This model demonstrates question-answering capabilities in the field of biomedicine that rival those of human experts and achieves state-of-the-art performance in cross-modal question-answering tasks involving natural language, molecules, and proteins. Additionally, the collaboration resulted in the joint open-source release of the world's first freely available and commercially usable biomedical-specific language model, Llama 2, named BioMedGPT-LM-7B. The "AIR-Intelligence Source Health Computing Joint Research Center" also collaborated on open-sourcing the DrugFM, a basic model for small molecule drug research. These open-sourced biomedical foundational models, emphasizing scientific research and commercial usability, provide a large-scale model foundation for biomedical research and applications.
Download：
