Add .gitignore and implement model weight loading with quantization support c250d8c Running manoskary commited on Dec 1, 2025
