环境配置

  1. Win + R,并输入sysdm.cpl
  1. 选择高级,然后点击环境变量
  1. 在系统变量区中选择Path,点击编辑
  1. 点击新建,输入程序的执行路径(绝对路径)。
  1. 若未关闭当前CMD。输入echo %PATH%会显示原来的 PATH 值。

  2. 若已关闭当前CMD。每个正在运行的 Windows 程序都会有自己已加载的 PATH,可以使用Process Explorer来查看当前正在运行的程序的环境变量。例如:如果你之前打开了 Chrome,且一直未关闭,按Ctrl+O打开C:\Windows\System32\cmd.exe,然后输入echo %PATH%会显示原来的 PATH 值。恢复之后删除C:\Program Files\Google\Chrome\Application和用户变量中的PATH值即可。

  3. 若已重启电脑。手动恢复 PATH 到默认值%SystemRoot%\system32;%SystemRoot%;%SystemRoot%\System32\Wbem;%SYSTEMROOT%\System32\WindowsPowerShell\v1.0,其他值已丢失。

参考:如何恢复我删除的Path环境变量?

Visual Studio Code 官网

  1. 安装编译器
    • Windows(二选一)
    • Linux: sudo apt install gcc g++ gdb build-essential
    • Mac OS: xcode-select --install
  2. 安装 VS Code C/C++ 扩展ms-vscode.cpptools
  3. asd
  4. 解决头文件找不到
    • 修改c_cpp_properties.json
    • Linux:gcc -v -E -x c++ -
Windows 10 + Visual Studio 2019
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
{
    "configurations": [
        {
            "name": "Win32",
            "includePath": [
                "${workspaceFolder}/**",
                "C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.29.30037/include/**"
            ],
            "defines": [
                "_DEBUG",
                "UNICODE",
                "_UNICODE"
            ],
            "compilerPath": "C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.29.30037/bin/Hostx64/x64/cl.exe",
            "cStandard": "c17",
            "cppStandard": "c++17",
            "intelliSenseMode": "windows-msvc-x64"
        }
    ],
    "version": 4
}
Windows 10 + Mingw-w64
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
{
    "configurations": [
        {
            "name": "Win32",
            "includePath": [
                "${workspaceFolder}/**",
                "C:/Portable/mingw64/lib/gcc/x86_64-w64-mingw32/8.1.0/include/**",
                "C:/Portable/mingw64/lib/gcc/x86_64-w64-mingw32/8.1.0/include/c++/**"
            ],
            "defines": [
                "_DEBUG",
                "UNICODE",
                "_UNICODE"
            ],
            "compilerPath": "C:/Portable/mingw64/bin/g++.exe",
            "cStandard": "c17",
            "cppStandard": "c++17",
            "intelliSenseMode": "gcc-x64"
        }
    ],
    "version": 4
}
  1. 打开设置。
    • C_Cpp: Clang_format_fallback Style设为{BasedOnStyle: LLVM, UseTab: Never, IndentWidth: 4, TabWidth: 4}
    • C_Cpp › Default: Cpp Standard设为``
    • C_Cpp › Default: C Standard设为``

参考

关闭 Windows Defender “首次看到时阻止”:

问题描述:

编译运行程序的时候总是弹出一个 Microsoft Defender 防病毒程序窗口,提示“需要扫描当前程序”。

解决方法:

