Update README.md

README.md

@@ -16,9 +16,9 @@ The SAD dataset is our gold standard dataset of tweets labelled for sarcasm. The
 
 # Data Splits
 
-|            | Train   | Valid | Test |
-| -----      | ------ | -----  | ---- |
-| SAD   | 1638 | 351 | 351|
+- Train: 1638
+- Valid: 351
+- Test: 351
 
 # Languages