NVLM: Open Frontier-Class Multimodal LLMs

35 pointsposted a year ago
by andsoitis
(research.nvidia.com)

2 Comments

e1gen-v

a year ago

Has anyone figured out how to do “visual” chunking for rag? I’m curious how this would be used in place of an OCR service.