feat: Add model loading status prints and remove `theme` from Gradio examples.

feat: Configure model loading with `BitsAndBytesConfig` for 4-bit quantization.

Deploy Sultan Math AI: Add Gradio interface and model configuration

initial commit