The simplest implementation of recent Sparse Attention patterns for efficient LLM inference.
语言:Jupyter Notebook · GitHub 星标:92 · Fork:5
在 GitHub 上查看原仓库