😄 😥 🙃 😭 😡
安裝conda 【本身就不需要root權限】
1.下載anconda https://www.anaconda.com/products/individual
或者直接清華園 https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/ 但是（有時候）出現不穩定抽風：wget https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/Anaconda3-2020.02-Linux-x86_64.sh
下載到本地之后，上傳到服務器，改文件 chmod 777 ** ，否則出現xftp上傳錯誤
2. sh Anaconda....sh
3.會提示安裝目錄，輸入一個【還沒有創建的安裝目錄】"Anaconda3 will now be installed into this location": 如 /4Tdisk/用戶名/Anaconda
4.裝完它自己會添加path到~/.bashrc
5. source ~/.bashrc【更新一下用戶配置文件】，通過conda info --e檢查是否安裝好了
換清華源 win/linux下：
建議直接參考官網命令: , 設置清華源鏡像
按提示更改 ~/.condarc。再運行索引緩存 conda clean -i
查看鏡像中是否有要安裝的pytorch版本
或者直接參考下面的（個人覺得不好用，總是抽風(╬▔皿▔)凸）
https://blog.csdn.net/wujialaoer/article/details/84977796
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/msys2/
conda config --set show_channel_urls yes
【有時候網絡還是有問題可參考】https://blog.csdn.net/ebzxw/article/details/80702506
【如果conda 安裝總是 HTTP 錯誤，可使用 pip】 https://blog.csdn.net/lsf_007/article/details/87931823
關於conda 和 pip 安裝
1.首選 conda install ***
2.如果conda太慢或失敗，備選 pip install ***
3.如果pip安裝超時，連接出錯，可以使用 pip install *** -i https://pypi.tuna.tsinghua.edu.cn/simple
安裝cuda
1.查看版本顯卡nvidia-smi,驅動版本是440.44
2.搜“nvidia driver cuda version”，可以看到可以安裝cuda 10.1版本
3.sh cuda***.run
https://blog.csdn.net/hizengbiao/article/details/88625044
4.或者直接使用別人的cuda，寫入到 ~/.profile
PS. 關於找合適的cuda
1.【查看驅動和cuda匹配版本】nvidia-smi 是440.44版本
2.【查看內核和cuda匹配版本】gcc -v 是16.04 5.4.0版本
https://docs.nvidia.com/deploy/cuda-compatibility/index.html
【那我就安裝10.1版本的】
自定義目錄安裝cuda和cudnn
自定義目錄：先創建安裝路徑 /4Tdisk/***/software/cuda/cuda-10.1/
https://blog.csdn.net/hizengbiao/article/details/88625044
【安裝結果】
(base) ***@hp:/4Tdisk/***/download$ sh cuda_10.1.243_418.87.00_linux.run
===========
= Summary =
===========
Driver: Not Selected
Toolkit: Installed in /4Tdisk/***/software/cuda/cuda-10.1/
Samples: Installed in /home/***/
Please make sure that
- PATH includes /4Tdisk/***/software/cuda/cuda-10.1/bin
- LD_LIBRARY_PATH includes /4Tdisk/***/software/cuda/cuda-10.1/lib64, or, add /4Tdisk/***/software/cuda/cuda-10.1/lib64 to /etc/ld.so.conf and run ldconfig as root
To uninstall the CUDA Toolkit, run cuda-uninstaller in /4Tdisk/***/software/cuda/cuda-10.1/bin
Please see CUDA_Installation_Guide_Linux.pdf in /4Tdisk/***/software/cuda/cuda-10.1/doc/pdf for detailed information on setting up CUDA.
***WARNING: Incomplete installation! This installation did not install the CUDA Driver. A driver of version at least 418.00 is required for CUDA 10.1 functionality to work.
To install the driver using this installer, run the following command, replacing <CudaInstaller> with the name of this run file:
sudo <CudaInstaller>.run --silent --driver
Logfile is /tmp/cuda-installer.log
【按照Summary配置PATH】
export PATH="/4Tdisk/**/cuda/cuda-10.1/bin:$PATH"
export LD_LIBRARY_PATH="/4Tdisk/**/cuda/cuda-10.1/lib64:$LD_LIBRARY_PATH"
下載解壓cudnn與復制
tar -xzvf cudnn-10.1-linux-x64-v7.6.5.32.tgz -C /4Tdisk/***/software/cuda/cuda-10.1/tem
cp /4Tdisk/***/software/cuda/cuda-10.1/tem/cuda/include/cudnn.h /4Tdisk/***/software/cuda/cuda-10.1/include
cp /4Tdisk/***/software/cuda/cuda-10.1/tem/cuda/lib64/libcudnn* /4Tdisk/***/software/cuda/cuda-10.1/lib64
chmod a+r /4Tdisk/***/software/cuda/cuda-10.1/include/cudnn.h /4Tdisk/***/software/cuda/cuda-10.1/lib64/libcudnn*
查看版本 nvcc -V 或者 nvcc --version
安裝pytorch
1.創環境 conda create --name pytorch
conda install pytorch torchvision cudatoolkit=10.1
或者指定pytorch版本
conda install pytorch=1.5 torchvision cudatoolkit=10.1 -c pytorch 【兩種凡是都試試尤其是當網絡抽風的時候😫】
檢查cuda版本和pytorch版本是否一致
$ python -c "import torch; print(torch.version.cuda)"
>>> 10.1
$ nvcc --version
>>> 10.1
測試是否成功
import torch
print(torch.__version__)
print(torch.cuda.is_available())
a = torch.ones(1,1)
print(a.cuda())
莫名其妙 —— 一切隨緣🐶
- 報錯段錯誤
清楚package緩存： conda clean -y --all - Solving environment: failed with initial frozen solve
conda config --set channel_priority flexible
Conde設置代理：https://blog.csdn.net/leviopku/article/details/98766822