Team Sentinals: CancerCare Chatbot

Welcome to the official repository of Team Sentinals from MGMs College of Engineering, Nanded! We are excited to present our chatbot designed to answer basic questions related to cancer. Our chatbot provides accurate and concise responses to your inquiries.

Problem Statement

"Introduction to GenAI and Simple LLM Inference on CPU and finetuning of LLM Model to create a Custom Chatbot"

Project Overview

Our solution leverages Intel's Neural Chat LLM for local inference on CPU, making it highly accessible and efficient without requiring GPU resources. The chatbot ingests data from PDF files, converts the text into embeddings using BGE Embeddings, and implements Retrieval-Augmented Generation (RAG) using an open-source stack.

Key Components

• Intel's Neural Chat LLM: Chosen for its capability to run independently on CPUs.
• BGE Embeddings: Utilized for text embedding.
• Chroma DB: Serves as the vector store.
• Langchain & CTransformers: Frameworks used for orchestration.

Getting Started

Prerequisites

1. Clone the Repository:

   git clone https://github.com/Aditya19110/Team_Sentinals_Unnati-2024
   cd Team_Sentinals_Unnati-2024

2. Create a Virtual Environment:

   python -m venv venv

3. Activate the Virtual Environment:

   • On Windows:

     venv\Scripts\activate

   • On macOS/Linux:

     source venv/bin/activate

4. Install Dependencies:

   pip install -r requirements.txt

5. Additional Dependency:

   pip install protobuf==3.20.0

Download the LLM

Due to space constraints, the LLM is not included in this repository. Please download the Intel Neural Chat LLM from Hugging Face and place it in the appropriate directory.

Data Preparation

Convert the data to embeddings by running:

python ingest.py

Note: This process might take some time as it converts the data into embeddings.

Running the Chatbot

Start the Flask application:

python app.py

Visit the provided URL in your browser to interact with the chatbot and ask any cancer-related questions. All the documents(pdfs) and ppts are uploaded here...

Conclusion

Thank you for exploring our project! We hope our chatbot provides valuable assistance in answering your cancer-related queries.

Team Members

Name	LinkedIn	Email
Aditya Kulkarni
Kashish Aswani
Rasika Rakhewar
Nagesh Ballurkar
Pallavi Lagludkar

Our Mentor

Name	LinkedIn	Email
Mr. Shivprasad Titare