ubuntu22.04 pytorch gtx1080ti

Recently, I have updated my Ubuntu Operating System to version 22.04 and I find out that the installation of CUDA and Cudnn is much more simpler that earlier version. Therefore, I have decided to create this blog to help others setting up the environment easily.

Nvidia Drivers

Let's us start with Installation of Nvidia Driver. Basically this driver is installed when we update our Ubuntu Operation System. Basically we need to know which driver has been used by our Operating System. In Ubuntu 22.04, we can click the "show application" on the bottom left and type in "additional driver" to check for the Nvidia Driver that we are using.

additional driver

Nvidia Drivers

As we can see, there are multiple Nvidia drivers and I have selected Nvidia-driver-525 to use. To install for multiple drivers, we can run these commands.

复制代码
sudo apt update
sudo apt upgrade
sudo ubuntu-drivers autoinstall
reboot
nvidia-smi

Cuda Toolkits

Once we have chosen the Nvidia Driver (Nvidia-driver 525 for my case) suitable for us, we could start installing the Cuda toolkits.

复制代码
sudo apt update
sudo apt upgrade
sudo apt install nvidia-cuda-toolkit

After installing the toolkit, we need to know the supported CuDNN version for the installed Cuda toolkits. We could run the following command.

复制代码
nvcc --version 

this is the output :

Suitable CuDNN version

CuDNN

From the output, we get to know the installed Cuda toolkits is version 11.x. Therefore, the corresponding CuDNN is Local Installer for Ubuntu22.04 x86_64 (Deb). If we go the nvidia CuDNN website, we will notice that there are 2 versions of CuDNN for Cuda 12.x and Cuda 11.x. We could choose another version of CuDNN Local Installer for Ubuntu22.04 x86_64 (Deb) if the installed Cuda is 12.x.

List of CuDNN

Once we have downloaded the suitable CuDNN, we could run the following command to install the CuDNN.

复制代码
sudo dpkg -i cudnn-local-repo-ubuntu2204-8.9.3.28_1.0-1_amd64.deb
sudo cp /var/cudnn-local-repo-ubuntu2204-8.9.3.28/cudnn-local-7F7A158C-keyring.gpg /usr/share/keyrings/ 

After the command finish running, we are done with the installation! Before we celebrate the success, let's us test the CUDA & CuDNN installation from virtual environment using torch library.

Virtual Environment

Installation of virtual environment can be done using the following scripts.

复制代码
sudo apt-get install python3-pip
sudo pip3 install virtualenv
virtualenv -p py3.10 venv
source venv/bin/activate

CUDA & CuDNN test : pytorch & tensorflow

Installation of pytorch library can be done using the following scripts.

复制代码
import torch
print(torch.cuda.is_available()) # should be True

in additional, we could test it using tensorflow.

复制代码
pip3 install tensorflow

This python script can be used to do the test.

复制代码
import tensorflow as tf
print("Num GPUs Available: ", tf.config.list_physical_devices('GPU') , len(tf.config.list_physical_devices('GPU')))

Once we see the following output, we can start celebrating by giving me a clap!!!

复制代码
Num GPUs Available:  [PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')] 1

Before leaving, I have a shameless promotion on my Udemy course : Practical Real-World SQL and Data Visualization. I have been working in the top travel platform company for years and I find that the free visualization tool, Metabase, is very useful. Therefore, I spend weekends creating the course and I wish you could benefit from the course. Lastly, I really appreciate if you could take a look on the course and even better, please help me to share the course with your friends just like I share the knowledge to you. Thank you!!!

相关推荐
产品经理独孤虾6 分钟前
人工智能大模型如何助力电商产品经理打造高效的商品工业属性画像
人工智能·机器学习·ai·大模型·产品经理·商品画像·商品工业属性
老任与码16 分钟前
Spring AI Alibaba(1)——基本使用
java·人工智能·后端·springaialibaba
蹦蹦跳跳真可爱58929 分钟前
Python----OpenCV(图像増强——高通滤波(索贝尔算子、沙尔算子、拉普拉斯算子),图像浮雕与特效处理)
人工智能·python·opencv·计算机视觉
nananaij34 分钟前
【Python进阶篇 面向对象程序设计(3) 继承】
开发语言·python·神经网络·pycharm
雷羿 LexChien39 分钟前
从 Prompt 管理到人格稳定:探索 Cursor AI 编辑器如何赋能 Prompt 工程与人格风格设计(上)
人工智能·python·llm·编辑器·prompt
两棵雪松1 小时前
如何通过向量化技术比较两段文本是否相似?
人工智能
heart000_11 小时前
128K 长文本处理实战:腾讯混元 + 云函数 SCF 构建 PDF 摘要生成器
人工智能·自然语言处理·pdf
敲键盘的小夜猫1 小时前
LLM复杂记忆存储-多会话隔离案例实战
人工智能·python·langchain
开开心心_Every2 小时前
便捷的Office批量转PDF工具
开发语言·人工智能·r语言·pdf·c#·音视频·symfony
高压锅_12202 小时前
Django Channels WebSocket实时通信实战:从聊天功能到消息推送
python·websocket·django