now i'm doing my continuous acoustic model, and i'm doing the baum_welch passage to create my CI models. I have no ERRORS, but some WARNINGS
INFO: main.c(253): Reading /home/aenima1891/tutorial/numeri/model_architecture/numeri.ci.mdef
INFO: model_def_io.c(587): Model definition info:
INFO: model_def_io.c(588): 18 total models defined (18 base, 0 tri)
INFO: model_def_io.c(589): 72 total states
INFO: model_def_io.c(590): 54 total tied states
INFO: model_def_io.c(591): 54 total tied CI states
INFO: model_def_io.c(592): 18 total tied transition matrices
INFO: model_def_io.c(593): 4 max state/model
INFO: model_def_io.c(594): 4 min state/model
INFO: s3mixw_io.c(116): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/mixture_weights [54x1x1 array]
WARNING: "mod_inv.c", line 368: Model inventory n_density not set; setting to value in mixw file, 1.
INFO: s3tmat_io.c(115): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/transition_matrices [18x3x4 array]
INFO: mod_inv.c(296): inserting tprob floor 1.000000e-04 and renormalizing
INFO: s3gau_io.c(166): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/means [57x1x1 array]
INFO: s3gau_io.c(166): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/variances [57x1x1 array]
INFO: gauden.c(183): 57 total mgau
INFO: gauden.c(155): 1 feature streams (|0|=39 )
INFO: gauden.c(194): 1 total densities
INFO: gauden.c(98): min_var=1.000000e-04
WARNING: "mod_inv.c", line 256: n_top 4 > n_density 1. n_top <- 1
INFO: gauden.c(172): compute 1 densities/frame
of codebooks in mean/var files, 57, inconsistent with ts2cb mapping 54
INFO: main.c(361): Will reestimate mixing weights.
INFO: main.c(363): Will reestimate means.
INFO: main.c(365): Will reestimate variances.
INFO: main.c(367): WIll NOT optionally delete silence in Baum Welch or Viterbi.
INFO: main.c(375): Will reestimate transition matrices
INFO: main.c(388): Reading main lexicon: /home/aenima1891/tutorial/numeri/etc/numeri.dic
INFO: lexicon.c(233): 10 entries added from /home/aenima1891/tutorial/numeri/etc/numeri.dic
INFO: main.c(400): Reading filler lexicon: /home/aenima1891/tutorial/numeri/etc/numeri.filler
INFO: lexicon.c(233): 3 entries added from /home/aenima1891/tutorial/numeri/etc/numeri.filler
INFO: main.c(421): Silence Tag SIL
INFO: corpus.c(1343): Will process all remaining utts starting at 0
INFO: main.c(620): Reestimation: Baum-Welch
I have done 6 iterations, and every time there are these 2 warnings.
indeed, in "norm" iterations too there are some "recursive" warnings
WARNING: "gauden.c", line 1604: (mgau= 54, feat= 0, density= 0) never observed
WARNING: "gauden.c", line 1604: (mgau= 55, feat= 0, density= 0) never observed
WARNING: "gauden.c", line 1604: (mgau= 56, feat= 0, density= 0) never observed
but i think that these depend by a lack of training data
thanks to all for any suggestion and clarification
this forum is my lifeline...
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
now i'm doing my continuous acoustic model, and i'm doing the baum_welch passage to create my CI models. I have no ERRORS, but some WARNINGS
INFO: main.c(253): Reading /home/aenima1891/tutorial/numeri/model_architecture/numeri.ci.mdef
INFO: model_def_io.c(587): Model definition info:
INFO: model_def_io.c(588): 18 total models defined (18 base, 0 tri)
INFO: model_def_io.c(589): 72 total states
INFO: model_def_io.c(590): 54 total tied states
INFO: model_def_io.c(591): 54 total tied CI states
INFO: model_def_io.c(592): 18 total tied transition matrices
INFO: model_def_io.c(593): 4 max state/model
INFO: model_def_io.c(594): 4 min state/model
INFO: s3mixw_io.c(116): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/mixture_weights [54x1x1 array]
WARNING: "mod_inv.c", line 368: Model inventory n_density not set; setting to value in mixw file, 1.
INFO: s3tmat_io.c(115): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/transition_matrices [18x3x4 array]
INFO: mod_inv.c(296): inserting tprob floor 1.000000e-04 and renormalizing
INFO: s3gau_io.c(166): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/means [57x1x1 array]
INFO: s3gau_io.c(166): Read /home/aenima1891/tutorial/numeri/model_parameters/numeri.ci_cont_flatinitial/variances [57x1x1 array]
INFO: gauden.c(183): 57 total mgau
INFO: gauden.c(155): 1 feature streams (|0|=39 )
INFO: gauden.c(194): 1 total densities
INFO: gauden.c(98): min_var=1.000000e-04
WARNING: "mod_inv.c", line 256: n_top 4 > n_density 1. n_top <- 1
INFO: gauden.c(172): compute 1 densities/frame
of codebooks in mean/var files, 57, inconsistent with ts2cb mapping 54
INFO: main.c(361): Will reestimate mixing weights.
INFO: main.c(363): Will reestimate means.
INFO: main.c(365): Will reestimate variances.
INFO: main.c(367): WIll NOT optionally delete silence in Baum Welch or Viterbi.
INFO: main.c(375): Will reestimate transition matrices
INFO: main.c(388): Reading main lexicon: /home/aenima1891/tutorial/numeri/etc/numeri.dic
INFO: lexicon.c(233): 10 entries added from /home/aenima1891/tutorial/numeri/etc/numeri.dic
INFO: main.c(400): Reading filler lexicon: /home/aenima1891/tutorial/numeri/etc/numeri.filler
INFO: lexicon.c(233): 3 entries added from /home/aenima1891/tutorial/numeri/etc/numeri.filler
INFO: main.c(421): Silence Tag SIL
INFO: corpus.c(1343): Will process all remaining utts starting at 0
INFO: main.c(620): Reestimation: Baum-Welch
I have done 6 iterations, and every time there are these 2 warnings.
indeed, in "norm" iterations too there are some "recursive" warnings
WARNING: "gauden.c", line 1604: (mgau= 54, feat= 0, density= 0) never observed
WARNING: "gauden.c", line 1604: (mgau= 55, feat= 0, density= 0) never observed
WARNING: "gauden.c", line 1604: (mgau= 56, feat= 0, density= 0) never observed
but i think that these depend by a lack of training data
thanks to all for any suggestion and clarification
this forum is my lifeline...