Thank u sir,
>Most probably you've made a mistake on a previous step. Adaptation even with a single class should improve accuracy
yes sir,Now it is working fine after i followed the procedure which u suggested.
Now i would like to work on Multi class MLLR..Plz say how to proceed....
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
-cb2mllrfn it is one of the argument in BW,MLLLR_SOLVE and Many ....
it s default value is '1cls.'
if i want to change the value to multiclass....
how to generate regression class...
whether it is possible to generate using
mk_mllr_class
if so ,how?.....
> how to generate regression class
> whether it is possible to generate using mk_mllr_class
mk_mllr_class just converts a text file into a binary format used by other tools. The text file you need to write it yourself.
A side note about your previous post:
> Since I got negative result in adapting small amount of utterences..
Most probably you've made a mistake on a previous step. Adaptation even with a single class should improve accuracy.
Thank u sir,
>Most probably you've made a mistake on a previous step. Adaptation even with a single class should improve accuracy
yes sir,Now it is working fine after i followed the procedure which u suggested.
Now i would like to work on Multi class MLLR..Plz say how to proceed....