FlirtyHuh

FlirtyHuh is a fun and interactive flirt bot that brings a touch of humor and charm to your conversations. This project involves creating a custom dataset, fine-tuning a T5-small model, and developing a user-friendly front-end using HTML, CSS, and JavaScript.

Features

• A custom-trained T5-small model for generating flirty responses.
• User-friendly interface built with HTML, CSS, and JavaScript.
• Fun and engaging interaction with users.
• Responsive design for accessibility across various devices.
• Real-time message exchange with dynamic bot responses.

Project Overview

1. Dataset Creation

• Created a custom dataset tailored for flirtatious responses.
• The dataset includes a variety of prompts and witty replies designed to emulate a charming conversationalist.
• Employed techniques to ensure diversity and creativity in the dataset.

2. Model Training

• Fine-tuned the T5-small model using the custom dataset.
• Leveraged Hugging Face Transformers library for training and evaluation.
• Optimized the model to generate creative, relevant, and engaging responses.
• Incorporated evaluation metrics such as BLEU and Rouge to measure response quality.

3. Front-End Development

• Designed a clean and intuitive front-end using HTML, CSS, and JavaScript.
• Implemented a responsive design for seamless usage across desktops, tablets, and smartphones.
• Built features to send messages, display bot responses dynamically, and manage session states.

Tech Stack

• Model: T5-small
• Libraries: Hugging Face Transformers, PyTorch
• Front-End: HTML, CSS, JavaScript
• Backend: Flask or FastAPI (for hosting the bot logic)

Setup Instructions

Backend

1. Clone the repository:

   git clone <repository-url>
   cd flirtyhuh

2. Install dependencies:

   pip install torch transformers flask

3. Load the fine-tuned T5 model and set up the backend script.
4. Run the backend server:

   python app.py

Front-End

1. Open the frontend folder.
2. Edit the config.js file to point to the backend server URL if necessary.
3. Open bot.html in a web browser.
4. Start chatting with the bot and enjoy!

Usage

1. Open the web interface.
2. Type a message into the input box and hit send.
3. The bot will respond with a witty or flirty reply in real time.

Future Enhancements

• Add support for multiple languages to cater to a wider audience.
• Integrate with popular messaging platforms like WhatsApp and Telegram.
• Expand the dataset for more nuanced and varied responses.
• Implement additional personality modes for the bot (e.g., humorous, poetic, etc.).
• Add user analytics to improve bot performance and user satisfaction.

Contributing

Feel free to contribute to FlirtyHuh! If you have suggestions for improvement or encounter any issues, please submit a pull request or create an issue in the repository.

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Enjoy chatting with FlirtyHuh and spread the charm!