replace LongLLMLingua hyperlink with its github repo

README.md

@@ -15,7 +15,7 @@ pipeline_tag: text-generation
 
 ## TL;DR
 _Refiner_ is an end-to-end extract-and-restructure paradigm that incorporates query-relevant contents, contexts, and sectionalizes interconnected information, ensuring information distinction and alignment with the original context. _Refiner_ achieves a 80.5% tokens reduction and a 1.6-7.0% improvement margin in
-multi-hop tasks, rivaling with next best solution, [LongLLMLingua](https://arxiv.org/abs/2310.06839).
+multi-hop tasks, rivaling with next best solution, [LongLLMLingua](https://github.com/microsoft/LLMLingua).
 
 **How _Refiner_ Works**
 _Refiner_ integrates in post-retrieval process seamlessly with RAG systems, leveraging a single decoder-only LLM to: