cloudlab for llm服务器配置

以下配置不为通用模板需要根据服务器架构此为x86_64显卡操作系统和项目需求调整#查看信息 lscpu df -h lsblk #扩展根文件系统谨慎 wget -O setup-grow-rootfs.sh https://gitlab.flux.utah.edu/johnsond/openstack-build-ubuntu/-/raw/master/setup-grow-rootfs.sh?inlinefalse sudo RESIZEROOT0 sh ./setup-grow-rootfs.sh #挂载 sudo mkfs.ext4 /dev/sdb sudo mount -t ext4 /dev/sdb /disk-sdb #更新包管理器 sudo apt-get update sudo apt-get upgrade -y #安装pip sudo apt install python3-pip -y #安装conda wget -c https://repo.anaconda.com/archive/Anaconda3-2023.03-1-Linux-x86_64.sh bash Anaconda3-2023.03-1-Linux-x86_64.sh vim ~/.bashrc export PATH~/anaconda3/bin:$PATH source ~/.bashrc #conda环境 conda create -n vllm python3.10 -y conda activate vllm #nvidia驱动安装 https://blog.csdn.net/Hello0_WorlD/article/details/143951225?spm1001.2014.3001.5501 nvidia-smi #安装cuda-toolkit https://developer.nvidia.com/cuda-toolkit-archive nvcc -V #若nvcc command not found vim ~/.bashrc export LD_LIBRARY_PATH/usr/local/cuda/lib export PATH$PATH:/usr/local/cuda/bin source ~/.bashrc nvcc -V #安装nccl https://developer.nvidia.com/nccl/nccl-legacy-downloads dpkg -l | grep nccl python -c import torch;print(torch.cuda.nccl.version()) #安装pytorch https://pytorch.org/ import torch print(torch.__version__) print(torch.cuda.is_available())# 系统盘配置 # 99盘占据swap导致无法扩充希望扩充sda3 sudo parted /dev/sda print sudo swapoff /dev/sda99 sudo parted /dev/sda rm 99 # Fix, 99 sudo parted /dev/sda resizepart 3 100% sudo resize2fs /dev/sda3 # docker echo /dev/nvme0n1 /mnt/nvm ext4 defaults 0 2 | sudo tee -a /etc/fstab sudo mkdir -p /mnt/nvm/docker sudo chmod 777 /mnt/nvm/docker curl -fsSL https://get.docker.com | sudo sh sudo mkdir -p /etc/docker sudo tee /etc/docker/daemon.json EOF { data-root: /mnt/nvm/docker } EOF sudo systemctl enable docker sudo systemctl start docker sudo usermod -aG docker $USER newgrp docker sudo systemctl stop docker sudo mkdir -p /mnt/nvm/docker sudo rsync -aqxP /var/lib/docker/ /mnt/nvm/docker/ sudo systemctl restart docker docker info | grep Docker Root Dir sudo systemctl stop docker sudo systemctl stop containerd sudo mkdir -p /mnt/nvm/containerd sudo rsync -aqxP /var/lib/containerd/ /mnt/nvm/containerd/ sudo mv /var/lib/containerd /var/lib/containerd.bak sudo ln -s /mnt/nvm/containerd /var/lib/containerd sudo systemctl start containerd sudo systemctl start docker sudo rm -rf /var/lib/containerd.bak