Interactive Story Generator with WebSocket and GPT-2

This project is an Interactive Story Generator that allows users to participate in a dynamic storytelling experience powered by GPT-2. It uses WebSockets for real-time communication, enabling user choices to influence the direction of the story. The system narrates the story using pyttsx3 for text-to-speech and leverages torch for deep learning inference with GPT-2.

Project Structure

• main.py: Runs the main interactive session, initializes the story generator, and handles user input.
• story_generator.py: Manages story generation using GPT-2, processes user input via WebSocket, and generates responses.
• web_socket_server.py: Sets up a WebSocket server to manage user interactions, processing choices to drive the story forward.
• utils.py: Contains a text-to-speech function using pyttsx3 for narrating story segments.
• requirements.txt: Lists the required libraries.

Key Features

1. GPT-2 Story Generation: Uses the pre-trained GPT-2 model to generate story content, with user choices affecting the narrative flow.
2. Real-Time Interaction: WebSocket-based server allows real-time interaction and choice-driven storytelling.
3. Narration: Implements pyttsx3 for narrating story segments to enhance the immersive experience.
4. User-Driven Plot: Responds to user choices to create a customized story experience.

Requirements

• Python 3.x
• Libraries: Transformers, Rich, pyttsx3, WebSockets, torch

Install dependencies using:

pip install -r requirements.txt

Usage

1. Start the WebSocket Server: Run the WebSocket server in web_socket_server.py to handle incoming user connections and responses.

   python web_socket_server.py

2. Launch the Main Program: Start the main interactive story session in main.py, where you can provide inputs to steer the story.

   python main.py

3. Interact with the Story:

   • The program will prompt you to make choices as the story progresses.
   • Enter "A" or "B" to choose different paths, and the system will generate the next story segment based on your choice.

Example

• Initial story prompt: “In a distant land, there was a mystical forest where strange creatures roamed freely...”
• User selects option "A" or "B".
• The program responds with a new story segment based on the chosen option, narrates it, and displays the text in the console.

File Details

• main.py: Handles the main program flow, initializes the story generator, and waits for user inputs.
• story_generator.py: Loads GPT-2, generates story text, manages WebSocket connection, and processes user choices.
• web_socket_server.py: Hosts the WebSocket server, processes user options, and sends responses back to story_generator.py.
• utils.py: Defines a narrate_text function that uses pyttsx3 to convert story text to speech.

Contributing

Contributions are welcome to expand the choice options, enhance story generation prompts, or improve real-time interaction.

License

This project is open-source and available under the MIT License.