pytorch安装（Linux）

安装cuda

官网链接：https://developer.nvidia.com/cuda-downloads

先前版本：https://developer.nvidia.com/cuda-toolkit-archive

#CUDA Toolkit 11.8
#WSL2 Linux Ubuntu 18.04 x86_64
#创建一个文件夹
mkdir -p ~/env/cuda && cd ~/env/cuda
wget https://developer.download.nvidia.com/compute/cuda/repos/wsl-ubuntu/x86_64/cuda-wsl-ubuntu.pin
sudo mv cuda-wsl-ubuntu.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/11.8.0/local_installers/cuda-repo-wsl-ubuntu-11-8-local_11.8.0-1_amd64.deb
sudo dpkg -i cuda-repo-wsl-ubuntu-11-8-local_11.8.0-1_amd64.deb
sudo cp /var/cuda-repo-wsl-ubuntu-11-8-local/cuda-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cuda

此时如果nvidia-smi无法使用应该是显卡驱动掉了，重新安装一遍显卡驱动

写入环境变量

# 编辑文件
nano ~/.bashrc
# 写入下面命令
export PATH=/usr/local/cuda-11.8/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64:$LD_LIBRARY_PATH

# 使其生效
source ~/.bashrc
nvcc --version

安装conda管理环境

miniconda：https://docs.conda.io/projects/miniconda/en/latest/

mkdir -p ~/env/miniconda3 && cd ~/env/cuda
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh
bash miniconda.sh -b -u -p ~/env/miniconda3
rm -rf ~/env/miniconda3/miniconda.sh

~/env/miniconda3/bin/conda init bash
~/env/miniconda3/bin/conda init zsh
#关闭，重启终端

conda环境安装pytorch

conda create -n pytorch python=3.8
conda activate pytorch

root用户无法pip，会出现WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager

nano ~/.bashrc
export PIP_ROOT_USER_ACTION=ignore
source ~/.bashrc

pytorch官网：https://pytorch.org/

先前版本：https://pytorch.org/get-started/previous-versions/

conda install pytorch==2.0.0 torchvision==0.15.0 torchaudio==2.0.0 pytorch-cuda=11.8 -c pytorch -c nvidia

验证安装

python -c "import torch; print(f'PyTorch version: {torch.__version__}'); print(f'CUDA available: {torch.cuda.is_available()}')"