DocRAG : Talk with your PDFs

An intelligent document interaction system that lets you have natural conversations with your PDF documents using state-of-the-art large language models and semantic search.

Features

• Command line interface
• Natural conversation with your PDFs
• Language understanding with Groq's LLama 3 70B model

Setup

1. clone

git clone https://github.com/DocRAG.git
cd DocRAG

2. Add your Docs Create a DATA folder in the project root Drop your PDF files into the DATA folder

3. Set your API key Create a .env file in the project root

Add your Groq API key to the .env file

GROQ_API_KEY="your api key"

4. Run DocRAG

python main.py

Advanced Config

Change the behaviour of DocRAG by adjusting

# Document Processing
CHUNK_SIZE = 200
CHUNK_OVERLAP = 20

# LLM Settings
TEMPERATURE = 0.7
MAX_TOKENS = 2048

# Agent Config
MAX_ITERATIONS = 4