A complete pipeline for training, inference, and evaluation of a large language model (LLM) using Hugging Face. This repository includes notebooks, scripts, and configuration to take a model from raw data → fine-tuning/training → inference (with examples) plus evaluation and deployment readiness.