22085081bb
Co-authored-by: Ying Sheng <sqy1415@gmail.com> Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com> Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu> Co-authored-by: parasol-aser <3848358+parasol-aser@users.noreply.github.com> Co-authored-by: LiviaSun <33578456+ChuyueSun@users.noreply.github.com> Co-authored-by: Cody Yu <hao.yu.cody@gmail.com>
515 B
515 B
Flashinfer Mode
flashinfer is a kernel library for LLM serving; we use it here to support our attention computation.
Install flashinfer
git submodule update --init --recursive
pip install 3rdparty/flashinfer/python
Run Sever With Flashinfer Mode
Add through --model_mode argument from the command line.
Example:
python -m sglang.launch_server --model-path meta-llama/Llama-2-7b-chat-hf --port 30000 --model-mode flashinfer