All code public
Projects
After a self-paced LLM foundations prelude: an LLM evaluation harness to start, five projects in GPU kernels and kernel evaluation that build on it, and one small platform project that runs the evaluations. Each has a write-up here and its code on GitHub.
Before the projects · self-paced
LLM foundations: reading and code examples
The big picture, then gradient descent, backpropagation and attention built by hand, training in practice, a tiny GPT, inference memory, and evaluating LLMs. Each exercise is in the llm-foundations repo and written up as a foundations post.
Project write-ups will appear here as each one goes live. Follow the RSS feed to catch them.