机房部署wiki

272次阅读
没有评论

共计 1615 个字符,预计需要花费 5 分钟才能阅读完成。

机房部署配置:

系统:centos7

显卡:rtx3090*8

安装目标:

sklearn

tqdm

numpy

torch

安装步骤:

1、安装显卡驱动

查询显卡型号

lspci -nn | grep NV

查看后几位代码:例如NVIDIA Corporation TU104GL [Tesla T4] [10de:1eb8]

在网站搜索:http://pci-ids.ucw.cz/mods/PC/10de?action=help?help=pci

确认3090无误

https://blog.csdn.net/weixin_40579389/article/details/121481083

驱动安装升级依赖为gcc rpm格式

查看gcc版本 gcc –version

下载地址: https://centos.pkgs.org/7/centos-x86_64/

gcc需要依赖rpm

安装 rpm -ivh kernel-devel-3.10.0-1160.el7.x86_64.rpm

N卡驱动库

下载驱动,上传

https://download.nvidia.com/XFree86/Linux-x86_64/530.30.02/

运行 sh NVIDIA-Linux-x86_64-530.30.02.run

安装cuda11.7(配合torch1.13.1)

下载地址:

https://developer.nvidia.com/cuda-11-7-0-download-archive?target_os=Linux&target_arch=x86_64&Distribution=CentOS&target_version=7&target_type=runfile_local

运行 sh cuda_11.7.0_515.43.04_linux.run

安装anaconda

下载:https://www.anaconda.com/products/distribution

上传安装:sh Anaconda3-2021.05-Linux-x86_64.sh

配置环境:

避免ImportError: /lib64/libstdc++.so.6: version `CXXABI_1.3.9′ not found

vi ~/.bashrc 后面添加保存

export PATH=”/public/home/lq1202010794/anaconda3/bin:$PATH”

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/root/anaconda3/lib/

source ~/.bashrc

https://blog.csdn.net/qq_44004676/article/details/126511759

2、申请 pip 清华源访问,安装库

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple torch

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple sklearn

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tqdm

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple numpy

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple transformers

查看python3 whitchpython

3、运行操作

显示所有日志:journalctl -r

查看进程:ps aux 查看pid

kill -9 pid

查看显卡负载:nvidia-smi

linux 会话保持:

screen -ls

screen -r 18245

screen -d

https://blog.csdn.net/zhang_j_mail/article/details/127091772

正文完
 
评论(没有评论)