PyTorch 第一个简单例子

import torch.nn as nn
import torch.nn as nn
import torch.nn.functional as F
from torch.autograd import Variable 
import torch.optim as optim

class Net(nn.Module):
   def __init__(self):
    super(Net, 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, 10)

   def forward( self, x):
     x = F.max_pool2d(F.relu(self.conv1(x)), (2,2))
     x = F.max_pool2d(F.relu(self.conv2(x)), 2)
     x = x.view(x.size()[0], -1)  #reshape '-1'means auto
     x = F.relu(self.fc1(x))
     x = F.relu(self.fc2(x))
     x = self.fc3(x)
    return x

input = Variable(t.randn(1,1,32,32))
net = Net()

#print('net.conv1.bias.grad')
#print(net.conv1.bias.grad)
weight = 0
learning_rate = 0.01
#weight = weight - learning_rate*gradient
optimizer = optim.SGD(net.parameters(), lr = 0.01)
optimizer.zero_grad()
output = net(input)
#print(out.size)
#print(net)
target = Variable(t.arange(0,10))
target = target.view([1,10])

criterion = nn.MSELoss()
loss = criterion(output, target)
net.zero_grad()
#print('net.conv1.bias.grad')
#print(net.conv1.bias.grad)
loss.backward()
for var in net.parameters():
    #print(var)
    var.data.sub_(var.grad.data * learning_rate) #inplace

optimizer.step()