Running 140 Natural Language Processing before LLMs ๐ 140 From n-gram counts to attention, learning by doing