Has been eval BiomedCLIP as vision model in LLaVA-Med-1.5？

[chuansuo12]

Excuse me, i find LLaVa-Med-1.5 use "openai/clip-vit-large-patch14-336", why not use BiomedCLIP ? Has been eval it ? if not, i want to trains projectors with BiomedCLIP