Comment to 6:
Requirements say you have to download the german audio files. If you follow the link you will see a huge list of recorded voices/sentences but you don't know which you have to download. For testing I only downloaded files from one user and configured later only to train with these.
In the script copy_and_convert_audio.sh: Many user, me included, can't use the script until we change the line
The last task in the list is training. But what to do next?