Quantized Inference

A technique that reduces memory footprint and accelerates computation by lowering the numerical precision of model weights or ac