1 Change source
Domestic sources have obvious speed advantages. Commonly used ones include Tsinghua University, Alibaba, etc.
# 1) System software source # vim /etc/apt/sources.list # The source code image is commented by default to improve the speed of apt update. You can uncomment it yourself if necessary. ---------- deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-updates main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-updates main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-backports main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-backports main restricted universe multiverse deb http://security.ubuntu.com/ubuntu/ jammy-security main restricted universe multiverse # deb-src http://security.ubuntu.com/ubuntu/ jammy-security main restricted universe multiverse # Pre-release software source, not recommended to be enabled # deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-proposed main restricted universe multiverse # # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-proposed main restricted universe multiverse ------------------- # 2) pip source $ sudo apt update $ sudo apt install python3-pip $ pip install pip -U #Upgrade to the latest version $ pip --version $ pip 22.0.2 from /usr/lib/python3/dist-packages/pip (python 3.10) # Modify pip source to Tsinghua source pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple
2 Download Anaconda
wget https://repo.anaconda.com/archive/Anaconda3-2023.09-0-Linux-x86_64.sh bash Anaconda3-2023.09-0-Linux-x86_64.sh source ~/.bashrc vim ~/.bashrc ------------------- # >>> conda initialize >>> # !! Contents within this block are managed by 'conda init' !! __conda_setup="$('/home/woodman/anaconda3/bin/conda' 'shell.bash' 'hook' 2> /dev/null)" if [ $? -eq 0 ]; then eval "$__conda_setup" else if [ -f "/home/woodman/anaconda3/etc/profile.d/conda.sh" ]; then . "/home/woodman/anaconda3/etc/profile.d/conda.sh" else export PATH="/home/woodman/anaconda3/bin:$PATH" fi fi unset __conda_setup # <<< conda initialize <<< conda deactivate #export PATH=/home/woodman/anaconda3/bin:$PATH # The following are my terminal interface settings, personal habits, changes or comments export PS1="\[\e]0;\u@ \w\a\]${debian_chroot: + ($debian_chroot)}\[\033[01;32m\]\u\[\033[01;32m \]:[\W]$" ---------------------------------- # Conda also changed its source to Tsinghua University $ conda config --set show_channel_urls yes #Modify it after generating the ~/.condarc. ---- channels: -defaults show_channel_urls: true default_channels: - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/r - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/msys2 custom_channels: conda-forge: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud msys2: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud bioconda: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud menpo: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud pytorch: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud pytorch-lts: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud simpleitk: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud ---- conda clean -i clears the index cache conda config --show-sources
3 Creation and deletion of Anaconda environment
# Commonly used software is installed when creating $conda create -n woodman python=3.11 ipykernel psutil jupyter jupyterlab nodejs numpy matplotlib $conda activate woodman $python -m ipykernel install --user --name woodman --display-name "torch2.1" $cat ~/.local/share/jupyter/kernels/woodman/kernel.json $jupyter kernelspec list $jupyter kernelspec remove kernel_name # delete $ conda remove -n your_env_name (virtual environment name) --all $ conda remove --name your_env_name package_name # Delete a package in the environment # Back up installed software conda list -e > requirements.txt conda install --yes --file requirements.txt pip freeze > requirements.txt pip install -r requirements.txt #Smoothly install torch again. Correspondence between torch, torchvision, torchaudio and torchtext versions in #PyTorch #https://blog.csdn.net/shiwanghualuo/article/details/122860521 #CPU $pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu #GPU,cu121 is nvidia-smi: CUDA Version: 12.0 $pip install torch==2.1.0 + cu121 torchvision==0.16.0 + cu121 torchaudio==2.1.0 + cu121 -f https://download.pytorch.org/whl/torch_stable.html
4 Nvidia f u
The driver installation of the graphics card should have been placed second, but considering its complexity, I put it last.
Let’s go through the various links first. In fact, all the required URLs and commands are listed later.
https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html https://developer.nvidia.com/cuda-downloads https://developer.nvidia.com/CUDA-toolkit-archive # Version correspondence table between CUDA and NVIDIA Driver https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html
To download a new version of cuda, I usually use the method (1)
(1) # Driver Installer # Check the driver version supported by the graphics card, generally download "recommended" # The server version can better adapt to the alchemy environment $ubuntu-drivers devices ---- driver: nvidia-driver-535-server-open - distro non-free recommended -------- $ubuntu-drivers install nvidia-driver-535-server-open # Base Installer # “recommended” # Installation Instructions: wget https://developer.download.nvidia.com/compute/cuda/12.3.0/local_installers/cuda_12.3.0_545.23.06_linux.run # Only select Cuda Toolkit XX.X and Driver installed in front, others are not required. sudo sh cuda_12.3.0_545.23.06_linux.run # After the installation is complete, there are cuda and cuda-12.1 in the /usr/local/ path. # Note that cuda is a soft link, pointing to cuda-12.1 $cd /usr/local/ $file cuda
(2) # Base Installer # Installation Instructions: wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600 wget https://developer.download.nvidia.com/compute/cuda/12.3.0/local_installers/cuda-repo-ubuntu2204-12-3-local_12.3.0-545.23.06-1_amd64.deb sudo dpkg -i cuda-repo-ubuntu2204-12-3-local_12.3.0-545.23.06-1_amd64.deb sudo cp /var/cuda-repo-ubuntu2204-12-3-local/cuda-*-keyring.gpg /usr/share/keyrings/ sudo apt-get update sudo apt-get -y install cuda-toolkit-12-3 #DriverInstaller #NVIDIA Driver Instructions (choose one option) #To install the legacy kernel module flavor: sudo apt-get install -y cuda-drivers #To install the open kernel module flavor: sudo apt-get install -y nvidia-kernel-open-545 sudo apt-get install -y cuda-drivers-545 ------- # If it is already installed, delete it first and then install it. sudo apt-get --purge remove nvidia-kernel-source-XXX sudo apt-get install --verbose-versions nvidia-kernel-open-XXX sudo apt-get install --verbose-versions cuda-drivers-XXX
System environment configuration
$ sudo vim ~/.bashrc ---------- #cuda-20231111 export CUDA_HOME=$CUDA_HOME:/usr/local/cuda export PATH=/usr/local/cuda/bin:$PATH export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH ---------------- $ source ~/.bashrc $ nvcc -V
Install cuDnn
It will be faster to surf the Internet scientifically. Just register an account with QQ email.
tar -xf cudnn-linux-x86_64-8.9.5.30_cuda12-archive.tar.xz mv cudnn-linux-x86_64-8.9.5.30_cuda12-archive cuda sudo cp cuda/include/cudnn*.h /usr/local/cuda/include sudo cp cuda/lib/libcudnn* /usr/local/cuda/lib64 sudo chmod a + r /usr/local/cuda/include/cudnn*.h /usr/local/cuda/lib64/libcudnn* #Verify whether the installation is successful cat /usr/local/cuda/include/cudnn_version.h | grep CUDNN_MAJOR -A 2
Reference:
3090 graphics card + CUDA11.0 + torch1.7.1_torch1.7.1 and cuda11.0 how to download – CSDN blog
[NVIDIA] Ubuntu 20.04 installation nvidia-460 + cuda-11.2_a100 460 cuda11.2-CSDN Blog
U20.4 upgrade pytorch 1.11_pip upgrade torch_woodman718’s blog-CSDN blog