How does model quantization affect performance?

Model quantization reduces size and complexity, often leading to improved inference speed without significant loss in accuracy.