Geeky Ghost Writer

A powerful book generation application that leverages Ollama's local LLMs to create complete books through an intuitive Gradio interface. This tool helps writers outline, generate, and edit full-length books with minimal effort.

✨ Features

• 📚 Complete Book Generation: Create entire books from a single prompt with chapter-by-chapter generation
• 🧠 Local LLM Integration: Powered by Ollama for private, local AI inference
• 🖥️ User-friendly Interface: Intuitive Gradio UI with tabs for different aspects of book creation
• 📝 Project Management: Create, save, and load book projects with automatic UI synchronization
• 🌐 World Building Tools: Define and manage world elements for consistent storytelling
• 👤 Character Management: Create and track characters throughout your narrative
• ✏️ Advanced Editing: Edit individual chapters and perform find-and-replace across your book
• 🔍 Outline Generation: Create detailed chapter outlines with key events, character developments, settings, and tone
• 📊 Progress Tracking: Monitor book generation progress with time estimates
• 📤 Export Options: Combine chapters and export your book in text format

📋 Requirements

• Python 3.8 or higher
• Ollama installed and running locally
• At least one model loaded in Ollama (e.g., mistral, llama2, mixtral)
• Windows, macOS, or Linux operating system

🚀 Getting Started

Windows Quick Setup

1. Download or clone this repository
2. Run simple_setup.bat which will:
   • Check for Python and Ollama installation
   • Create a virtual environment
   • Install required dependencies
   • Start the application

Manual Installation

1. Ensure Python 3.8+ and Ollama are installed

2. Clone the repository:

   git clone https://github.com/GeekyGhost/Geeky-Ghost-Writer.git
   cd Geeky-Ghost-Writer

3. Create and activate a virtual environment:

   # Windows
   python -m venv venv
   venv\Scripts\activate
   
   # macOS/Linux
   python -m venv venv
   source venv/bin/activate

4. Install dependencies:

   pip install gradio>=4.0.0 requests>=2.25.0 markdown>=3.3.0 pyyaml>=6.0

5. Run the application:

   python simple_ollama_app.py

🧩 How to Use

The application interface is divided into several tabs:

1. Project Management

• Create new book projects with titles and writing style descriptions
• Load and manage existing projects
• View current project information

2. World Building & Characters

• Create detailed world elements (geography, magic systems, technology, etc.)
• Add and manage character profiles
• Edit or delete world-building elements and characters

3. Book Generation

• Enter your story prompt
• Select the number of chapters and Ollama model to use
• Generate a detailed chapter outline
• Generate the full book chapter by chapter
• Monitor generation progress
• Combine chapters into a complete book

4. Editor

• Load and edit individual chapters
• Perform find-and-replace operations across the entire book
• Save chapter changes

5. View Results

• Browse and view generated files
• Export your book in text format (with more formats planned for future updates)

💻 Advanced Usage

Recommended Models

For best results, use one of these Ollama models:

• mistral (balanced performance and quality)
• mixtral (highest quality but slower)
• llama2 (good all-around performance)
• phi (faster, lighter model)
• gemma (Google's efficient model)

💡 Model Management Tip: For an easy way to download, manage, and configure your Ollama models, check out Little Geeky's Learning UI - a companion tool that provides a user-friendly interface for Ollama model management.

Project Structure

Generated content is saved in the book_output directory with this structure:

book_output/
├── YourBookTitle_UUID_Timestamp/
    ├── book_metadata.json
    ├── outline.txt
    ├── chapter_01.txt
    ├── chapter_02.txt
    └── full_book.txt

Generation Process

The application uses a multi-step process:

1. Project Creation: Set up book details, world, and characters
2. Outline Generation: Create chapter outlines with key events
3. Chapter Generation: Generate each chapter based on the outline
4. Editing: Refine and improve content
5. Export: Combine chapters and export the book

🧰 Technical Details

The application consists of:

• Gradio-based UI with multiple interactive tabs
• Direct integration with the Ollama API for text generation
• Project management with metadata persistence
• Thread-based chapter generation to prevent UI freezing
• Progress tracking and estimation

📣 Limitations

• Generation time increases with chapter count
• Quality depends on the Ollama model used
• May require significant system resources for larger models
• Currently only exports to text format

🔮 Upcoming Features

• Additional export formats (PDF, EPUB)
• Text-to-speech integration
• Illustration generation
• Enhanced editing tools
• Custom writing styles and templates

📝 License

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

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

🔗 Related Projects

• Little Geeky's Learning UI: For now, you can use this as a companion tool for managing your Ollama models with a user-friendly interface

---

Geeky Ghost Writer - Let AI help craft your next literary masterpiece