From discussion #725, I will delete all model that may be legal problems and replace with dataset that open source compatible.