ERROR: data/lang/G.fst is empty

[Abhishek Srivastava]

Hi,

I ran into the following error while training the model: ERROR: data/lang/G.fst is empty. I have built Kaldi on my system following the instructions on official Kaldi github repo.

Did anyone else run into this same issue or has any pointers on how to resolve this?

Thanks.

Output with more context

.
.
.
Checking data/lang/oov.{txt, int} ...
--> text seems to be UTF-8 or ASCII, checking whitespaces
--> text contains only allowed whitespaces
--> 1 entry/entries in data/lang/oov.txt
--> data/lang/oov.int corresponds to data/lang/oov.txt
--> data/lang/oov.{txt, int} are OK
--> data/lang/L.fst is olabel sorted
--> data/lang/L_disambig.fst is olabel sorted
--> data/lang/G.fst is ilabel sorted
--> ERROR: data/lang/G.fst is empty
fstdeterminizestar data/lang/G.fst /dev/null 
--> data/lang/G.fst is determinizable
--> utils/lang/check_g_properties.pl successfully validated data/lang/G.fst
--> utils/lang/check_g_properties.pl succeeded.
--> Testing determinizability of L_disambig . G
fstdeterminizestar 
fsttablecompose data/lang/L_disambig.fst data/lang/G.fst 
--> ERROR: fail to determinize L_disambig . G.  Output is:
fst type                                          vector
arc type                                          standard
input symbol table                                none
output symbol table                               none
# of states                                       0
# of arcs                                         0
initial state                                     -1
# of final states                                 0
# of input/output epsilons                        0
# of input epsilons                               0
# of output epsilons                              0
input label multiplicity                          0
output label multiplicity                         0
# of accessible states                            0
# of coaccessible states                          0
# of connected states                             0
# of connected components                         0
# of strongly conn components                     0
input matcher                                     y
output matcher                                    y
input lookahead                                   n
output lookahead                                  n
expanded                                          y
mutable                                           y
error                                             n
acceptor                                          y
input deterministic                               y
output deterministic                              y
input/output epsilons                             n
input epsilons                                    n
output epsilons                                   n
input label sorted                                y
output label sorted                               y
weighted                                          n
cyclic                                            n
cyclic at initial state                           n
top sorted                                        y
accessible                                        y
coaccessible                                      y
string                                            y
weighted cycles                                   n
--> ERROR (see error messages above)
prepare_lang.sh: error validating output
+ exit 1
.
.
.

[Abhishek Srivastava]

Found a solution:

1. Delete data/lang
2. Rerun

https://groups.google.com/g/kaldi-help/c/OYBhUkJFWVA/m/f0j7Z_7mBwAJ