Update README.md

README.md

@@ -9,6 +9,13 @@ license: mit
 
 With the addition of chatting data, the model is trained to handle the semantics of sequences longer than KoBART.
 
+```python
+from transformers import reTrainedTokenizerFast, BartModel
+
+tokenizer = PreTrainedTokenizerFast.from_pretrained('gogamza/kobart-base-v2')
+model = BartModel.from_pretrained('gogamza/kobart-base-v2')
+```
+
 ### Performance 
 
 NSMC