add padding and truncation arguments to `pipe` example

based on feedback that the current example doesn't work for batch inference

README.md

@@ -45,7 +45,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_name)
 pipe = pipeline("text-classification", model=model, tokenizer=tokenizer, function_to_apply="sigmoid")
 
 
-pipe(example)
+pipe(example, padding=True, truncation=True)
 
 >>> [{'label': 'NZT', 'score': 0.9142044186592102}]
 ```