Click here to register.

Speech Recognition Engines

Flat
Re: sphinx 3 Forward-Backward error
User: Visitor
Date: 5/20/2010 4:23 am
Views: 73
Rating: 1

my error are:

MODULE: 20 Training Context Independent models
    Phase 1: Cleaning up directories:
    accumulator...logs...qmanager...models...
    Phase 2: Flat initialize
    Phase 3: Forward-Backward
        Baum welch starting for 1 Gaussian(s), iteration: 1 (1 of 1)
        0%
WARNING: This step had 0 ERROR messages and 1 WARNING messages.  Please check the log file for details.
        Normalization for iteration: 1
WARNING: This step had 0 ERROR messages and 39 WARNING messages.  Please check the log file for details.
        Current Overall Likelihood Per Frame = -6.21991534781008
        Baum welch starting for 1 Gaussian(s), iteration: 2 (1 of 1)
        0%
WARNING: This step had 0 ERROR messages and 1 WARNING messages.  Please check the log file for details.
        Normalization for iteration: 2
WARNING: This step had 0 ERROR messages and 39 WARNING messages.  Please check the log file for details.
        Current Overall Likelihood Per Frame = -2.51032940743467
        Convergence Ratio = 0.596404570310026
        Baum welch starting for 1 Gaussian(s), iteration: 3 (1 of 1)
        0%
This step had 2 ERROR messages and 1 WARNING messages.  Please check the log file for details.
        Normalization for iteration: 3
WARNING: This step had 0 ERROR messages and 45 WARNING messages.  Please check the log file for details.
        Current Overall Likelihood Per Frame = 2.23426168958743
        Convergence Ratio = 1.89002729401582
        Baum welch starting for 1 Gaussian(s), iteration: 4 (1 of 1)
        0%
This step had 2 ERROR messages and 1 WARNING messages.  Please check the log file for details.
        Normalization for iteration: 4
WARNING: This step had 0 ERROR messages and 45 WARNING messages.  Please check the log file for details.
        Current Overall Likelihood Per Frame = 4.60733202357564
        Convergence Ratio = 1.0621272991645
        Baum welch starting for 1 Gaussian(s), iteration: 5 (1 of 1)
        0%
This step had 2 ERROR messages and 1 WARNING messages.  Please check the log file for details.
        Normalization for iteration: 5
WARNING: This step had 0 ERROR messages and 45 WARNING messages.  Please check the log file for details.
        Current Overall Likelihood Per Frame = 5.08089587426326
        Convergence Ratio = 0.102784832580852
        Baum welch starting for 1 Gaussian(s), iteration: 6 (1 of 1)
        0%
This step had 2 ERROR messages and 1 WARNING messages.  Please check the log file for details.
        Normalization for iteration: 6
WARNING: This step had 0 ERROR messages and 45 WARNING messages.  Please check the log file for details.
        Current Overall Likelihood Per Frame = 5.21273870333988
        Training completed after 6 iterations
MODULE: 30 Training Context Dependent models
    Phase 1: Cleaning up directories:
    accumulator...logs...qmanager...
    Phase 2: Initialization
WARNING: This step had 0 ERROR messages and 1 WARNING messages.  Please check the log file for details.
FATAL_ERROR: "main.c", line 334: src mean n_cb (== 159) inconsistent w/ ts2cb mapping 123
Something failed: (/home/tonynguyen/test/test/scripts_pl/30.cd_hmm_untied/slave_convg.pl)

I has change my senones to 500 but not any change

Please help me

Thanks

 

--- (Edited on 5/20/2010 4:24 am [GMT-0500] by Visitor) ---

Re: sphinx 3 Forward-Backward error
User: nsh
Date: 5/20/2010 12:26 pm
Views: 2579
Rating: 2

It means your phone file is not quite correct. It either has duplicated phones or has windows-style newlines or empty lines.

You can easily find the discussions of this problem by searching CMUSphinx forum.

http://sourceforge.net/search/?group_id=1904&type_of_search=forums

--- (Edited on 5/20/2010 21:26 [GMT+0400] by nsh) ---

PreviousNext