Autoencoder

• MNIST
• Neural Network
• 1 hidden layers

!pip install torch torchvision

Requirement already satisfied: torch in /usr/local/lib/python3.6/dist-packages (1.1.0)
Requirement already satisfied: torchvision in /usr/local/lib/python3.6/dist-packages (0.3.0)
Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from torch) (1.16.4)
Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from torchvision) (1.12.0)
Requirement already satisfied: pillow>=4.1.1 in /usr/local/lib/python3.6/dist-packages (from torchvision) (4.3.0)
Requirement already satisfied: olefile in /usr/local/lib/python3.6/dist-packages (from pillow>=4.1.1->torchvision) (0.46)

1. Settings

1) Import required libraries

import numpy as np
import torch
import torch.nn as nn
import torch.optim as optim
import torch.nn.init as init
import torchvision.datasets as dset
import torchvision.transforms as transforms
from torch.utils.data import DataLoader
import matplotlib.pyplot as plt

2) Set hyperparameters

batch_size = 256
learning_rate = 0.0002
num_epoch = 5

2. Data

1) Download Data

mnist_train = dset.MNIST("./", train=True, transform=transforms.ToTensor(), target_transform=None, download=True)
mnist_test = dset.MNIST("./", train=False, transform=transforms.ToTensor(), target_transform=None, download=True)

2) Set DataLoader

train_loader = torch.utils.data.DataLoader(mnist_train,batch_size=batch_size, shuffle=True,num_workers=2,drop_last=True)
test_loader = torch.utils.data.DataLoader(mnist_test,batch_size=batch_size, shuffle=False,num_workers=2,drop_last=True)

4. Model & Optimizer

1) Model

# Àΰø½Å°æ¸ÁÀ¸·Î ÀÌ·ç¾îÁø ¿ÀÅä¿£ÄÚ´õ¸¦ »ý¼ºÇÕ´Ï´Ù.
# ´Ü¼øÇÏ°Ô Çϱâ À§ÇØ È°¼ºÈ­ ÇÔ¼ö´Â »ý·«Çß½À´Ï´Ù.
class Autoencoder(nn.Module):
    def __init__(self):
        super(Autoencoder,self).__init__()
        self.encoder = nn.Linear(28*28,20)
        self.decoder = nn.Linear(20,28*28)   
                
    def forward(self,x):
        x = x.view(batch_size,-1)
        encoded = self.encoder(x)
        out = self.decoder(encoded).view(batch_size,1,28,28)
        return out

2) Loss func & Optimizer

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print(device)

model = Autoencoder().to(device)
loss_func = nn.MSELoss()
optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)

cuda:0

5. Train

loss_arr =[]
for i in range(num_epoch):
    for j,[image,label] in enumerate(train_loader):
        x = image.to(device)
        
        optimizer.zero_grad()
        output = model.forward(x)
        loss = loss_func(output,x)
        loss.backward()
        optimizer.step()
        
    if j % 1000 == 0:
        print(loss)
        loss_arr.append(loss.cpu().data.numpy()[0])

6. Check with Train Image

out_img = torch.squeeze(output.cpu().data)
print(out_img.size())

for i in range(10):
    plt.imshow(torch.squeeze(image[i]).numpy(),cmap='gray')
    plt.show()
    plt.imshow(out_img[i].numpy(),cmap='gray')
    plt.show()

torch.Size([256, 28, 28])

with torch.no_grad():
  for i in range(1):
      for j,[image,label] in enumerate(test_loader):
          x = image.to(device)

          optimizer.zero_grad()
          output = model.forward(x)

      if j % 1000 == 0:
          print(loss)        

out_img = torch.squeeze(output.cpu().data)
print(out_img.size())

for i in range(10):
    plt.imshow(torch.squeeze(image[i]).numpy(),cmap='gray')
    plt.show()
    plt.imshow(out_img[i].numpy(),cmap='gray')
    plt.show()

torch.Size([256, 28, 28])