more than this, somethings likely wrong.
(none/max) Type of AGC to apply to input files
$CFG_AGC = 'none';
(current/none) Type of cepstral mean subtraction/normalization
to apply to input files
$CFG_CMN = 'current';
(yes/no) Normalize variance of input files to 1.0
$CFG_VARNORM = 'no';
(yes/no) Train full covariance matrices
$CFG_FULLVAR = 'no';
(yes/no) Use diagonals only of full covariance matrices for
Forward-Backward evaluation (recommended if CFG_FULLVAR is yes)
$CFG_DIAGFULL = 'no';
(yes/no) Perform vocal tract length normalization in training. This
will result in a "normalized" model which requires VTLN to be done
during decoding as well.
$CFG_VTLN = 'no';
Starting warp factor for VTLN
$CFG_VTLN_START = 0.80;
Ending warp factor for VTLN
$CFG_VTLN_END = 1.40;
Step size of warping factors
$CFG_VTLN_STEP = 0.05;
Directory to write queue manager logs to
$CFG_QMGR_DIR = "$CFG_BASE_DIR/qmanager";
Directory to write training logs to
$CFG_LOG_DIR = "$CFG_BASE_DIR/logdir";
Directory for re-estimation counts
$CFG_BWACCUM_DIR = "$CFG_BASE_DIR/bwaccumdir";
Directory to write model parameter files to
$CFG_MODEL_DIR = "$CFG_BASE_DIR/model_parameters";
Directory containing transcripts and control files for
speaker-adaptive training
$CFG_LIST_DIR = "$CFG_BASE_DIR/etc";
Decoding variables for MMIE training
$CFG_LANGUAGEWEIGHT = "11.5";
$CFG_BEAMWIDTH = "1e-100";
$CFG_WORDBEAM = "1e-80";
$CFG_LANGUAGEMODEL = "$CFG_LIST_DIR/$CFG_DB_NAME.lm.DMP";
$CFG_WORDPENALTY = "0.2";
Lattice pruning variables
$CFG_ABEAM = "1e-50";
$CFG_NBEAM = "1e-10";
$CFG_PRUNED_DENLAT_DIR = "$CFG_BASE_DIR/pruned_denlat";
MMIE training related variables
$CFG_MMIE = "no";
$CFG_MMIE_MAX_ITERATIONS = 5;
$CFG_LATTICE_DIR = "$CFG_BASE_DIR/lattice";
$CFG_MMIE_TYPE = "rand"; # Valid values are "rand", "best" or "ci"
$CFG_MMIE_CONSTE = "3.0";
$CFG_NUMLAT_DIR = "$CFG_BASE_DIR/numlat";
$CFG_DENLAT_DIR = "$CFG_BASE_DIR/denlat";
Variables used in main training of models
$CFG_DICTIONARY = "$CFG_LIST_DIR/$CFG_DB_NAME.dic";
$CFG_RAWPHONEFILE = "$CFG_LIST_DIR/$CFG_DB_NAME.phone";
$CFG_FILLERDICT = "$CFG_LIST_DIR/$CFG_DB_NAME.filler";
$CFG_LISTOFFILES = "$CFG_LIST_DIR/${CFG_DB_NAME}_train.fileids";
$CFG_TRANSCRIPTFILE = "$CFG_LIST_DIR/${CFG_DB_NAME}_train.transcription";
$CFG_FEATPARAMS = "$CFG_LIST_DIR/feat.params";
Variables used in characterizing models
$CFG_HMM_TYPE = '.cont.'; # Sphinx 4, PocketSphinx
$CFG_HMM_TYPE = '.semi.'; # PocketSphinx
$CFG_HMM_TYPE = '.ptm.'; # PocketSphinx (larger data sets)
if (($CFG_HMM_TYPE ne ".semi.")
and ($CFG_HMM_TYPE ne ".ptm.")
and ($CFG_HMM_TYPE ne ".cont.")) {
die "Please choose one CFG_HMM_TYPE out of '.cont.', '.ptm.', or '.semi.', " .
"currently $CFG_HMM_TYPE\n";
}
This configuration is fastest and best for most acoustic models in
PocketSphinx and Sphinx-III. See below for Sphinx-II.
$CFG_STATESPERHMM = 3;
$CFG_SKIPSTATE = 'no';
if ($CFG_HMM_TYPE eq '.semi.') {
$CFG_DIRLABEL = 'semi';
Four stream features for PocketSphinx
$CFG_FEATURE = "s2_4x";
$CFG_NUM_STREAMS = 4;
$CFG_INITIAL_NUM_DENSITIES = 256;
$CFG_FINAL_NUM_DENSITIES = 256;
die "For semi continuous models, the initial and final models have the same density"
if ($CFG_INITIAL_NUM_DENSITIES != $CFG_FINAL_NUM_DENSITIES);
} elsif ($CFG_HMM_TYPE eq '.ptm.') {
$CFG_DIRLABEL = 'ptm';
Four stream features for PocketSphinx
$CFG_FEATURE = "s2_4x";
$CFG_NUM_STREAMS = 4;
$CFG_INITIAL_NUM_DENSITIES = 64;
$CFG_FINAL_NUM_DENSITIES = 64;
die "For phonetically tied models, the initial and final models have the same density"
if ($CFG_INITIAL_NUM_DENSITIES != $CFG_FINAL_NUM_DENSITIES);
} elsif ($CFG_HMM_TYPE eq '.cont.') {
$CFG_DIRLABEL = 'cont';
Single stream features - Sphinx 3
$CFG_FEATURE = "1s_c_d_dd";
$CFG_NUM_STREAMS = 1;
$CFG_INITIAL_NUM_DENSITIES = 1;
$CFG_FINAL_NUM_DENSITIES = 8;
die "The initial has to be less than the final number of densities"
if ($CFG_INITIAL_NUM_DENSITIES > $CFG_FINAL_NUM_DENSITIES);
}
Number of top gaussians to score a frame. A little bit less accurate computations
make training significantly faster. Uncomment to apply this during the training
For good accuracy make sure you are using the same setting in decoder
In theory this can be different for various training stages. For example 4 for
CI stage and 16 for CD stage
$CFG_CI_TOPN = 4;
$CFG_CD_TOPN = 16;
(yes/no) Train multiple-gaussian context-independent models (useful
for alignment, use 'no' otherwise) in the models created
specifically for forced alignment
$CFG_FALIGN_CI_MGAU = 'no';
(yes/no) Train multiple-gaussian context-independent models (useful
for alignment, use 'no' otherwise)
$CFG_CI_MGAU = 'no';
(yes/no) Train context-dependent models
$CFG_CD_TRAIN = 'yes';
Number of tied states (senones) to create in decision-tree clustering
$CFG_N_TIED_STATES = 200;
How many parts to run Forward-Backward estimatinon in
$CFG_NPART = 1;
(yes/no) Train a single decision tree for all phones (actually one
per state) (useful for grapheme-based models, use 'no' otherwise)
$CFG_CROSS_PHONE_TREES = 'no';
Use force-aligned transcripts (if available) as input to training
$CFG_FORCEDALIGN = 'no';
Use a specific set of models for force alignment. If not defined,
context-independent models for the current experiment will be used.
$CFG_FORCE_ALIGN_MODELDIR = "$CFG_MODEL_DIR/$CFG_EXPTNAME.falign_ci_$CFG_DIRLABEL";
Use a specific dictionary and filler dictionary for force alignment.
If these are not defined, a dictionary and filler dictionary will be
created from $CFG_DICTIONARY and $CFG_FILLERDICT, with noise words
removed from the filler dictionary and added to the dictionary (this
is because the force alignment is not very good at inserting them)
$CFG_FORCE_ALIGN_DICTIONARY = "$ST::CFG_BASE_DIR/falignout$ST::CFG_EXPTNAME.falign.dict";;
$CFG_FORCE_ALIGN_FILLERDICT = "$ST::CFG_BASE_DIR/falignout/$ST::CFG_EXPTNAME.falign.fdict";;
Use a particular beam width for force alignment. The wider
(i.e. smaller numerically) the beam, the fewer sentences will be
rejected for bad alignment.
$CFG_FORCE_ALIGN_BEAM = 1e-60;
Calculate an LDA/MLLT transform?
$CFG_LDA_MLLT = 'no';
Dimensionality of LDA/MLLT output
$CFG_LDA_DIMENSION = 29;
This is actually just a difference in log space (it doesn't make
sense otherwise, because different feature parameters have very
different likelihoods)
$CFG_CONVERGENCE_RATIO = 0.1;
Queue::POSIX for multiple CPUs on a local machine
Queue::PBS to use a PBS/TORQUE queue
$CFG_QUEUE_TYPE = "Queue";
Name of queue to use for PBS/TORQUE
$CFG_QUEUE_NAME = "workq";
(yes/no) Build questions for decision tree clustering automatically
$CFG_MAKE_QUESTS = "yes";
If CFG_MAKE_QUESTS is yes, questions are written to this file.
