Search for a command to run...
VATT: Transformers for Multimodal Self-Supervised Learning from Raw Video, Audio and Text