Update README.md

README.md

@@ -17,8 +17,12 @@ Illustrated below are the gradients used for this specific L2 recipe, with the t
 
 This model primarily uses (and was heavily tested with) Alpaca formatting, so for optimal model performance, use:
 ```
+<System prompt/Character Card>
+
 ### Instruction:
 Your instruction or question here.
+For roleplay purposes, I suggest the following - Write <CHAR NAME>'s next reply in a chat between <YOUR NAME> and <CHAR NAME>. Write a single reply only.
+
 ### Response:
 ```