Update README.md

README.md

@@ -4,6 +4,15 @@ license: cc-by-nc-3.0
 
 PragS2: Pragmatic Masked Language Modeling with Emoji_any dataset followed by Hashtag-Based Surrogate Fine-Tuning
 
+You can load these model and use for downstream fine-tuning. For example:
+```python
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+
+tokenizer = AutoTokenizer.from_pretrained('UBC-NLP/prags1', use_fast = True)
+model = AutoModelForSequenceClassification.from_pretrained('UBC-NLP/prags1',num_labels=lable_size)
+```
+
+
 More details are in our paper:
 ```
 @inproceedings{zhang-abdul-mageed-2022-improving,