WSL安装CUDA独显 PyTorch 深度学习环境配置，测速为什么要用WSL？

前言

为了预习大三课程，想提前学习下PyTorch。
于是我遇到了神仙学习教程《动手学深度学习》，同时以此为参考完成了环境配置，感谢大佬们无私奉献Thanks♪(･ω･)ﾉ

本教程展示了独显windows电脑用WSL Ubuntu子系统跑PyTorch深度学习的环境配置，至于为什么用WSL嘛。因为自己测试一番发现性能比win强很多。篇幅有限，省略WSL安装过程。

安装

设备：

Windows11: WSL-Ubuntu-22.04
1660ti-6g独显

Miniconda安装

Miniconda官网 wget方式配合链接下载对应Linux版本

wget https://repo.anaconda.com/miniconda/Miniconda3-py310_23.3.1-0-Linux-x86_64.sh

sh指令默认安装

sh Miniconda3-py310_23.3.1-0-Linux-x86_64.sh -b

初始化环境

~/miniconda3/bin/conda init

提示关闭该Terminal，重新打开一个。
创建新环境，名称d2l可更改。

conda create --name d2l python=3.9 -y

切换到d2l环境

conda activate d2l

注意每次运行都要运行此指令。例如每次source .bashrc 之后。

如要退出当前环境：conda deactivate。
如要完整删除名为dal的环境：conda remove -n d2l --all

CUDA安装

CUDA(官网)是英伟达官方的深度学习工具包，如图是WSL的选项。运行下载代码。
cuda_download

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/12.1.1/local_installers/cuda-repo-wsl-ubuntu-12-1-local_12.1.1-1_amd64.deb
sudo dpkg -i cuda-repo-wsl-ubuntu-12-1-local_12.1.1-1_amd64.deb
sudo cp /var/cuda-repo-wsl-ubuntu-12-1-local/cuda-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cuda

添加完成后需更新~/.bashrc文件

sudo vi ~/.bashrc
conda activate d2l

i进入insert模式，添加以下代码到文件最后，注意修改为对应版本，此处使用cuda12.1。

export PATH=/usr/local/cuda-12.1/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-12.1/lib64\${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

Esc, :wq，回车保存。

source ~/.bashrc
conda activate d2l

测试

wsl中运行以下代码，输出应如图：

nvcc -V

nvcc_v_printout

wsl中运行以下代码，输出应如图：

nvidia-smi

nvidia-msi_printout

若not found，同时windows terminal运行有效，请考虑以下两种原因：

错误原因：

wsl版本低，需升级成wsl2
nvidia显卡驱动版本过低，可通过Nvidia Experience升级

安装PyTorch框架

GPU版本的需要在PyTorch官网选择，我下的是Preview版本的刚好适配CUDA12.1如图，不过向下兼容性也不错。
pytorch_download
确保以下代码在d2l环境下运行！

conda install pytorch torchvision torchaudio pytorch-cuda=12.1 -c pytorch-nightly -c nvidia

下载d2l包

pip install d2l==0.17.6

显卡测试

克隆测试文件

git clone https://github.com/pytorch/examples.git
cd examples/mnist/

替换main.py文件内容为以下：

from __future__ import print_function
import argparse
import torch
import time
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
from torch.optim.lr_scheduler import StepLR


class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 32, 3, 1)
        self.conv2 = nn.Conv2d(32, 64, 3, 1)
        self.dropout1 = nn.Dropout(0.25)
        self.dropout2 = nn.Dropout(0.5)
        self.fc1 = nn.Linear(9216, 128)
        self.fc2 = nn.Linear(128, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = F.relu(x)
        x = self.conv2(x)
        x = F.relu(x)
        x = F.max_pool2d(x, 2)
        x = self.dropout1(x)
        x = torch.flatten(x, 1)
        x = self.fc1(x)
        x = F.relu(x)
        x = self.dropout2(x)
        x = self.fc2(x)
        output = F.log_softmax(x, dim=1)
        return output


def train(args, model, device, train_loader, optimizer, epoch):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        data, target = data.to(device), target.to(device)
        optimizer.zero_grad()
        output = model(data)
        loss = F.nll_loss(output, target)
        loss.backward()
        optimizer.step()
        if batch_idx % args.log_interval == 0:
            print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
                epoch, batch_idx * len(data), len(train_loader.dataset),
                100. * batch_idx / len(train_loader), loss.item()))
            if args.dry_run:
                break


def test(model, device, test_loader):
    model.eval()
    test_loss = 0
    correct = 0
    with torch.no_grad():
        for data, target in test_loader:
            data, target = data.to(device), target.to(device)
            output = model(data)
            test_loss += F.nll_loss(output, target, reduction='sum').item()  # sum up batch loss
            pred = output.argmax(dim=1, keepdim=True)  # get the index of the max log-probability
            correct += pred.eq(target.view_as(pred)).sum().item()

    test_loss /= len(test_loader.dataset)

    print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
        test_loss, correct, len(test_loader.dataset),
        100. * correct / len(test_loader.dataset)))


def main():
    # Training settings
    parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
    parser.add_argument('--batch-size', type=int, default=64, metavar='N',
                        help='input batch size for training (default: 64)')
    parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N',
                        help='input batch size for testing (default: 1000)')
    parser.add_argument('--epochs', type=int, default=14, metavar='N',
                        help='number of epochs to train (default: 14)')
    parser.add_argument('--lr', type=float, default=1.0, metavar='LR',
                        help='learning rate (default: 1.0)')
    parser.add_argument('--gamma', type=float, default=0.7, metavar='M',
                        help='Learning rate step gamma (default: 0.7)')
    parser.add_argument('--no-cuda', action='store_true', default=False,
                        help='disables CUDA training')
    parser.add_argument('--no-mps', action='store_true', default=False,
                        help='disables macOS GPU training')
    parser.add_argument('--dry-run', action='store_true', default=False,
                        help='quickly check a single pass')
    parser.add_argument('--seed', type=int, default=1, metavar='S',
                        help='random seed (default: 1)')
    parser.add_argument('--log-interval', type=int, default=10, metavar='N',
                        help='how many batches to wait before logging training status')
    parser.add_argument('--save-model', action='store_true', default=False,
                        help='For Saving the current Model')
    args = parser.parse_args()
    use_cuda = not args.no_cuda and torch.cuda.is_available()
    use_mps = not args.no_mps and torch.backends.mps.is_available()

    torch.manual_seed(args.seed)

    if use_cuda:
        device = torch.device("cuda")
    elif use_mps:
        device = torch.device("mps")
    else:
        device = torch.device("cpu")

    train_kwargs = {'batch_size': args.batch_size}
    test_kwargs = {'batch_size': args.test_batch_size}
    if use_cuda:
        cuda_kwargs = {'num_workers': 5,	#线程数
                       'pin_memory': True,
                       'shuffle': True}
        train_kwargs.update(cuda_kwargs)
        test_kwargs.update(cuda_kwargs)

    transform=transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize((0.1307,), (0.3081,))
        ])
    dataset1 = datasets.MNIST('../data', train=True, download=True,
                       transform=transform)
    dataset2 = datasets.MNIST('../data', train=False,
                       transform=transform)
    train_loader = torch.utils.data.DataLoader(dataset1,**train_kwargs)
    test_loader = torch.utils.data.DataLoader(dataset2, **test_kwargs)

    model = Net().to(device)
    optimizer = optim.Adadelta(model.parameters(), lr=args.lr)

    scheduler = StepLR(optimizer, step_size=1, gamma=args.gamma)
    start_time = time.time()
    for epoch in range(1, args.epochs + 1):
        train(args, model, device, train_loader, optimizer, epoch)
        test(model, device, test_loader)
        scheduler.step()
    end_time = time.time()
    total_time = end_time - start_time
    print(f"Total time: {total_time:.2f} seconds.")
    if args.save_model:
        torch.save(model.state_dict(), "mnist_cnn.pt")


if __name__ == '__main__':
    main()

此处采用5线程，可自行修改。
运行对应模式测速，根据配置可能cpu和gpu运行代码刚好相反。

$ python main.py	#cpu模式
$ CUDA_VISIBLE_DEVICES=2 python main.py  #gpu模式

测试代码来源及参考：
pytorch/examples - github
深度学习：Windows11 VS WSL2 VS Ubuntu 性能对比，pytorch2.0性能测试！

完成！

至此环境配置完毕，我继续跟随《动手学深度学习》了。

cd ~
mkdir d2l-zh && cd d2l-zh
curl https://zh-v2.d2l.ai/d2l-zh-2.0.0.zip -o d2l-zh.zip
unzip d2l-zh.zip && rm d2l-zh.zip
cd pytorch