Story-Generation-using-GPT2

Objective

Creatify Stories is a Flask application that uses AI to generate short stories based on user inputs. The primary goal of the application is to allow users to creatively express ideas by providing a prompt, which the AI then expands into a narrative. This project showcases how artificial intelligence can be integrated into web applications to enhance user creativity and engagement.

Features

• User-Friendly Interface: The application has a clean and intuitive UI built with Bootstrap, making it accessible for all users.
• Text Generation: Utilizes the GPT-2 model from Hugging Face's Transformers library to generate coherent stories based on user prompts.
• Loading Animation: Displays a loading spinner while the story is being generated, improving user experience during processing delays.
• Responsive Design: The application is responsive and works seamlessly on various devices.

UI of the application

Demo Video

Story_Generation_Recording.mp4

Technologies Used

• Frontend: HTML, CSS (Bootstrap), JavaScript
• Backend: Flask (Python)
• AI Model: GPT-2 from Hugging Face Transformers

Installation

Follow these steps to set up the project locally:

1. Clone the Repository:

git clone https://github.com/theiturhs/Story-Generation-using-GPT2.git
cd Story-Generation-using-GPT2

2. Create a Virtual Environment (optional but recommended)

python -m venv venv
venv\Scripts\activate

3. Run the Application:

python App.py

4. Access the Application:

Open your web browser and go to http://127.0.0.1:8000.

Usage

1. Input a prompt in the provided text area. For example: "I was wondering what if I..."
2. Click the "Create Story" button.
3. Wait for the AI to generate a story (the loading spinner will indicate that the process is ongoing).
4. Once generated, the story will be displayed below the input form.

Code Explanation

1. Flask Backend: The backend is powered by Flask, which handles HTTP requests and renders HTML templates.
2. generate_story function: Processes the user input and generates a story using the generate_text function from the textGenerationCode module.
3. Text Generation: The generate_text function uses the Hugging Face Transformers pipeline to generate text based on the user's prompt.
4. HTML Template: The front end is styled using Bootstrap for a modern and responsive design. It includes a form for user input and a section to display the generated story.

References

• HuggingFace - GPT2

Contact

For any inquiries or feedback, feel free to reach out: