When i do a batch decoding for one partiuclar audio file i get this error
ERROR: "ngram_search.c", line 1207: Failed to find DAG node corresponding to
how do i solve it? please help
When i decode that single file i get the decoded sentence
But when i do a batch decoding of many file that respective audio file gives me the error
Last edit: Tania Mendonca 2017-04-19
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
config.set_float('-beam', 1e-80)
config.set_float('-fwdflatbeam', 1e-80)
config.set_float('-fwdflatlw', 10)
config.set_float('-fwdflatwbeam',1e-40)
config.set_float('-lpbeam',1e-80)
config.set_float('-lponlybeam',1e-80)
config.set_float('-lw',10)
config.set_float('-pbeam',1e-80)
config.set_float('-wbeam',1e-40)
config.set_float('-wip',2e-01)
which particular parameter i need to change and by how much value?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
When i do a batch decoding for one partiuclar audio file i get this error
ERROR: "ngram_search.c", line 1207: Failed to find DAG node corresponding to
how do i solve it? please help
When i decode that single file i get the decoded sentence
But when i do a batch decoding of many file that respective audio file gives me the error
Last edit: Tania Mendonca 2017-04-19
Increase beam width.
config.set_float('-beam', 1e-80)
config.set_float('-fwdflatbeam', 1e-80)
config.set_float('-fwdflatlw', 10)
config.set_float('-fwdflatwbeam',1e-40)
config.set_float('-lpbeam',1e-80)
config.set_float('-lponlybeam',1e-80)
config.set_float('-lw',10)
config.set_float('-pbeam',1e-80)
config.set_float('-wbeam',1e-40)
config.set_float('-wip',2e-01)
which particular parameter i need to change and by how much value?
All of them
I tried changing the lw to 6.5 the error doesn't exist now duribg batch decoding. But the accuracy dropped by 2%. Why is the accuracy dropping?