Pinned Loading

1. swiftLLM swiftLLM Public

   A tiny yet powerful LLM inference system tailored for researching purpose. vLLM-equivalent performance with only 2k lines of code (2% of vLLM).

   Python 126 7

2. LLMServe/DistServe LLMServe/DistServe Public

   Disaggregated serving system for Large Language Models (LLMs).

   Jupyter Notebook 403 50

3. hpcgame.minesweeper hpcgame.minesweeper Public

   C++ 12 1

4. CUDA-From-Correctness-To-Performance-Code CUDA-From-Correctness-To-Performance-Code Public

   Codes & examples for "CUDA - From Correctness to Performance"

   C++ 74 18