Update README.md

README.md

@@ -73,7 +73,7 @@ The testing is performed on a suite of 38 datasets. See our paper for more detai
 
 ## Results
 
-The model achieves a 72.7% zero-shot top-1 accuracy on ImageNet-1k and 62.4% image retrieval Recall@5 on COCO captions. 
+The model achieves a 72.7% zero-shot top-1 accuracy on ImageNet-1k,  64.4% image retrieval 80.7% text retrieval recall@5 on COCO captions. 
 
 
 # Citation