ChatBot-RAG

A powerful chatbot implementation using Retrieval Augmented Generation (RAG) to provide context-aware responses based on your data.

Features

• 🔍 Retrieval Augmented Generation: Enhances LLM responses with relevant context from your data
• 🧠 Ollama Support: Run models locally with Ollama for privacy and customization
• 🔗 LangChain Integration: Built on the powerful LangChain framework for advanced chains and pipelines

Installation

pip install chatbot-rag

Requirements

• Python 3.12
• Ollama (for local model hosting)
• Tesseract-OCR (for image-based data extraction)

Tesseract Installation

To enable image-based data extraction, you need to install Tesseract-OCR.
Follow the installation instructions provided at this link.

After installation, ensure the tesseract executable is accessible in your system's PATH.
For example, on Windows, you can verify by running:

tesseract --version

Ollama Installation

Ollama is required for hosting models locally.
Refer to the official Ollama documentation for installation instructions: Ollama Installation Guide.

Once installed, verify it is working by running:

ollama --version

Use

Quick Start

from chatbot_rag.chat import Chatbot 
from chatbot_rag.RAG import RAG

# Use a specific Ollama model
rag = RAG(path="./data/")
rag()
bot = Chatbot(name="deepseek-r1:8b")

# Query with specific parameters
question = "Summarize my recent research on climate change"
context  = rag._search_context(question,k=5)
response = bot(context,question)
print(response)

Using temporal paths

from chatbot_rag.chat import Chatbot 
from chatbot_rag.RAG import RAG


with tempfile.TemporaryDirectory() as tmpdirname:
    persistent_dir = os.path.join(tmpdirname, "all_info/")
    os.makedirs(persistent_dir, exist_ok=True)
    rag = RAG(path="./data/",base_persist_path=persistent_dir)
    rag()
    chatbot = Chatbot(name="llama3.1:8b")

    question = "What is the main topic of the document?"
    context = rag._search_context(question)
    answer = chatbot(context=context, question=question)
    print(f"Answer: {answer}")

Using other preprocessing (PyMuPDFPreprocessing)

from chatbot_rag.chat import Chatbot 
from chatbot_rag.RAG import RAG
from src.chatbot_rag.preprocessing import PyMuPDFPreprocessing


kwargs = {"tesseract_path": "C:/Program Files/Tesseract-OCR/tesseract"}
rag = RAG(path="./data/",preprocessing=PyMuPDFPreprocessing,**kwargs)
rag()
chatbot = Chatbot(name="llama3.1:8b")

question = "What is the main topic of the document?"
context = rag._search_context(question)
answer = chatbot(context=context, question=question)
print(f"Question: {answer}")

By default, the system will attempt to extract information from images using Tesseract-OCR. You can disable image extraction by adding the following to the kwargs:

kwargs = {"extract_images": False}

and passing it directly to the RAG component.

Projects Using ChatBot-RAG

🌟 ChatBot-RAG App

ChatBot-RAG App is a chatbot framework leveraging Retrieval-Augmented Generation (RAG) to deliver context-aware responses.
It integrates 🔗 LangChain for advanced pipelines and supports 🧠 Ollama for local model hosting, ensuring enhanced privacy and customization.

Key Features:

• 🔍 Context-Aware Responses: Uses RAG to provide accurate and relevant answers.
• 🧠 Local Model Hosting: Powered by Ollama for privacy and flexibility.
• 🔗 Advanced Pipelines: Built on LangChain for seamless integration and extensibility.

Explore the project here: ChatBot-RAG App

License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.