update example

README.md

@@ -272,7 +272,7 @@ output = tokenizer.batch_decode(output)
 
 # loop over the batch to print, in this example the batch size is 1
 for i in output:
-    print(output)
+    print(i)
 ```
 
 ## Training Data