If CFG_MAKE_QUESTS is no, questions are read from this file.
$CFG_QUESTION_SET = "${CFG_BASE_DIR}/model_architecture/${CFG_EXPTNAME}.tree_questions";
$CFG_QUESTION_SET = "${CFG_BASE_DIR}/linguistic_questions";
$CFG_CP_OPERATION = "${CFG_BASE_DIR}/model_architecture/${CFG_EXPTNAME}.cpmeanvar";
Configuration for grapheme-to-phoneme model
$CFG_G2P_MODEL= 'no';
Configuration script for sphinx decoder
Variables starting with $DEC_CFG_ refer to decoder specific
arguments, those starting with $CFG_ refer to trainer arguments,
some of them also used by the decoder.
$DEC_CFG_VERBOSE = 1; # Determines how much goes to the screen.
These are filled in at configuration time
Name of the decoding script to use (psdecode.pl or s3decode.pl, probably)
$DEC_CFG_SCRIPT = 'psdecode.pl';
$DEC_CFG_EXPTNAME = "$CFG_EXPTNAME";
$DEC_CFG_JOBNAME = "$CFG_EXPTNAME"."_job";
Models to use.
$DEC_CFG_MODEL_NAME = "$CFG_EXPTNAME.cd_${CFG_DIRLABEL}_${CFG_N_TIED_STATES}";
$DEC_CFG_FEATFILES_DIR = "$CFG_BASE_DIR/feat";
$DEC_CFG_FEATFILE_EXTENSION = '.mfc';
$DEC_CFG_AGC = $CFG_AGC;
$DEC_CFG_CMN = $CFG_CMN;
$DEC_CFG_VARNORM = $CFG_VARNORM;
$DEC_CFG_QMGR_DIR = "$CFG_BASE_DIR/qmanager";
$DEC_CFG_LOG_DIR = "$CFG_BASE_DIR/logdir";
$DEC_CFG_MODEL_DIR = "$CFG_MODEL_DIR";
$DEC_CFG_DICTIONARY = "$CFG_BASE_DIR/etc/$CFG_DB_NAME.dic";
$DEC_CFG_FILLERDICT = "$CFG_BASE_DIR/etc/$CFG_DB_NAME.filler";
$DEC_CFG_LISTOFFILES = "$CFG_BASE_DIR/etc/${CFG_DB_NAME}_test.fileids";
$DEC_CFG_TRANSCRIPTFILE = "$CFG_BASE_DIR/etc/${CFG_DB_NAME}_test.transcription";
$DEC_CFG_RESULT_DIR = "$CFG_BASE_DIR/result";
$DEC_CFG_PRESULT_DIR = "$CFG_BASE_DIR/presult";
This variables, used by the decoder, have to be user defined, and
may affect the decoder output
$DEC_CFG_LANGUAGEMODEL = "$CFG_BASE_DIR/etc/${CFG_DB_NAME}.lm.DMP";
Or can be JSGF or FSG too, used if uncommented
$DEC_CFG_GRAMMAR = "$CFG_BASE_DIR/etc/${CFG_DB_NAME}.jsgf";
$DEC_CFG_FSG = "$CFG_BASE_DIR/etc/${CFG_DB_NAME}.fsg";
$DEC_CFG_LANGUAGEWEIGHT = "10";
$DEC_CFG_BEAMWIDTH = "1e-80";
$DEC_CFG_WORDBEAM = "1e-40";
$DEC_CFG_ALIGN = "builtin";
$DEC_CFG_NPART = 1; # Define how many pieces to split decode in
This variable has to be defined, otherwise utils.pl will not load.
$CFG_DONE = 1;
return 1;
~~~~~~~~~~~~~~~
I would like to understand why the RFT value is always 0.02 and wherein it is measured RFT. It is in seconds?
Last edit: Nickolay V. Shmyrev 2015-11-22
It is polite to add files as an archive in attachment.
Your model is very small so decoding is quite fast. "Always" is not applicable here since you didn't change any decoding parameters. If you increase the BEAMWIDTH in config file decoding will be slower.
real-time factor is a ratio between lenght of audio (in seconds) and decoding time (in seconds also) so it is unitless (dimensionless).
Last edit: Nickolay V. Shmyrev 2015-11-22