Convert a large audio wav file to text using cmu sphinx

I need to do a project which converts video to text. I've converted avi files to wav using JAVE encoder/decoder library. I need to convert wav files to text using cmu sphinx. I've tried the transcriber demo and it works fine. But when i give the SpeakerId demo(it contains text instead of numbers like Transcriber), there is no output! How do I give a wav file and convert it to text??

Could you tell me how to construct an acoustic model and train the system? I have cmu sphinx4 alpha.