Update app.py

app.py

@@ -104,7 +104,7 @@ md = """
 
 A Document-Retrieval QA application built by training **[RoBERTa model](https://arxiv.org/pdf/1907.11692)** on **[SQuAD 2.0](https://rajpurkar.github.io/SQuAD-explorer/)** dataset for efficient answer extraction and
 the system is augmented by using NLP based **[BM25](https://www.researchgate.net/publication/220613776_The_Probabilistic_Relevance_Framework_BM25_and_Beyond)** retriever for information retrieval from a large text corpus.
-The project is a brief enhancement and augmentation to the work done in the research paper **[Encoder-based LLMs: Building QA systems and Comparative Analysis](https://drive.google.com/file/d/1Ztd6x46g21ufoewmKZMoElMxViNfd_2P/view?usp=sharing)**.
+The project is a brief enhancement and augmentation to the work done in the research paper **Encoder-based LLMs: Building QA systems and Comparative Analysis**.
 In this paper we study about BERT and its advanced variants and learn to build an efficient answer extraction QA system from scratch.
 The built system can be used in information retrieval system and search engines.