Closed Holzhaus closed 7 years ago
Same problem with revision: 6e5329-dirty.
Switching to phonetisaurus-g2p (older revision) and using the fst built with revision 6e5329 do output pronunciations.
Using openfst-1.4.1.
I believe this should be resolved in the current version of master tied to 1.6.1. However you would need to rebuild/retrain your models.
The result from
phonetisaurus-g2pfst
seems to be always empty.I looks like
path_filter.ordered_paths.size()
inPhonetisaurusScript.h
is always0
.