This repository is a structured tutorial series organizing LLMs and their applications across nine layers. It covers foundation models (ChatGLM, LLaMA, Alpaca, Vicuna, Baichuan, BLOOMz, MOSS) with implementation guides, parameter-efficient fine-tuning techniques (QLoRA, PEFT), inference acceleration methods, and agent tooling including langchain and AutoGPT. The repository also addresses multimodal models like Stable Diffusion and VisualGLM, serving as a comprehensive reference for practitioners building with LLMs.