Pytorch中实现CPU和GPU之间的切换的两种方法

2023-01-28 12:03

短信预约 -IT技能 免费直播动态提醒

# 创建网络模型
tudui = Tudui()
if torch.cuda.is_available():
   tudui = tudui.cuda()
 
# 损失函数
loss_fn = nn.CrossEntropyLoss()
if torch.cuda.is_available():
    loss_fn = loss_fn.cuda()
 
# 数据输入   包括训练和测试的代码，二者都需要添加此代码
if torch.cuda.is_available():
   imgs = imgs.cuda()
   targets = targets.cuda()

方法一：.to(device)

1.不知道电脑GPU可不可用时：

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu' )
a.to(device)

第一行代码的意思是判断电脑GPU可不可用，如果可用的话device就采用cuda()即调用GPU，不可用的话就采用cpu()即调用CPU。

第二行代码的意思就是把变量放到对应的device上（当然如果你用的是CPU的话就不用这一步了，因为变量默认是存在CPU上的，调用GPU的话要先把变量放到GPU上跑，跑完之后再调回CPU上）

２.指定GPU时

# 定义训练的设备
device = torch.device("cuda:0")
 
# 网络模型创建
tudui = Tudui()
tudui = tudui.to(device)
 
# 损失函数
loss_fn = nn.CrossEntropyLoss()
loss_fn = loss_fn.to(device)
 
# 训练步骤开始
    tudui.train()
    for data in train_dataloader:
        imgs, targets=data
        imgs = imgs.to(device)
        targets = targets.to(device)
        outputs = tudui(imgs)
        loss = loss_fn(outputs, targets)
 
# 测试步骤开始
    tudui.eval()
    total_test_loss = 0
    total_accuracy = 0
    with torch.no_grad():
        for data in test_dataloader:
            imgs, targets=data
            imgs = imgs.to(device)
            targets = targets.to(device)
            outputs = tudui(imgs)
            loss = loss_fn(outputs, targets)
            total_test_loss = total_test_loss + loss.item()
            accuracy = (outputs.argmax(1)==targets).sum()
            total_accuracy = total_accuracy + accuracy

3.指定cpu时：

device = torch.device('cpu')

方法二：

1、需要修改的

# 三种常见的写法
device = torch.device('cuda')
device = torch.device('cuda: 0')
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

2、代码

# 创建模型
tudui = Tudui()
if torch.cuda.is_available():
   tudui = tudui.cuda()
 
# 损失函数
loss_fn = nn.CrossEntropyLoss()
if torch.cuda.is_available():
    loss_fn = loss_fn.cuda()
 
# 训练步骤开始
    tudui.train()
    for data in train_dataloader:
        imgs, targets=data
        if torch.cuda.is_available():
            imgs = imgs.cuda()
            targets = targets.cuda()
        outputs = tudui(imgs)
        loss = loss_fn(outputs, targets)
 
 # 测试步骤开始
    tudui.eval()
    total_test_loss = 0
    total_accuracy = 0
    with torch.no_grad():
        for data in test_dataloader:
            imgs, targets=data
            if torch.cuda.is_available():
                imgs = imgs.cuda()
                targets = targets.cuda()
            outputs = tudui(imgs)
            loss = loss_fn(outputs, targets)
            total_test_loss = total_test_loss + loss.item()
            accuracy = (outputs.argmax(1)==targets).sum()
            total_accuracy = total_accuracy + accuracy