聽說pytorch使用比TensorFlow簡單，加之pytorch現已支援windows，所以今天裝了pytorch玩玩，第一件事還是寫了個簡單的CNN在MNIST上實驗，初步體驗的確比TensorFlow方便。

參考程式碼（在莫煩python的教程程式碼基礎上修改）如下：

import torch  
import torch.nn as nn  
from torch.autograd import Variable  
import torch.utils.data as Data  
import torchvision  
import time
#import matplotlib.pyplot as plt  
  
torch.manual_seed(1)  
  
EPOCH = 1  
BATCH_SIZE = 50  
LR = 0.001  
DOWNLOAD_MNIST = False  
if_use_gpu = 1
  
# 獲取訓練集dataset  
training_data = torchvision.datasets.MNIST(  
             root='./mnist/', # dataset儲存路徑  
             train=True, # True表示是train訓練集，False表示test測試集  
             transform=torchvision.transforms.ToTensor(), # 將原資料規範化到（0,1）區間  
             download=DOWNLOAD_MNIST,  
             )  
  
# 列印MNIST資料集的訓練集及測試集的尺寸  
print(training_data.train_data.size())  
print(training_data.train_labels.size())  
# torch.Size([60000, 28, 28])  
# torch.Size([60000])  
  
#plt.imshow(training_data.train_data[0].numpy(), cmap='gray')  
#plt.title('%i' % training_data.train_labels[0])  
#plt.show()  
  
# 通過torchvision.datasets獲取的dataset格式可直接可置於DataLoader  
train_loader = Data.DataLoader(dataset=training_data, batch_size=BATCH_SIZE,  
                               shuffle=True)  
  
# 獲取測試集dataset  

test_data = torchvision.datasets.MNIST(  
             root='./mnist/', # dataset儲存路徑  
             train=False, # True表示是train訓練集，False表示test測試集  
             transform=torchvision.transforms.ToTensor(), # 將原資料規範化到（0,1）區間  
             download=DOWNLOAD_MNIST,  
             )  
# 取前全部10000個測試集樣本  
test_x = Variable(torch.unsqueeze(test_data.test_data, dim=1).float(), requires_grad=False)
#test_x = test_x.cuda()
## (~, 28, 28) to (~, 1, 28, 28), in range(0,1)  
test_y = test_data.test_labels
#test_y = test_y.cuda()  
class CNN(nn.Module):  
    def __init__(self):  
        super(CNN, self).__init__()  
        self.conv1 = nn.Sequential( # (1,28,28)  
                     nn.Conv2d(in_channels=1, out_channels=16, kernel_size=5,  
                               stride=1, padding=2), # (16,28,28)  
        # 想要con2d卷積出來的圖片尺寸沒有變化, padding=(kernel_size-1)/2  
                     nn.ReLU(),  
                     nn.MaxPool2d(kernel_size=2) # (16,14,14)  
                     )  
        self.conv2 = nn.Sequential( # (16,14,14)  
                     nn.Conv2d(16, 32, 5, 1, 2), # (32,14,14)  
                     nn.ReLU(),  
                     nn.MaxPool2d(2) # (32,7,7)  
                     )  
        self.out = nn.Linear(32*7*7, 10)  
  
    def forward(self, x):  
        x = self.conv1(x)  
        x = self.conv2(x)  
        x = x.view(x.size(0), -1) # 將（batch，32,7,7）展平為（batch，32*7*7）  
        output = self.out(x)  
        return output  
  
cnn = CNN()  
if if_use_gpu:
    cnn = cnn.cuda()

optimizer = torch.optim.Adam(cnn.parameters(), lr=LR)  
loss_function = nn.CrossEntropyLoss()  
 


for epoch in range(EPOCH):  
    start = time.time() 
    for step, (x, y) in enumerate(train_loader):  
        b_x = Variable(x, requires_grad=False) 
        b_y = Variable(y, requires_grad=False)  
        if if_use_gpu:
            b_x = b_x.cuda()
            b_y = b_y.cuda()
  
        output = cnn(b_x)  
        loss = loss_function(output, b_y)  
        optimizer.zero_grad()  
        loss.backward()  
        optimizer.step()  
  
        if step % 100 == 0:  
            print('Epoch:', epoch, '|Step:', step,  
                  '|train loss:%.4f'%loss.data[0])  
    duration = time.time() - start 
    print('Training duation: %.4f'%duration)
    
cnn = cnn.cpu()
test_output = cnn(test_x)  
pred_y = torch.max(test_output, 1)[1].data.squeeze()
accuracy = sum(pred_y == test_y) / test_y.size(0) 
print('Test Acc: %.4f'%accuracy)