Search for a command to run...
VALOR: Vision-Audio-Language Omni-Perception Pretraining Model and Dataset