Search for a command to run...
RoboLLM: Robotic Vision Tasks Grounded on Multimodal Large Language Models