Add model card for LServe

README.md

@@ -0,0 +1,15 @@
+---
+pipeline_tag: text-generation
+library_name: transformers
+license: apache-2.0
+---
+
+# LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention
+
+This repository contains the LServe model, as presented in the paper [LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention](https://hf.co/papers/2502.14866). LServe is an efficient system that accelerates long-sequence LLM serving via hybrid sparse attention.
+
+**Paper:** [LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention](https://hf.co/papers/2502.14866)
+
+**Code:** [https://github.com/mit-han-lab/omniserve](https://github.com/mit-han-lab/omniserve)
+
+**Project Page:** [https://hanlab.mit.edu/projects/lserve](https://hanlab.mit.edu/projects/lserve)