>>73493121LLM LoRAs aren't a thing as far as I'm aware (at least yet), at most you have different finetunes of public models, but no way to simply and easily combine them in a modular way like you can do with SD LoRAs. At most you can make a finetune combining several finetunes if you know how to mess with the stuff, but I don't know how well that works and it'd probably just end up bloating the models without actually making them any better. And besides, those finetunes don't seem to aimed at introducing new concepts or knowledge, but trying to make those public models less dumb and more efficient, because in the end they're still very far behind corporate models. And for corporate models, those are definitely not going to let users tinker with anything fun any time soon, if ever.
tl;dr: just use lorebooks/world infos like it has already been suggested.
