AI CookBook : LLM Guide .PY

dev.io

About this app

The AI CookBook is an interactive companion for engineers and AI enthusiasts who want to move beyond pre-built chatbots and create their own specialized intelligence. The app breaks down the complex AI lifecycle into four distinct, manageable stages:

Build: Fine-tune base models (like Qwen) using specific training modes such as Reasoning, Function Calling, or RAG-optimized retrieval.

Convert: Transform trained LoRA adapters into optimized .GGUF files for high-speed, quantized inference on consumer-grade hardware.

Local Run: Seamlessly integrate your models into Ollama for local testing and validation.

Serve: Deploy your custom intelligence as a professional REST API using Python and Flask, complete with enterprise-grade features like IP-based rate limiting and detailed hardware requirement cheat sheets.

With a focus on "Core Concepts" and "Essentials," the app simplifies complex math (like low-rank matrix decomposition) into actionable code snippets and hardware-specific advice, ensuring users know not just how to build, but why specific parameters work.

What’s new

Added 4 New Cookbooks to Get Started

- What Are Large Language Models? A Simple Guide to AI Models and Their Types
- Inside an LLM: How AI Models Are Trained and How They Process Information
- LLM Basics Explained: Parameters, Precision (Floating Points), and Model Size
- Using LLMs in Everyday Tasks: Practical Uses and Hardware Requirements