Update README.md

README.md

@@ -31,6 +31,20 @@ It is a FLAN-T5-xl model (3B parameters) finetuned on:
 
 ## Usage
 
+The input text should be of the format:
+
+```
+POST: { the context, such as the 'history' column in SHP }
+
+RESPONSE A: { first possible continuation }
+
+RESPONSE B: { second possible continuation }
+
+Which response is better? RESPONSE
+```
+
+The output generated by SteamSHP-XL will either be `A` or `B`.
+
 Here's how to use the model:
 
 ```python
@@ -48,20 +62,6 @@ Here's how to use the model:
 ['A']
 ```
 
-The input text should be of the format:
-
-```
-POST: { the context, such as the 'history' column in SHP }
-
-RESPONSE A: { first possible continuation }
-
-RESPONSE B: { second possible continuation }
-
-Which response is better? RESPONSE
-```
-
-The output generated by SteamSHP-XL will either be `A` or `B`.
-
 If the input exceeds the 512 token limit, you can use [pybsd](https://github.com/nipunsadvilkar/pySBD) to break the input up into sentences and only include what fits into 512 tokens.
 When trying to cram an example into 512 tokens, we recommend truncating the context as much as possible and leaving the responses as untouched as possible.