I was doing fine until Steps 3 and 4

Generated with self hosted ollama llama2-uncensored:7b (the small model since I have a small rig)