Update README.md

README.md

@@ -38,7 +38,7 @@ The total size of additional training data is **47.44 B** tokens.
 
 The inference can be done using the following snippet of code:
 
-```{python}
+```python
 from transformers import AutoTokenizer, pipeline
 
 tokenizer = AutoTokenizer.from_pretrained("cjvt/OPT_sl")