31 lines
1.0 KiB
Markdown
31 lines
1.0 KiB
Markdown
# Simple RAG Agent
|
||
|
||
This repository contains a simple Retrieval-Augmented Generation (RAG) agent implemented with **LangChain** and **OpenAI** embeddings. It demonstrates how to build a question‑answering system that retrieves relevant documents from a local vector store and generates answers using OpenAI’s GPT model.
|
||
|
||
## Features
|
||
|
||
- **Document ingestion**: Load text files and create embeddings.
|
||
- **Vector store**: Uses FAISS as the local vector store.
|
||
- **Retriever**: Retrieves top‑k relevant documents for a user query.
|
||
- **LLM wrapper**: Uses OpenAI GPT‑3.5‑Turbo or GPT‑4.
|
||
- **Prompt template**: Combines retrieved context with the user question.
|
||
- **Simple CLI**: Interact with the agent from the command line.
|
||
|
||
## Installation
|
||
|
||
```bash
|
||
pip install -r requirements.txt
|
||
```
|
||
|
||
## Usage
|
||
|
||
```bash
|
||
python rag_agent.py --data_dir path/to/documents --query "What is the capital of France?"
|
||
```
|
||
|
||
The agent will print the generated answer.
|
||
|
||
## Dependencies
|
||
|
||
All required packages are listed in `requirements.txt`.
|