Our carefully selected list of quantized models, designed to run efficiently within the constraints of limited VRAM.