update model card

README.md

@@ -10,7 +10,7 @@ datasets:
 
 # Margin-MSE Trained DistilBERT-Cat (vanilla/mono/concatenated DistilBERT re-ranker)
 
-We provide a retrieval trained DistilBERT-Cat model (https://arxiv.org/pdf/2004.12832.pdf). Our model is trained with Margin-MSE using a 3 teacher BERT_Cat (concatenated BERT scoring) ensemble on MSMARCO-Passage.
+We provide a retrieval trained DistilBERT-Cat model. Our model is trained with Margin-MSE using a 3 teacher BERT_Cat (concatenated BERT scoring) ensemble on MSMARCO-Passage.
 
 This instance can be used to **re-rank a candidate set**. The architecure is a 6-layer DistilBERT, with an additional single linear layer at the end.