# 🦙 LLaMA 3 8B LoRA Fine-Tuning: Customer Support Chatbot

This project demonstrates how to fine-tune [Meta-LLaMA-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) using LoRA (Low-Rank Adaptation) on a customer support dataset. The result is a lightweight, domain-specific LLM that can assist with FAQs and support queries.

---

## 🚀 Project Overview

| Feature          | Description                                         |
|------------------|-----------------------------------------------------|
| 🧠 Base Model     | Meta-LLaMA-3-8B-Instruct                            |
| 🔧 Method         | LoRA fine-tuning (PEFT)                             |
| 🗃️ Dataset        | Custom JSONL customer service dataset              |
| 🧰 Libraries      | HuggingFace Transformers, Datasets, PEFT, BnB       |
| 🖥️ Platform       | Local GPU (RTX 4060 Ti) + HuggingFace CLI Support  |

---

## 📁 Directory Structure