wehub-resource-sync
48 lines
951 B
Markdown
48 lines
951 B
Markdown
# Simple RAG Workflow with LlamaIndex
|
|
|
|
A basic implementation guide for building a Retrieval-Augmented Generation (RAG) system using LlamaIndex.
|
|
|
|
## Prerequisites
|
|
|
|
- Python 3.10+
|
|
- Ollama
|
|
|
|
## Installation
|
|
|
|
1. Install Ollama:
|
|
|
|
**macOS**
|
|
```bash
|
|
curl -fsSL https://ollama.com/install.sh | sh
|
|
```
|
|
|
|
**Linux**
|
|
```bash
|
|
curl -fsSL https://ollama.com/install.sh | sh
|
|
```
|
|
|
|
2. Pull the Llama 2 model:
|
|
```bash
|
|
ollama pull llama3.2
|
|
```
|
|
|
|
## Project Overview
|
|
|
|
This project demonstrates how to:
|
|
- Set up a basic RAG system using LlamaIndex
|
|
- Integrate with Ollama for local LLM inference
|
|
- Process and index documents for retrieval
|
|
- Generate contextual responses using the indexed knowledge
|
|
|
|
## Getting Started
|
|
|
|
1. Clone this repository
|
|
2. Follow the installation steps above
|
|
3. Run the Jupyter notebook `workflow.ipynb` to see the RAG system in action
|
|
|
|
## Note
|
|
|
|
Make sure Ollama is running in the background before executing the notebook:
|
|
```bash
|
|
ollama serve
|
|
``` |