please make sure you have following installed :

tensorflow 1.3
scipy.io.wavfile 
python_speech_features
jupyter notebook


My setup used ubuntu 16.04 , Dell inspiron 15 5559 laptop ( tensorflow CPU )

there is a file called words.txt , in folder named words , modify that file to create your own
vocabulary, but please reserve last line for space words (" ") , otherwise you will get an error

This project is an ongoing effort, and by no means is complete, the purpose is for me to learn state of the art techniques , concering speech recognition, please feel free to make suggestions and give feedback.