Alice in Wonderland Chatbot

Introduction

This project implements a chatbot based on the text of Alice in Wonderland using a Retrieval-Augmented Generation (RAG) workflow. The chatbot processes the book's contents, stores them in a Chroma vector store, and uses a fine-tuned transformer model (BART) to generate responses based on user queries.

Components Used:

• Chroma: A vector database for storing and retrieving chunks of text.
• SentenceTransformers: For generating embeddings of text for efficient similarity search.
• Hugging Face's BART model: For summarization and generating coherent responses.

The system allows users to ask questions based on the text of Alice in Wonderland and will provide a relevant answer derived from the book's content.

Setup Instructions

Prerequisites

Before you start, ensure you have the following installed:

• Python 3.7 or higher
• pip for package management
• git for version control

Clone the Repository

First, clone the repository to your local machine:

git clone https://github.com/baync180705/rag_chatbot.git
cd rag_chatbot

Running the project using Docker (Recommended Approach)

If you do not have the Tarball Image:

You will first have to build the docker image and then run it inside a docker container.

For building the image (This process might take a long time to execute depending on your device specifications):

sudo docker build -t alice_in_wonderland_chatbot .

Now you can simply run the image in interactive mode:

sudo docker run -it alice_in_wonderland_chatbot -q "Your query here" #or you can use --query_text in place of -q
# You may also run it without the query. As soon as the container runs, you will eventually be prompted to enter it.
sudo docker run -it alice_in_wonderland_chatbot

If you have the Tarball Image:

Load the Image:

docker load < alice_in_wonderland_chatbot.tar

Run the image in interactive mode:

sudo docker run -it alice_in_wonderland_chatbot -q "Your query here" #or you can use --query_text in place of -q
# You may also run it without the query. As soon as the container runs, you will eventually be prompted to enter it.
sudo docker run -it alice_in_wonderland_chatbot

Manual Setup

Create a Virtual Environment

Next, we create a virtual environment and activate it

For linux/macOS:

python3 -m venv venv #or python -m venv venv
source venv/bin/activate

For Windows:

python -m venv venv
venv\Scripts\activate

Install Dependencies

After activating the virtual environment, we install the dependencies

pip install -r requirements.txt

Run the Project

Now since all the dependencies have been installed, we are good to run the project

python main.py -q "Your query here" #or you can use --query_text in place of -q
# You may also run it without the query. As soon as the container runs, you will eventually be prompted to enter it.
sudo docker run -it alice_in_wonderland_chatbot

Conclusion

This project demonstrates a basic RAG-based chatbot using the text from Alice in Wonderland. By leveraging vector-based similarity search and a generative transformer model, it can answer questions from the book with contextually relevant responses.