如何修复 pytorch 神经网络中不正确的通道大小？

Question

我正在使用频谱图形式的 Google 话语数据集。每个数据点都有维度 (160, 101)。在我的数据加载器中，我使用了 batch_size=128。因此，每个批次的维度为 (128, 160, 101)。

我使用了一个LeNet模型作为模型，代码如下：

class LeNet(nn.Module):
    def __init__(self):
        super(LeNet, self).__init__()
        self.conv1 = nn.Conv2d(1, 6, 5)
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1   = nn.Linear(16*5*5, 120)
        self.fc2   = nn.Linear(120, 84)
        self.fc3   = nn.Linear(84, 30)

    def forward(self, x):
        out = F.relu(self.conv1(x))
        out = F.max_pool2d(out, 2)
        out = F.relu(self.conv2(out))
        out = F.max_pool2d(out, 2)
        out = out.view(out.size(0), -1)
        out = F.relu(self.fc1(out))
        out = F.relu(self.fc2(out))
        out = self.fc3(out)
        return out

我尝试使用 dim=3 解压缩数据，但出现此错误：

Traceback (most recent call last):
  File "train_speech.py", line 359, in <module>
    train_loss, reg_loss, train_acc, cost = train(epoch)
  File "train_speech.py", line 258, in train
    outputs = (net(inputs))['out']
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/parallel/data_parallel.py", line 166, in forward
    return self.module(*inputs[0], **kwargs[0])
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/content/gdrive/My Drive/Colab Notebooks/mixup_erm-master/models/lenet.py", line 15, in forward
    out = F.relu(self.conv1(x))
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py", line 1051, in _call_impl
    return forward_call(*input, **kwargs)
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/conv.py", line 443, in forward
    return self._conv_forward(input, self.weight, self.bias)
  File "/usr/local/lib/python3.7/dist-packages/torch/nn/modules/conv.py", line 440, in _conv_forward
    self.padding, self.dilation, self.groups)
RuntimeError: Given groups=1, weight of size [6, 1, 5, 5], expected input[128, 160, 101, 1] to have 1 channels, but got 160 channels instead

我该如何解决这个问题？

编辑：下面的新错误消息

torch.Size([128, 160, 101])
torch.Size([128, 1, 160, 101])
/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:718: UserWarning: Named tensors and all their associated APIs are an experimental feature and subject to change. Please do not use them for anything important until they are released as stable. (Triggered internally at  /pytorch/c10/core/TensorImpl.h:1156.)
  return torch.max_pool2d(input, kernel_size, stride, padding, dilation, ceil_mode)
Traceback (most recent call last):
  File "train_speech.py", line 363, in <module>
    train_loss, reg_loss, train_acc, cost = train(epoch)
  File "train_speech.py", line 262, in train
    outputs = (net(inputs))['out']
IndexError: too many indices for tensor of dimension 2

我正在解压每批数据。我的培训代码的相关部分如下。 inputs 类似于 x.

print(inputs.shape)
inputs = inputs.unsqueeze(1)
print(inputs.shape)
outputs = (net(inputs))['out']

编辑 2：新错误

Traceback (most recent call last):
  File "train_speech.py", line 361, in <module>
    train_loss, reg_loss, train_acc, cost = train(epoch)
  File "train_speech.py", line 270, in train
    loss.backward()
  File "/usr/local/lib/python3.7/dist-packages/torch/_tensor.py", line 255, in backward
    torch.autograd.backward(self, gradient, retain_graph, create_graph, inputs=inputs)
  File "/usr/local/lib/python3.7/dist-packages/torch/autograd/__init__.py", line 149, in backward
    allow_unreachable=True, accumulate_grad=True)  # allow_unreachable flag
RuntimeError: Function AddmmBackward returned an invalid gradient at index 1 - got [128, 400] but expected shape compatible with [128, 13024]

编辑 3：在下方训练循环

def train(epoch):
    print('\nEpoch: %d' % epoch)
    net.train()
    train_loss = 0
    reg_loss = 0
    correct = 0
    total = 0
    for batch_idx, (inputs, targets) in enumerate(trainloader):
        if use_cuda:
            inputs, targets = inputs.cuda(), targets.cuda()
        inputs, targets_a, targets_b, lam,layer, cost = mixup_data(inputs, targets,
                                                       args.alpha,args.mixupBatch, use_cuda)
        inputs, targets_a, targets_b = map(Variable, (inputs,
                                                      targets_a, targets_b))
        outputs = net(inputs)
        loss = mixup_criterion(criterion, outputs, targets_a, targets_b, lam)
    
        train_loss += loss.item()
        _, predicted = torch.max(outputs.data, 1)
        total += targets.size(0)
        correct += (lam * predicted.eq(targets_a.data).cpu().sum().float()
                    + (1 - lam) * predicted.eq(targets_b.data).cpu().sum().float())
    
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

    return (train_loss/batch_idx, reg_loss/batch_idx, 100.*correct/total, cost/batch_idx)

Answer 1

您应该在 axis=1 a.k.a. 频道轴上展开：

>>> x = x.unsqueeze(1)

如果您在数据集 __getitem__ 内，则它对应于 axis=0。

如何修复 pytorch 神经网络中不正确的通道大小？

How to fix incorrect channel size in pytorch neural network?

neural-network

conv-neural-network

pytorch