Update README.md

README.md

@@ -20,11 +20,10 @@ This tagging system allows models to focus on the specific ambiguous words durin
 
 ## Data Format
 
-The dataset is organized as a dictionary structure where:
-- Keys are sense IDs
-- Values are the corresponding definitions
-
-Each entry in the original dataset was processed from the `senses.txt` file, which contained entries separated by empty lines. Each entry included multiple fields, from which we extracted the sense IDs and definitions.
+The dataset is organized to suit for the alpaca_prompt:
+- Instruction
+- Input
+- Output
 
 ## Usage