Update README.md

README.md

@@ -14,6 +14,6 @@ dataset_info:
 ---
 # Dataset Card for "coco-30-val-2014"
 
-This is 30k randomly sampled image-captioned pairs from the COCO 2014 `val` split. This is useful for image generation benchmarks (FID, CLIPScore, etc.).
+This is 30k randomly sampled image-captioned pairs from the [COCO](https://cocodataset.org/) 2014 `val` split. This is useful for image generation benchmarks (FID, CLIPScore, etc.).
 
 Refer to the gist to know how the dataset was created: https://gist.github.com/sayakpaul/0c4435a1df6eb6193f824f9198cabaa5.