  1. Win+R,输入gpedit.msc,打开本地组策略编辑器。
  2. 左侧选择计算机配置->管理模板->Windows 组件-Microsoft Defender 防病毒->MAPS
  3. 右侧双击配置“首次看到时阻止”功能,选择已禁用,然后点击确定,保存退出。
  • .vscode/c_cpp_properties.json
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
{
    "configurations": [
        {
            "name": "Linux",
            "includePath": [
                "${workspaceFolder}/**"
            ],
            "defines": [],
            "compilerPath": "/usr/bin/g++",
            "cStandard": "c11",
            "cppStandard": "c++11",
            "intelliSenseMode": "linux-gcc-x64"
        }
    ],
    "version": 4
}
  • .vscode/tasks.json
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
{
    // See https://go.microsoft.com/fwlink/?LinkId=733558
    // for the documentation about the tasks.json format
    "version": "2.0.0",
    "tasks": [
        {
            "label": "C/C++: g++ build active file",
            "type": "shell",
            "command": "/usr/bin/g++",
            "args": [
                "-g",
                "-std=c++11",
                "${file}",
                "-o",
                "${fileDirname}/${fileBasenameNoExtension}"
            ],
            "options": {
                "cwd": "/usr/bin"
            },
            "problemMatcher": [
                "$gcc"
            ],
            "group": {
                "kind": "build",
                "isDefault": true
            },
            "detail": "Task generated by Debugger."
        }
    ]
}
  • .vscode/launch.json
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
{
    // Use IntelliSense to learn about possible attributes.
    // Hover to view descriptions of existing attributes.
    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
    "version": "0.2.0",
    "configurations": [
        {
            "name": "C/C++: g++ build and debug active file",
            "type": "cppdbg",
            "request": "launch",
            "program": "${fileDirname}/${fileBasenameNoExtension}",
            "args": [],
            "stopAtEntry": false,
            "cwd": "${workspaceFolder}",
            "environment": [],
            "externalConsole": false,
            "MIMode": "gdb",
            "miDebuggerPath": "/usr/bin/gdb",
            "setupCommands": [
                {
                    "description": "Enable pretty-printing for gdb",
                    "text": "-enable-pretty-printing",
                    "ignoreFailures": true
                }
            ],
            "preLaunchTask": "C/C++: g++ build active file"
        }
    ]
}
  1. Python官网
  2. Miniconda官网
  3. Anaconda官网
  • Windows:%APPDATA%\pip\pip.ini
1
2
[global]
proxy = http://user:password@proxy_name:port

更新:conda update

查看环境:conda env listconda info -e

创建环境:conda create -n <ENVNAME> python=3.X -y

删除环境:conda remove -n <ENVNAME> --all -y

设置代理:conda config --set proxy_servers.http http://127.0.0.1:10809

PyTorch 官网

CPU:

Anaconda:

conda install pytorch torchvision torchaudio cpuonly -c pytorch

Pip:

pip install torch==1.8.1+cpu torchvision==0.9.1+cpu torchaudio===0.8.1 -f https://download.pytorch.org/whl/torch_stable.html

GPU (CUDA 11.0):

Anaconda:

conda install pytorch torchvision torchaudio cudatoolkit=11.0 -c pytorch

Pip:

pip install torch==1.7.1+cu110 torchvision==0.8.2+cu110 torchaudio===0.7.2 -f https://download.pytorch.org/whl/torch_stable.html

检查是否安装成功:

1
2
3
4
5
import torch
# 检查 pytorch 是否安装成功
print(torch.__version__)
# 检查 CUDA 是否可用
print(torch.cuda.is_available())

TensorFlow 官网

Pip:

CPU and GPU: pip install tensorflow

Wheel:

1
2
3
4
# https://storage.googleapis.com/tensorflow/windows/cpu/tensorflow_cpu-2.4.0-cp38-cp38-win_amd64.whl
pip install tensorflow_cpu-2.4.1-cp38-cp38-win_amd64.whl  # CPU
# https://storage.googleapis.com/tensorflow/windows/gpu/tensorflow_gpu-2.4.0-cp38-cp38-win_amd64.whl
pip install tensorflow_gpu-2.4.0-cp38-cp38-win_amd64.whl  # GPU

检查是否安装成功:

1
2
3
4
5
6
import tensorflow as tf
# 检查 tensorflow 是否安装成功
print(tf.__version__)
# 检查 CUDA 是否可用
# 输出最后一行显示 [PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]
print(tf.config.experimental.list_physical_devices('GPU'))

问题描述:

安装 CUDA 11.1 + cuDNN 8.1,tensorflow 2.4.1 检查 GPU 时报错:

Could not load dynamic library 'cusolver64_10.dll'; dlerror: cusolver64_10.dll not found

解决方案:

卸载重新安装 CUDA 11.0 + cuDNN 8.0,参考 Windows 经过测试的构建配置

  1. NVIDIA 驱动程序下载,选择对应版本,下载安装。
NVIDIA 驱动程序下载
  1. CUDA 工具包下载,下载对应版本安装。
CUDA 工具包下载

CMD输入nvcc -V,出现如下输出表示安装成功。

1
2
3
4
5
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2020 NVIDIA Corporation
Built on Tue_Sep_15_19:12:04_Pacific_Daylight_Time_2020
Cuda compilation tools, release 11.1, V11.1.74
Build cuda_11.1.relgpu_drvr455TC455_06.29069683_0
  1. cuDNN 下载,需要登陆账号,登陆后下载对应版本,解压将binincludelib三个文件夹的内容复制到 CUDA 安装目录下。