QuantVLA introduces a training-free quantization framework for Vision-Language-Action models, enabling efficient, low-bit deployment in robotics without arch...
Level: advanced
By Jingxuan Zhang and 6 other authors
Category: education