Sugar AI is a central hub for all AI-powered chatbots and assistants integrated within Sugar. Designed to enhance interactive learning, it brings the power of AI and LLMs into various educational activities, making learning more engaging and accessible for children. By seamlessly integrating AI into Sugar activities, Sugar AI fosters creativity, problem-solving, and interactive learning. Children can engage with AI chatbots for conversations, receive coding assistance through intelligent programming guides, and explore new concepts in an intuitive and friendly environment.
1. Chatbot for the Chat Activity💬
The Chat Activity chatbot is an interactive gen-AI bot that helps kids learn through engaging conversations.
2. AI Coding Assistant for the Pippy Activity👩💻
The Pippy Activity coding assistant is a gen-AI collaboration tool for children to learn and explore Python programming. Pippy's AI-assistant uses a Llama3.1 model from Ollama with a RAG architecture for retrieving relevent examples from the documentations. Specially enhanced to handle Python, GTK, and Pygame queries, making it an essential tool for beginner programmers.
sugar-ai/
├── main.py - Child QA with GPT-2
├── rag_agent.py - Coding assistant with RAG
├── docs/ - Documentation PDFs
└── requirements.txt - Dependency list
✔️ Python 3.10+
✔️ Ollama (for Llama3.1 model)
✔️ GPU recommended for better performance
# Clone repository
git clone https://github.com/sugarlabs/sugar-ai.git
cd sugar-ai# Create virtual environment
python -m venv venv
source venv/bin/activate # Linux/Mac
# venv\Scripts\activate # WindowsAlternatively, you can use Conda. Refer to this documentation to setup conda locally on your system.
# Create and activate a Conda environment
conda create --name <env-name> python=<python-version>
conda activate <env-name># Install dependencies
pip install -r requirements.txtGNU General Public License v3.0
See LICENSE for full text.