Given the popularity of Agglomerative Vision Foundation Models such as RADIO, RADIOv2.5. It would be great if they can be added into TIMM. https://github.com/NVlabs/RADIO