Update README.md

README.md

@@ -34,8 +34,7 @@ This is the <a href="https://huggingface.co/facebook/bart-base">bart-base</a> (<
 * Krapivin (<a href = "http://eprints.biblio.unitn.it/1671/1/disi09055%2Dkrapivin%2Dautayeu%2Dmarchese.pdf">Krapivin et al., 2009</a>)
 * Inspec (<a href = "https://aclanthology.org/W03-1028.pdf">Hulth, 2003</a>)
 
-Inspired by <a href = "https://aclanthology.org/2020.findings-emnlp.428.pdf">(Cachola et al., 2020)</a>, we applied control codes to fine-tune BART in a multi-task manner. First, we create a training set containing comma-separated lists of keyphrases and titles as text generation targets. For this purpose, we form text-title and text-keyphrases pairs based on the original text corpus. Second, we append each source text in the training set with control codes <|TITLE|> and <|KEYPHRASES|> respectively. After that, the training set is shuffled in random order. Finally, the preprocessed training set is utilized to fine-tune the pre-trained BART model 
-
+Inspired by <a href = "https://aclanthology.org/2020.findings-emnlp.428.pdf">(Cachola et al., 2020)</a>, we applied control codes to fine-tune BART in a multi-task manner. First, we create a training set containing comma-separated lists of keyphrases and titles as text generation targets. For this purpose, we form text-title and text-keyphrases pairs based on the original text corpus. Second, we append each source text in the training set with control codes <|TITLE|> and <|KEYPHRASES|> respectively. After that, the training set is shuffled in random order. Finally, the preprocessed training set is utilized to fine-tune the pre-trained BART model.
 ```python
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 
@@ -48,7 +47,7 @@ text = "In this paper, we investigate cross-domain limitations of keyphrase gene
         namely scientific texts from computer science and biomedical domains and news texts. \
         We explore the role of transfer learning between different domains to improve the model performance on small text corpora."
 
-#generating keyphrases
+#generating \n-separated keyphrases
 tokenized_text = tokenizer.prepare_seq2seq_batch(["<|KEYPHRASES|> " + text], return_tensors='pt')
 translation = model.generate(**tokenized_text)
 translated_text = tokenizer.batch_decode(translation, skip_special_tokens=True)[0]