I'm interested in building a LLM trained on the corpus of hn's archive. I don't want to work on this by myself, and I'm wondering if anyone wants to do it with me?