Instruction finetuning based chatBot design

Project Code :TCMAPY1936

Objective

The project aims to develop an AI-driven wedding chatbot using advanced deep learning models to provide accurate, context-aware, and real-time assistance for wedding planning. It involves fine-tuning LLAMA-3-8B and DeepSeek-r1 models on JSONL wedding datasets, evaluating performance using metrics like exact match, ROUGE, and F1-score. The best model will integrate Retrieval Augmented Generation (RAG) for contextual responses and be deployed via a Flask web app for seamless interaction. Features include real-time query handling, administrative control, analytics, and reporting. The system will be scalable and adaptable to different wedding types, ensuring efficient event management and enhanced user experience.

Abstract

 

Consider the case of a wedding chatbot that does not only respond to the questions, but is aware of the details of each aspect of the wedding, such as the time when the ceremony should take place and the details of the dress code. Tuning the state-of-the-art deep learning models (LLAMA-3-8B and DeepSeek-r1) to specialized JSAONL wedding datasets, we develop an intelligent assistant that allows giving detailed and context-aware answers to the details regarding the wedding. These models are tested with powerful performance measures such as accuracy, precision, recall and F1-score, which guarantees the best and accurate performance of the chatbot.

This is done by fine-tuning the wedding-specific queries using the JSAONL wedding data to suit the LLAMA-3-8B and DeepSeek-r1 models. The data sets are broad in nature as they have a variety of wedding specifications such as date, venue, dresses, and RSVP. All the models are strictly tested to identify the capacity of each to provide the right and relevant response and the model with the best performance is identified to be deployed.

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.

Block Diagram

Specifications

Hardware Requirements

Processor                           - I7/Intel Processor

Hard Disk                          - 160GB

Key Board                         - Standard Windows Keyboard

Mouse                                           - Two or Three Button Mouse

Monitor                             - SVGA

RAM                                             - 8GB

Software Requirements:

Operating System             :  Windows 11

Server side Script              :  HTML, CSS, Bootstrap & JS

Programming Language   :  Python

Libraries                            :  Flask, Pandas, Mysql.connector, Os, Smtplib, Numpy

IDE/Workbench                :  PyCharm

Technology                                   :  Python 3.6+

Demo Video