Hugging Face Introduces Compact Versions of SmolVLM Vision Language Model That Can Run on Consumer Laptops