when i try to run mmie training from configuration
$CFG_MMIE = "yes";
$CFG_MMIE_TYPE = "best";
Valid values are "rand", "best" or "ci"
sphinx-train complete the following step
MODULE: 60 Lattice Generation
Phase 1: Cleaning up directories ...
log...qmanager...numlat...denlat...
Phase 2: Generating numerator lattices ...
Force alignment starting: (1 of 1)
0%
This step had 7 ERROR messages and 1 WARNING messages. Please check the log
file for details.
Phase 3: Generating denominator lattices ...
Decoding 403 segments starting at 0 (part 1 of 1)
0%
This step had 10 ERROR messages and 4 WARNING messages. Please check the log
file for details.
MODULE: 61 Lattice Pruning
Phase 1: Cleaning up directories ...
logs...qmanager...pruned_denlat...
Phase 2: Posterior probability lattice pruning ...
0%
Only 0 parts of 1 of lattices were successfully pruned
Parts
1
failed to run!
i can not solve this problem thanks;
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Traceback (most recent call last):
File "/home/muneer/tutorial/sphinxtrain/python/cmusphinx/mllt.py", line 139,
in <module>
mllt = m.train()
File "/home/muneer/tutorial/sphinxtrain/python/cmusphinx/mllt.py", line 110,
in train
AA, f, d = fmin_l_bfgs_b(self.objective, A.ravel(), args=A.shape, factr=10)
File "/usr/lib/python2.7/dist-packages/scipy/optimize/lbfgsb.py", line 187, in
fmin_l_bfgs_b
f, g = func_and_grad(x)
File "/usr/lib/python2.7/dist-packages/scipy/optimize/lbfgsb.py", line 138, in
func_and_grad
f, g = func(x, *args)
File "/home/muneer/tutorial/sphinxtrain/python/cmusphinx/mllt.py", line 83, in
objective
lg = lg - float(nj) * dot(dot(inv(diag(cl)), A), C)
File "/usr/lib/pymodules/python2.7/numpy/linalg/linalg.py", line 445, in inv
return wrap(solve(a, identity(a.shape, dtype=a.dtype)))
File "/usr/lib/pymodules/python2.7/numpy/linalg/linalg.py", line 328, in solve
raise LinAlgError, 'Singular matrix'
numpy.linalg.linalg.LinAlgError: Singular matrix
Thu Jan 26 11:45:50 2012</module>
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The main problem for you is you are using older version of sphinx train just
update the sphinxtrain and copy the be.exe file into the bin folder of your
project and run the module again you may get the problem of undefined
reference variable $n_gau but nothing will happen just ignore it
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
hi all
when i try to run mmie training from configuration
$CFG_MMIE = "yes";
$CFG_MMIE_TYPE = "best";
Valid values are "rand", "best" or "ci"
sphinx-train complete the following step
MODULE: 60 Lattice Generation
Phase 1: Cleaning up directories ...
log...qmanager...numlat...denlat...
Phase 2: Generating numerator lattices ...
Force alignment starting: (1 of 1)
0%
This step had 7 ERROR messages and 1 WARNING messages. Please check the log
file for details.
Phase 3: Generating denominator lattices ...
Decoding 403 segments starting at 0 (part 1 of 1)
0%
This step had 10 ERROR messages and 4 WARNING messages. Please check the log
file for details.
MODULE: 61 Lattice Pruning
Phase 1: Cleaning up directories ...
logs...qmanager...pruned_denlat...
Phase 2: Posterior probability lattice pruning ...
0%
Only 0 parts of 1 of lattices were successfully pruned
Parts
1
failed to run!
i can not solve this problem thanks;
You can find detailed description in the logs
Most likely issue is that you are missing python sphinxbase module properly
installed. So pythons script to prune lattices doesn't work.
this is the full log
Traceback (most recent call last):
File "/home/muneer/tutorial/sphinxtrain/python/cmusphinx/mllt.py", line 139,
in <module>
mllt = m.train()
File "/home/muneer/tutorial/sphinxtrain/python/cmusphinx/mllt.py", line 110,
in train
AA, f, d = fmin_l_bfgs_b(self.objective, A.ravel(), args=A.shape, factr=10)
File "/usr/lib/python2.7/dist-packages/scipy/optimize/lbfgsb.py", line 187, in
fmin_l_bfgs_b
f, g = func_and_grad(x)
File "/usr/lib/python2.7/dist-packages/scipy/optimize/lbfgsb.py", line 138, in
func_and_grad
f, g = func(x, *args)
File "/home/muneer/tutorial/sphinxtrain/python/cmusphinx/mllt.py", line 83, in
objective
lg = lg - float(nj) * dot(dot(inv(diag(cl)), A), C)
File "/usr/lib/pymodules/python2.7/numpy/linalg/linalg.py", line 445, in inv
return wrap(solve(a, identity(a.shape, dtype=a.dtype)))
File "/usr/lib/pymodules/python2.7/numpy/linalg/linalg.py", line 328, in solve
raise LinAlgError, 'Singular matrix'
numpy.linalg.linalg.LinAlgError: Singular matrix
Thu Jan 26 11:45:50 2012</module>
This log is about MLLT stage, it's somewhat unrelated to MMIE. Are you sure
that's the only issue you have found?
The main problem for you is you are using older version of sphinx train just
update the sphinxtrain and copy the be.exe file into the bin folder of your
project and run the module again you may get the problem of undefined
reference variable $n_gau but nothing will happen just ignore it