A modern versioning system for Artificial Intelligence projects.

ModelSync is a distributed version control tool inspired by Git, but entirely focused on the Artificial Intelligence ecosystem.

While Git is great for versioning code, it fails when dealing with:

• 🗂️ Large datasets
• 🧠 Trained models (binaries)
• 🧪 Experiments with multiple parameters
• 📈 Performance metrics and comparisons

The goal of ModelSync is to fill this gap, offering a data and experiment-oriented versioning environment.

• modelsync init → Initializes a .modelsync/ repository
• modelsync commit -m "message" → Saves a snapshot with metadata
• Simple versioning with file hashing
• Basic API for queries and external interactions

• Python 3.10+
• Linux, macOS or Windows (WSL recommended)

git clone https://github.com/your-username/modelsync.git
cd modelsync
pip install -r requirements.txt

# Initialize ModelSync repository
python modelsync/cli/main.py init

# Create a commit with message
python modelsync/cli/main.py commit -m "Model with 92% accuracy"