机器学习服务器准备

环境

Ubuntu：Server 20.04.4 LTS
CUDA：11.6
显卡：GeForce 1080 Ti

1. JDK8安装

sudo apt-get install openjdk-8-jdk
# 如果原本已安装jdk，可用以下命令进行切换
sudo update-alternatives --config java

2. 显卡驱动安装

2.1 官网查看并下载显卡驱动版本

https://www.nvidia.cn/Download/index.aspx?lang=cn

image.png
image.png

2.2 禁用The Nouveau kernel driver

sudo vi /etc/modprobe.d/blacklist.conf

  在文件 blacklist.conf 中加入如下内容：
  blacklist nouveau
  blacklist lbm-nouveau
  options nouveau modeset=0
sudo update-initramfs -u
sudo reboot

2.3 重启后安装

sudo apt-get install gcc make
sudo chmod +x NVIDIA-Linux-x86_64-510.60.02.run
sudo ./NVIDIA-Linux-x86_64-510.60.02.run

image.png

2.4 错误提示

如果不禁用Nouveau kernel driver会出现以下错误提示

image.png

如果不安装gcc和make会出现以下提示

image.png
image.png

2.5 查看显卡

nvidia-smi

image.png

3. 安装CUDA

https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=Ubuntu&target_version=20.04&target_type=deb_local

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/11.6.2/local_installers/cuda-repo-ubuntu2004-11-6-local_11.6.2-510.47.03-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu2004-11-6-local_11.6.2-510.47.03-1_amd64.deb
sudo apt-key add /var/cuda-repo-ubuntu2004-11-6-local/7fa2af80.pu
sudo apt-get update
sudo apt-get -y install cuda

4. Pytorch

https://pytorch.org/get-started/locally/

sudo apt install python3-pip 
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu113  -i https://pypi.tuna.tsinghua.edu.cn/simple