[jmorgan]

I really enjoyed Anrej Kaparthy's llama2.c project (https://github.com/karpathy/llama2.c), which runs through creating and running a miniature Llama2 architecture model from scratch.