Update README.md

README.md

@@ -4,6 +4,7 @@ IndoT5-base trained on translated PAWS.
 
 ## Model in action
 
+```python
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 
 tokenizer = AutoTokenizer.from_pretrained("Wikidepia/IndoT5-base-paraphrase")  
@@ -22,6 +23,7 @@ outputs = model.generate(
     early_stopping=True,
     num_return_sequences=5
 )
+```
 
 ## Limitations