Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
In this paper, we propose LayoutLMv3 to pre-train multimodal Transformers for Document AI with unified text and image masking.