VLMQ: Efficient Post-Training Quantization for Large Vision-Language Models via Hessian Augmentation

MQVLMRe-assign community