Search for a command to run...
VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model