yolov5 environment configuration series tutorials under Windows 11 (advanced part)

After configuring the most basic environment, start our formal yoloV5 test.
For details on the basic configuration tutorial: https://blog.csdn.net/u010416569/article/details/125981012

3. Check the cuda and cudnn versions suitable for your graphics card

1. Check the CUDA suitable for your graphics card
Open the Nvidia control panel,

enter the system information in the lower left corner,

click on the component to see the CUDA version suitable for your Nvidia graphics card

2. Check the cudnn corresponding to your cuda version

If there is no version you need in the picture, you can check CUDNN on Nvidia official website

4. Download the corresponding pytorch, cuda and cudnn versions

Change channel (for faster download)

1. Open the .condarc file in the directory of C:\Users\SKY in txt format

2. Copy all the following codes into the txt

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
  simpleitk: https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud`

Download the corresponding pytorch, cuda and cudnn
-1. Enter the official website to find the corresponding pytorch version
Official website link: https://pytorch.org/get-started/locally/
Choose according to your needs, as shown in the figure, we use cuda11.1 as an example

conda install pytorch torchvision torchaudio cudatoolkit=11.1 -c pytorch -c conda-forge

-2. Open cmd and activate the environment you configured for yolov5

activate 你的环境名

-3. Download pytorch
and enter the code copied in the previous step

downloading

Download completed

3. View the configuration in the environment
Enter the following code to view the configuration in the environment

conda list

The following is the configuration that should be in the environment at this step, mainly looking at these two

cudatoolkit               11.1.1               heb2d755_7    conda-forge

pytorch                   1.9.0           py3.8_cuda11.1_cudnn8_0    pytorch

(This cudatoolkit is included in the pytorch package downloaded from the official website, so there is no need to download another CUDA)

(yolo) C:\Users\SKY>conda list
packages in environment at C:\me\Anaconda3\envs\yolo:
Name                    Version                   Build  Channel
blas                      2.111                       mkl    conda-forge
blas-devel                3.9.0              11_win64_mkl    conda-forge
ca-certificates           2021.7.5             haa95532_1    https://repo.anaconda.com/pkgs/main
certifi                   2021.5.30        py38haa95532_0    https://repo.anaconda.com/pkgs/main
cudatoolkit               11.1.1               heb2d755_7    conda-forge
freetype                  2.10.4               h546665d_1    conda-forge
intel-openmp              2021.3.0          h57928b3_3372    conda-forge
jpeg                      9b                   hb83a4c4_2    defaults
libblas                   3.9.0              11_win64_mkl    conda-forge
libcblas                  3.9.0              11_win64_mkl    conda-forge
liblapack                 3.9.0              11_win64_mkl    conda-forge
liblapacke                3.9.0              11_win64_mkl    conda-forge
libpng                    1.6.37               h1d00b33_2    conda-forge
libtiff                   4.2.0                hd0e1b90_0    defaults
libuv                     1.42.0               h8ffe710_0    conda-forge
lz4-c                     1.9.3                h8ffe710_1    conda-forge
m2w64-gcc-libgfortran     5.3.0                         6    conda-forge
m2w64-gcc-libs            5.3.0                         7    conda-forge
m2w64-gcc-libs-core       5.3.0                         7    conda-forge
m2w64-gmp                 6.1.0                         2    conda-forge
m2w64-libwinpthread-git   5.0.0.4634.697f757               2    conda-forge
mkl                       2021.3.0           hb70f87d_564    conda-forge
mkl-devel                 2021.3.0           h57928b3_565    conda-forge
mkl-include               2021.3.0           hb70f87d_564    conda-forge
msys2-conda-epoch         20160418                      1    conda-forge
ninja                     1.10.2               h5362a0b_0    conda-forge
numpy                     1.21.2           py38h089cfbf_0    conda-forge
olefile                   0.46               pyh9f0ad1d_1    conda-forge
openssl                   1.1.1k               h8ffe710_1    conda-forge
pillow                    8.3.1            py38h4fa10fc_0    defaults
pip                       21.0.1           py38haa95532_0    https://repo.anaconda.com/pkgs/main
python                    3.8.11               h6244533_1    https://repo.anaconda.com/pkgs/main
python_abi                3.8                      2_cp38    conda-forge
pytorch                   1.9.0           py3.8_cuda11.1_cudnn8_0    pytorch
setuptools                52.0.0           py38haa95532_0    https://repo.anaconda.com/pkgs/main
sqlite                    3.36.0               h2bbff1b_0    https://repo.anaconda.com/pkgs/main
tbb                       2021.3.0             h2d74725_0    conda-forge
tk                        8.6.11               h8ffe710_0    conda-forge
torchaudio                0.9.0                      py38    pytorch
torchvision               0.10.0               py38_cu111    pytorch
typing_extensions         3.10.0.0           pyha770c72_0    conda-forge
vc                        14.2                 h21ff451_1    https://repo.anaconda.com/pkgs/main
vs2015_runtime            14.27.29016          h5e58377_2    https://repo.anaconda.com/pkgs/main
wheel                     0.37.0             pyhd3eb1b0_0    https://repo.anaconda.com/pkgs/main
wincertstore              0.2                      py38_0    https://repo.anaconda.com/pkgs/main
xz                        5.2.5                h62dcd97_1    conda-forge
zlib                      1.2.11            h62dcd97_1010    conda-forge
zstd                      1.4.9                h6255e5f_0    conda-forge

We can see that although there are cuda and pytorch in the environment, there is no cudnn, so we need to download a corresponding version of cudnn

enter

conda install cudnn==8.1.0
		（该cudnn版本应与你的cuda对应，具体看上方教程）

Five, configure Yolov5 related environment

1. Download Yolov5
can be downloaded through this link https://github.com/ultralytics/yolov5

You can also download the yolov5 version I am using through my Baidu cloud https://pan.baidu.com/s/1cG6Cle2N4G274TdnnOB0Wg
Extraction code: cqdy

2. Configure the environment required by Yolov5. We unzip the downloaded yolov5 compressed package and open it.
Note that the yolov5 file should be placed in a pure English path
. Find a txt file named requirements.txt

Copy this paragraph, as follows

pip install -r requirements.txt

Open CMD, activate your configured environment, and enter the yolov5 file path you decompressed

activate yolo
cd C:\Users\SKY\Downloads\yolov5-4.0(cd到你自己安装yolov5的文件路径)
输入 pip install -r requirements.txt

Then the environment we need for yolov5 is configured

3. Download Yolov5 weight file
Link: https://pan.baidu.com/s/1RkbcY07rpKSFcDZKzM5R0g
Extraction code: cqdy

Place the weight file in the yolov5 file

6. Test whether Yolov5 can be used normally

1. Download pycharm
official website download https://www.jetbrains.com/pycharm/download/#section=windows
Baidu cloud download link: https://pan.baidu.com/s/10clX9scYVFSwGrA9bR1Org
Extraction code: cqdy

2. Configure pycharm
to open pycharm

3. Test whether yolov5 can be used normally,
right click to run the program

You can see that our CUDA is in normal use

After the operation is completed, it tells us that the result is placed in exp3, let's click to see

The test was successful.


In summary, in this process, many problems will be encountered. And everyone's problems are different. For the sake of simplicity of the tutorial, we will not expand on the problems encountered. There will be an opportunity later to compile a post dedicated to common problems. thank you all!