CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
Agent sessions will appear here when coding agents work on this repository.