The preprint for our latest work, BIOMEDICA, is now online! Find > 24 million image-text pairs for VLM training!