After the model has been trained, what happens if I tweak the feat.params file. E.g. change the "-model" parameter. Does this make sense or is the params supposed to exactly match some training settings?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
After the model has been trained, what happens if I tweak the feat.params file. E.g. change the "-model" parameter. Does this make sense or is the params supposed to exactly match some training settings?
model parameter is only used by sphinx4, it will break the model type detection.
overall the parameters should stay as is, otherwise bad things will happen.