AutoAWQ

Update quantizer.py for saving vram while quantization

#757

Closed