Update README.md

README.md

@@ -18,6 +18,10 @@ original_sentence = tokenizer.encode_plus("Ülke genelinde bisiklet kullanımın
 paraphrased_sentences = model.generate(original_sentence['input_ids'], max_length=150, num_return_sequences=5, num_beams=5)
 tokenizer.batch_decode(paraphrased_sentences, skip_special_tokens=True)
 ```
+## Input
+```
+Ülke genelinde bisiklet kullanımının artması hakkında ne düşünüyorsun?
+```
 ## Outputs
 
 ```
@@ -28,12 +32,15 @@ tokenizer.batch_decode(paraphrased_sentences, skip_special_tokens=True)
  'ülke genelinde bisiklet kullanımının artması hakkında fikirleriniz nedir?']
  ```
  
- 
 ## Dataset
  
 We used 50994 question sentence pairs, which are created manually, to train our model. The dataset is provided our mentor. Sentences were extracted from the titles of topics in popular Turkish forum website donanimhaber.com. We augmented the dataset by writing ten thousand sentences per person.
  
 
- ## Authors
- <a href="https://github.com/metinbinbir">Metin Binbir</a>  
- <a href="https://github.com/sercaksoy">Sercan Aksoy</a>
+ ## Authors & Citation
+ 
+ <a href="https://github.com/metinbinbir">Metin Binbir</a></br>
+ <a href="https://github.com/sercaksoy">Sercan Aksoy</a>
+```
+Metin Binbir, Sercan Aksoy, Paraphrase Generation for Turkish Language, Computer Project, Advisor: Mehmet Fatih Amasyali, Yildiz Technical University, Computer Engineering Dept. , Istanbul, Turkey , 2021.
+```