Spaces:
Running
Apply for community grant: Academic project (gpu and storage)
Hi HF team and community,
We are excited to introduce Apollo (https://arxiv.org/abs/2403.03640), a family of lightweight, multilingual medical LLMs designed to democratize medical AI for over 6 billion people.
While most medical knowledge is in English, delivering effective healthcare requires models that speak local languages. Apollo bridges this gap by providing state-of-the-art performance across the world's six most spoken languages. Our 7B model is the top-performing multilingual medical LLM up to the 70B scale, demonstrating that accessible, high-quality medical AI is possible on a global scale.
Our goal is to make the entire Apollo project—including our training corpora, code, and model weights—fully open-source. We want to provide the community with direct hands-on experience through an interactive demo, allowing everyone to see the potential of multilingual medical AI for themselves.
To achieve this, we need GPU resources for continued training, fine-tuning across languages, and hosting the public demos. With your support, we can break down language barriers in healthcare and make advanced medical AI accessible to a global audience.
Thanks for considering our request!