unsloth/medgemma-4b-pt

MedGemma 4B is a 4.3 billion parameter multimodal variant of the Gemma 3 architecture, developed by Google, specifically pre-trained and optimized for performance on medical text and image comprehension. It utilizes a SigLIP image encoder trained on diverse de-identified medical data including chest X-rays, dermatology, ophthalmology, and histopathology images. This model excels at tasks requiring medical image classification, visual question answering, and report generation, providing a strong baseline for healthcare AI applications.