import torch
from torch import nn
from d2l import torch as d2l
net = nn.Sequential(
# (224-11+1+2)/4=54
nn.Conv2d(1,96,kernel_size=11,stride=4,padding=1),nn.ReLU(),
# (54-3+1)/2=26
nn.MaxPool2d(kernel_size=3,stride=2),
# (26+4-5+1)=26
nn.Conv2d(96,256,kernel_size=5,padding=2),nn.ReLU(),
# (26-3+1)/2=12
nn.MaxPool2d(kernel_size=3,stride=2),
# 12-3+1+2=12
nn.Conv2d(256,384,kernel_size=3,padding=1),nn.ReLU(),
# 12-3+1+2=12
nn.Conv2d(384,384,kernel_size=3,padding=1),nn.ReLU(),
# 12+2-3+1=12
nn.Conv2d(384, 384, kernel_size=3, padding=1), nn.ReLU(),
# 12-3+1+2=12
nn.Conv2d(384,256,kernel_size=3,padding=1),nn.ReLU(),
# (12-3+1)/2=5
nn.MaxPool2d(kernel_size=3,stride=2),nn.Flatten(),
# 256*5*5=6400
nn.Linear(6400,4096),nn.ReLU(),nn.Dropout(p=0.5),
nn.Linear(4096,4096),nn.ReLU(),nn.Dropout(p=0.5),
nn.Linear(4096,10)
)
x=torch.randn(1,1,224,224)
for layer in net:
x=layer(x)
print(layer.__class__.__name__,'output shape:\t',x.shape)
batch_size=128
train_iter,test_iter = d2l.load_data_fashion_mnist(batch_size=batch_size,resize=224)
lr,num_epochs = 0.01,10
d2l.train_ch6(net,train_iter,test_iter,num_epochs,lr,d2l.try_gpu())
标签:kernel,12,nn,卷积,ReLU,padding,pytorch,AlexNet,size From: https://www.cnblogs.com/jinbb/p/17609409.html