✨ Little Star App

"Twinkle, twinkle, little star, how I wonder what you are..."

Little Star App is a magical AI playground designed to bring the power of Large Language Models (LLMs) directly to your device. Powered by Twinkle AI and the robust llama.cpp engine, this app lets you explore, download, and chat with AI models completely offline.

🌟 Key Features

• 🚀 On-Device Inference: Run GGUF models locally with privacy and speed using llama.cpp.
• 💬 AI Chat: Interact with LLMs through a user-friendly chat interface.
• 📥 Model Manager: Integrated browser to discover and download GGUF models from Hugging Face.
• ⚡ Performance Testing: "Completion Mode" to test model raw performance and generation speed.
• 📱 Cross-Platform: Built with Flutter for Android, iOS, and Desktop (WIP).

📸 Screenshots

Platform	Home & Models	Chat Interface	Model Completion
Android
iOS

🛠️ Tech Stack

• Framework: Flutter
• Inference Engine: llama.cpp (via FFI)
• Local Storage: Hive
• Networking: Dio

🚀 Getting Started

Prerequisites

• Flutter SDK (3.7.0 or later)
• C++ Compiler (CMake, GCC/Clang) for building native libraries.
• Android NDK (for Android build) / Xcode (for iOS build).

Installation

1. Clone the repository (including submodules):

git clone --recursive https://github.com/your-org/little-star-app.git
cd little-star-app

2. Install dependencies:

flutter pub get

3. Build & Run:

flutter run

📚 Documentation

Check out our development notes for deep dives into the implementation:

• FFI Integration Guide
• Android Integration
• iOS Integration

🤝 Contributing

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

📄 License

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

🙏 Acknowledgments

• llama.cpp for the incredible inference engine.
• Hugging Face for the model ecosystem.