8000 GitHub - efeslab/Nanoflow: A throughput-oriented high-performance serving framework for LLMs
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

efeslab/Nanoflow

Repository files navigation

Nanoflow

Docker setup

mkdir -p ~/framework-test
docker run --gpus all --net=host --privileged -v /dev/shm:/dev/shm --name nanoflow -v ~/framework-test:/code -it nvcr.io/nvidia/cuda:12.8.1-cudnn-devel-ubuntu22.04

apt update
apt install pybind11-dev
apt install liburing-dev
apt install libopenmpi-dev
sysctl -w kernel.io_uring_disabled=0
sysctl -w vm.nr_hugepages=65536

Gurobi License Setup (for Docker)

Follow these steps to obtain a Gurobi license and configure it so your Docker container can use it.

1. Request a Gurobi License

  1. Go to the Gurobi website and create an account (https://www.gurobi.com/).
  2. After logging in, navigate to My Gurobi → Get License.
  3. Choose the "WLS Academic" license type and fill out any required fields.
  4. Gurobi will email you a license file named gurobi.lic (or provide you with a license key string).

2. Place the License on Your Host Machine

mkdir -p ~/gurobi/license
mv /path/to/downloaded/gurobi.lic ~/gurobi/license/
ls ~/gurobi/license

Install Dependencies

git clone git@github.com:efeslab/Nanoflow.git
cd Nanoflow
chmod +x ./installAnaconda.sh
./installAnaconda.sh
# restart the terminal
source ~/.bashrc

cd Nanoflow-python
yes | bash setup.sh

Build

cd pybind
mkdir -p build
cmake ..
make -j 256

End-to-end Test

cd entry
python run_llama3.py -load_hf_weight=True

About

A throughput-oriented high-performance serving framework for LLMs

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •  
0