速度起飞!

相关API

1. torch.cuda.is_available() # 是否存在GPU
2. torch.cuda.device_count() # 可用的GPU个数
3. torch.cuda.get_device_name() # 取得设备名
4. torch.device() # 使用选择的设备

样例demo

注意：所有的tensor和model需要手动调用to(device)设置为使用GPU资源，否则报错

device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu') # 使用GPU

class NeuralNetwork(nn.Module):
    def __init__(self):
        super().__init__()
        self.flatten = nn.Flatten()
        self.linear_relu_stack = nn.Sequential(
            nn.Linear(28*28, 512),
            nn.ReLU(),
            nn.Linear(512, 512),
            nn.ReLU(),
            nn.Linear(512, 10)
        )

    def forward(self, x):
        x = self.flatten(x)
        logits = self.linear_relu_stack(x)
        return logits

model = NeuralNetwork().to(device) # model指定GPU执行
print(model)

def train(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)
    model.train()
    for batch, (X, y) in enumerate(dataloader):
        X, y = X.to(device), y.to(device) # tensor指定GPU执行

        # Compute prediction error
        pred = model(X)
        loss = loss_fn(pred, y)

        # Backpropagation
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        if batch % 100 == 0:
            loss, current = loss.item(), batch * len(X)
            print(f"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]")

相关资料

Ptyroch Cuda官网