1.1、显卡配置:RTX3060 桌面系统:ubuntu18.04
1.2、根据1.1的配置选择对应的Python、CUDA、pytorch、torchvision,我选择Python = 3.8| CUDA 11.7 |pytorch =1.13 | torchvision= 0.11
注意:这几个安装包的版本需要对应起来,版本不对会产生错误!!!
网上教程很多,可以直接去官网安装即可。
要更换清华或者中科大的源,后期下载的速度才不会i被限制:
命令:查看主目录下所有文件
ls -a
编辑.condarc
gedit ./.conda
将下面的源拷贝到文件中并保存。
channels:
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/msys2/
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge
show_channel_urls: true
auto_activate_base: false
激活conda工作环境、查看python的版本、查看安装的CUDA的版本,命令如下:
meng@meng-PC:~$ conda activate pytorch_env
(pytorch_env) meng@meng-PC:~$ python --version
Python 3.8.0
(pytorch_env) meng@meng-PC:~$ nvidia-smi
结果如图:
方法一:通过pytorch官网选择安装,因为速度太慢,本文没有选择这种方式
方法二:去清华源直接下载需要的包,直接安装
https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/linux-64/
cd到自己的安装目录下运行下面指令,完成pytorch的安装
conda install pytorch-1.13.0-py3.8_cuda11.7_cudnn8.5.0_0.tar.bz2
同样的
conda intalll torchvision-0.11.0-py38_cu113.tar.bz2
(pytorch_env) meng@meng-PC:~$ python
Python 3.8.0 (default, Nov 6 2019, 21:49:08)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.cuda.is_available()
True
>>>
如果结果为ture恭喜你torch的环境安装成功,但是实际情况往往不是这样,下面是我的踩坑记录:
1、报错信息如下:
(pytorch_env) meng@meng-PC:~/下载$ python
Python 3.8.0 (default, Nov 6 2019, 21:49:08)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
Traceback (most recent call last):
File "" , line 1, in <module>
File "/home/meng/anaconda3/envs/pytorch_env/lib/python3.8/site-packages/torch/__init__.py", line 191, in <module>
_load_global_deps()
File "/home/meng/anaconda3/envs/pytorch_env/lib/python3.8/site-packages/torch/__init__.py", line 153, in _load_global_deps
ctypes.CDLL(lib_path, mode=ctypes.RTLD_GLOBAL)
File "/home/meng/anaconda3/envs/pytorch_env/lib/python3.8/ctypes/__init__.py", line 369, in __init__
self._handle = _dlopen(self._name, mode)
OSError: libmkl_intel_lp64.so: cannot open shared object file: No such file or directory
解决:conda的环境变量不在当前用户的搜索路径中,先打开当前用户环境变量配置文件bashrc,将
export LD_LIBRARY_PATH=/home/jc/anaconda3/lib:$LD_LIBRARY_PATH
其中jc为用户名
参考文章:https://blog.csdn.net/Christine_11/article/details/126943635
2、报错信息二
>>> import torch
Traceback (most recent call last):
File "" , line 1, in <module>
File "/home/meng/anaconda3/envs/pytorch_env/lib/python3.8/site-packages/torch/__init__.py", line 753, in <module>
from .serialization import save, load
File "/home/meng/anaconda3/envs/pytorch_env/lib/python3.8/site-packages/torch/serialization.py", line 18, in <module>
from typing_extensions import TypeAlias
ModuleNotFoundError: No module named 'typing_extensions'
这个简单,直接安装typing_extensions即可
指令;
pip install typing_extensions
https://mp.csdn.net/mp_download/manage/download/UpDetailed