Building Practical AI Applications with LLMs
The companion code for a Korean-language book on building practical LLM applications, covering transformer architecture, model training and serving, embeddings and retrieval-augmented generation with LlamaIndex.
This repository holds the Jupyter notebook exercises for a Korean book on building real-world LLM applications, running from transformer fundamentals through model training, embeddings, lightweight serving under limited compute, and building retrieval-augmented generation systems with LlamaIndex, plus more advanced topics such as multimodal input and agents.
The exercises run in Google Colab, which offers a free T4 GPU with a 12-hour runtime limit, and notebooks can either be downloaded and opened locally or opened directly from GitHub inside Colab.