Its just a model file, Tokenizer is not the same as model. You can use any Tokenizer from Transformers like DistilBertTokenizer or BertTokenizer

---
language:
- English 
tags
- Text 
- Sequence-Classification
- Sarcasm  
- DistilBert 
datasets:
- Kaggle Dataset on News Headline Sarcasm - https://www.kaggle.com/rmisra/news-headlines-dataset-for-sarcasm-detection