3 ms·NVLM: Open Frontier-Class Multimodal LLMs35 points by andsoitis 2y agorkwz 2y agoHuggingFace link: https://huggingface.co/nvidia/NVLM-D-72B https://huggingface.co/nvidia/NVLM-D-72Be1gen-v 2y agoHas anyone figured out how to do “visual” chunking for rag? I’m curious how this would be used in place of an OCR service.