Pytorch 自动编码器打印精度位置参数错误

如何解决Pytorch 自动编码器打印精度位置参数错误

我得到了这个自动编码器

class AE(nn.Module):
    def __init__(self,**kwargs):
        super(AE,self).__init__()
        self.encoder_hidden_layer = nn.Linear(
            in_features=kwargs["input_shape"],out_features=128
        )
        self.encoder_output_layer = nn.Linear(
            in_features=128,out_features=128
        )
        self.decoder_hidden_layer = nn.Linear(
            in_features=128,out_features=128
        )
        self.decoder_output_layer = nn.Linear(
            in_features=128,out_features=kwargs["input_shape"]
        )

    def forward(self,features):
        activation = self.encoder_hidden_layer(features)
        activation = torch.relu(activation)
        code = self.encoder_output_layer(activation)
        code = torch.sigmoid(code)
        activation = self.decoder_hidden_layer(code)
        activation = torch.relu(activation)
        activation = self.decoder_output_layer(activation)
        reconstructed = torch.sigmoid(activation)
        return reconstructed

我正在尝试构建一个准确度计算器

with torch.no_grad():
    for batch_features in test_loader:
        batch_features = batch_features[0]
        test_examples = batch_features.view(-1,784)
        reconstruction = model(test_examples)
        
        images,labels = data[0].to(device),data[1].to(device)
        print(labels)
        
        outputs_auto = AE(images.to(device))
        _,predicted_auto = torch.max(outputs_auto.data,1)
        total += labels.size(0)
        correct += (predicted_auto == labels).sum().item()
        break
    print('Accuracy of the network on the 10000 test images: %d %%' % (
    100 * correct / total))

但它显示了这种类型的错误。我曾尝试对代码进行一些修改，但失败了。我是自动编码器的新手

---> 21         outputs_auto = AE(images.to(device))
     22         _,1)
     23         total += labels.size(0)
TypeError: __init__() takes 1 positional argument but 2 were given

我只想以百分比形式打印准确度。如果您需要更多代码，请要求编辑此问题

解决方法

调用 outputs_auto = AE(images.to(device)) 时，您正在调用 AE 的 __init__ 函数。而您应该做的是调用模块的转发功能（即__call__）。因此，您应该首先初始化您的自动编码器，然后使用所需的输入调用它。您还需要根据 input_shape 模块的实现提供 AE。例如你可以这样做：

autoencoder = AE(input_shape=100).to(device)

这是让您开始使用虚拟数据进行测试循环和精度测量：

test_set = TensorDataset(torch.rand(16,28,28),torch.rand(16,1))
test_loader = DataLoader(test_set)
 
correct = 0
with torch.no_grad():
    for data in test_loader:
        images,labels = data[0].to(device),data[1].to(device)
        images = images.reshape(images.shape[0],-1)
        
        outputs_auto = autoencoder(images)
        predicted_class = torch.argmax(outputs_auto)
        correct += (predicted_class == labels).sum().item()
print('Accuracy of the network %.1f%%' % (100*correct/len(test_loader)))

用相关的数据加载器替换 test_loader。还要将 input_shape 更改为正确的大小。