PyTorch code for “TVLT: Textless Vision-Language Transformer” (NeurIPS 2022 Oral) - View it on GitHub