You need to explain more what is the step you have trouble with. From your description it is not quite clear what problem do you have.

If you have trouble to record there is no need to dictate, you can just take any recording like audiobook or tv show and use it for training.

http://spirit.blau.in/simon/2012/05/15/ralfs-modern-greek-speech-model/

The problem i have is that i cant insert my phonemelist into audacity like in the step 3 of the link. I wondered if someone could give me any idea about that or if i have to follow different method on creating my speech model.

Hey Ioan

I think you need to check a different tutorial, this one

http://cmusphinx.sourceforge.net/wiki/tutorialam

It describes how to prepare the data, how to run the training and how to use the trained model. You take an audio book and split it on sentences in audio editor. You also need to write a transcription for each sentence you created. Everything else is technical.

In case you have any questions you are welcome to ask me, my mail is [email protected]. You are also welcome to ask questions here or on CMUSphinx forum.