x = x.view(x.size(0), -1) 一般出现在前向传播过程中卷积层与全连接层的交替的位置。下面代码为Lenet网络训练cifar10数据集的情况。
class LeNet(nn.Module):
def init(self):
super(LeNet, self).init()
self.conv1 = nn.Conv2d(3, 6, 5)
self.conv2 = nn.Conv2d(6, 16, 5)
self.fc1 = nn.Linear(1655, 120)
self.fc2 = nn.Linear(120, 84)
self.fc3 = nn.Linear(84, 10)
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
out = out.view(out.size(0), -1)中out.size(0) 的out大小为[batchsize,16,5,5],因此out.size(0) = batchsize;如果为out.size(1), 则 = 16。我自己训练时batchsize设置为32,所以out.size(0) = 32。
out.view(out.size(0),-1)则是把out转化为行数为out.size(0) ,列数相应自动生成(对于我的代码列数则为1655 = 400)。
out = out.view(out.size(0), -1)函数功能相当于torch.flatten()