python - 类型为 torch.FloatTensor 的预期对象，但发现参数类型为 torch.cuda.FloatTensor #2 'weight'

Question

首先，我使用类似“model.cuda()”的方法将模型和数据转换为 cuda。但是它仍然存在这样的问题。我调试模型的每一层，每个模块的权重都有 iscuda=True。那么有谁知道为什么会出现这样的问题呢？

我有两个模型，一个是 resnet50，另一个包含第一个模型作为主干。

class FC_Resnet(nn.Module):
    def __init__(self, model, num_classes):
        super(FC_Resnet, self).__init__()

        # feature encoding
        self.features = nn.Sequential(
            model.conv1,
            model.bn1,
            model.relu,
            model.maxpool,
            model.layer1,
            model.layer2,
            model.layer3,
            model.layer4)

        # classifier
        num_features = model.layer4[1].conv1.in_channels
        self.classifier = nn.Sequential(
            nn.Conv2d(num_features, num_classes, kernel_size=1, bias=True))

    def forward(self, x):
        # children=self.features.children()
        # for child in children:
        #     if child.weight is not None:
        #         print(child.weight.device)
        x = self.features(x)
        x = self.classifier(x)
        return x

def fc_resnet50(num_classes=20, pre_trained=True):
    model = FC_Resnet(models.resnet50(pre_trained), num_classes)

    return model

还有一个:

class PeakResponseMapping(nn.Sequential):
    def __init__(self, *args, **kargs):
        super(PeakResponseMapping, self).__init__(*args)
        ...

    def forward(self, input, class_threshold=0, peak_threshold=30, retrieval_cfg=None):
        assert input.dim() == 4
        if self.inferencing:
            input.requires_grad_()

        class_response_maps = super(PeakResponseMapping, self).forward(input)

        return class_response_maps

主要的很简单:

def main():
    dataset = VOC(img_transform=image_transform())
    dataloader = DataLoader(dataset, batch_size=BATCH_SIZE, shuffle=True)

    model = peak_response_mapping(fc_resnet50(), win_size=3, sub_pixel_locating_factor=8, enable_peak_stimulation=True)
    model=model.cuda()

    for step, (b_x, b_y) in enumerate(dataloader):
        b_x.cuda()
        b_y.cuda()

        result = model.forward(b_x)

Answer 1

在堆栈跟踪的某处，Torch 需要一个 CPU 张量 (torch.FloatTensor)，但得到一个 GPU/CUDA 张量 (torch.cuda.FloatTensor) .

给定一个张量tensor:

• tensor.to('cpu') 返回张量的 CPU 版本
• tensor.to('cuda') 返回张量的 CUDA 版本

编写与硬件无关的代码:

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

然后你可以这样做:

tensor.to(device)

对于 OP，这变成:

result = model.forward(b_x.to(device))