bob.kaldi.plad_train(), return NULL plda mode

[captain...@gmail.com]

Hi,

    I am doing speaker recognition by using bob.kaldi. Currently, I plan to train ivector, then train plda, but it's very weird, I always get NULL plad model. 

   

    I have attached code below. Please give me any ideas and suggestions. 

   Thank you.