Fetching the paper…

VL-GPT: A Generative Pre-trained Transformer for Vision and Language Understanding and Generation · Around