一、pytorch搭建实战以及sequential的使用
一、pytorch搭建实战以及sequential的使用
- 1.A sequential container
- 2.搭建cifar10 model structure
- 3.创建实例进行测试（可以检查网络是否正确）
- 3.tensorboard图可视化
1.A sequential container
官网说明文档 ： https://pytorch.org/docs/stable/generated/torch.nn.Sequential.html#sequential
Example
model = nn.Sequential(
nn.Conv2d(1,20,5),
nn.ReLU(),
nn.Conv2d(20,64,5),
nn.ReLU()
)
# Using Sequential with OrderedDict. This is functionally the
# same as the above code
model = nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(1,20,5)),
('relu1', nn.ReLU()),
('conv2', nn.Conv2d(20,64,5)),
('relu2', nn.ReLU())
]))
2.搭建cifar10 model structure
共有9层网络结构，顺序如下：
(0): Conv2d(3, 32, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))
(1): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(2): Conv2d(32, 32, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))
(3): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(4): Conv2d(32, 64, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))
(5): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(6): Flatten(start_dim=1, end_dim=-1)
(7): Linear(in_features=1024, out_features=64, bias=True)
(8): Linear(in_features=64, out_features=10, bias=True)
搭建网络结构：
class Qu(nn.Module):
def __init__(self):
super(Qu, self).__init__()
self.model1 = Sequential(
Conv2d(3, 32, 5, padding=2),
MaxPool2d(2),
Conv2d(32, 32, 5, padding=2),
MaxPool2d(2),
Conv2d(32, 64, 5, padding=2),
MaxPool2d(2),
Flatten(),
Linear(1024, 64),
Linear(64, 10)
)
def forward(self, x):
x = self.model1(x)
return x
其中
torch.nn.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode=‘zeros’, device=None, dtype=None)
例如第一个卷积层conv2d的 输入通道=3，输出通道=32，卷积核个数=5x5, padding=2
由于卷积作用后HW不变，仍为32X32，故需计算padding,padding的计算方法如下：
3.创建实例进行测试（可以检查网络是否正确）
qu = Qu()
#print(qu)
input = torch.ones(64, 3, 32, 32)
output = qu(input)
print(output.shape)
输出结果
torch.Size([64, 10])
3.tensorboard图可视化
writer = SummaryWriter("./logs_seq")
writer.add_graph(qu, input)
writer.close()
