Search for a command to run...
Does the Performance of Text-to-Image Retrieval Models Generalize Beyond Captions-as-a-Query?