This GitHub Repo contains the draft of a new 1-week curriculum on LLM Evals being added to ARENA 3.0. The final content that will go into the exercises is inside the chapter3_llm_evals folder. The rest of the repo are code for replicating various materials needed to create the exercises (you can mostly ignore). When running the notebooks, make sure you set the working directory as ARENA_evals/chapter3_llm_evals.
If you are testing our materials, do the following:
- Clone the repo and make a new branch
- Make a PR to the jupyter notebooks