We introduce MMed-RAG, a powerful multimodal RAG system that boosts the factuality of Medical Vision-Language Models (Med-LVLMs) by up to 43.8%! 🩺 [Paper] [X(Twitter)] MMed-RAG enhances alignment ...

We tested the RAG using two language models: Meta Llama-3.1 Instruct 8B and Deepseek-Distill-QWen-14B You can run these two models by running generate_llama3.py or generate_deepseek.py ...