Qwen 27B works GREAT as a LORE MASTER!

I don't use LLMs to write. Never been an interest of mine, prefer my own voice, my own style. That said, I've always wished I had a second brain to help me analyze certain aspects of my story bible, which can get pretty complex. Local models just haven't been up to the task, and I have no intention of letting closed models train on my original ideas. I've been super pleased with Qwen 27B for long context analysis, so I thought I'd give it a try with one of my dense story bibles. So I fed it a concept-dense 80K token document and asked it for some analysis.