我的开源
pytorch之Fashion MNIST 图像分类
背景
学习一下如何使用pytorch来处理问题。
这里用Fashion MNIST例子。
数据集介绍
本文略过，可以参见前面的文章。
初体验
FC网络
输入图像大小为28x28，设计如下全连接网络：
INPUT(784) + FC1(1000) + Relu1 + FC2(500) + Relu2 + FC3(200) + Relu3 + FC4(10) + log_softmax
from __future__ import print_function # 从future版本导入print函数功能
import argparse # 加载处理命令行参数的库
import torch # 引入相关的包
import torch.nn as nn # 指定torch.nn别名nn
import torch.nn.functional as F # 引用神经网络常用函数包，不具有可学习的参数
import torch.optim as optim
from torchvision import datasets, transforms # 加载pytorch官方提供的dataset
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
self.fc1 = nn.Linear(784, 1000) # 784表示输入神经元数量，1000表示输出神经元数量
self.fc2 = nn.Linear(1000, 500)
self.fc3 = nn.Linear(500, 200)
self.fc4 = nn.Linear(200, 10)
def forward(self, x):
x = x.view(-1, 28*28)
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = F.relu(self.fc3(x))
x = self.fc4(x)
# Applies a softmax followed by a logarithm, output batch * classes tensor
return F.log_softmax(x, dim=1)
def train(args, model, device, train_loader, optimizer, epoch):
model.train()
for batch_idx, (data, target) in enumerate(train_loader):
data, target = data.to(device), target.to(device)
optimizer.zero_grad()
output = model(data)
# negative log likelihood loss(nll_loss), sum up batch cross entropy
loss = F.nll_loss(output, target)
loss.backward()
optimizer.step() # 根据parameter的梯度更新parameter的值
#print(epoch, batch_idx, type(batch_idx))
if batch_idx % args.log_interval == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
def test(args, model, device, test_loader):
model.eval()
test_loss = 0
correct = 0
with torch.no_grad(): # 无需计算梯度
for data, target in test_loader:
data, target = data.to(device), target.to(device)
output = model(data)
# sum up batch loss
test_loss += F.nll_loss(output, target, reduction='sum').item()
# get the index of the max log-probability
pred = output.argmax(dim=1, keepdim=True)
correct += pred.eq(target.view_as(pred)).sum().item()
test_loss /= len(test_loader.dataset)
print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
test_loss, correct, len(test_loader.dataset),
100. * correct / len(test_loader.dataset)))
def main():
# Training settings
parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
parser.add_argument('--batch-size', type=int, default=64, metavar='N',
help='input batch size for training (default: 64)')
parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N',
help='input batch size for testing (default: 1000)')
parser.add_argument('--epochs', type=int, default=10, metavar='N',
help='number of epochs to train (default: 10)')
parser.add_argument('--lr', type=float, default=0.01, metavar='LR',
help='learning rate (default: 0.01)')
parser.add_argument('--momentum', type=float, default=0.5, metavar='M',
help='SGD momentum (default: 0.5)')
parser.add_argument('--no-cuda', action='store_true', default=False,
help='disables CUDA training')
parser.add_argument('--seed', type=int, default=1, metavar='S',
help='random seed (default: 1)')
parser.add_argument('--log-interval', type=int, default=10, metavar='N',
help='how many batches to wait before logging training status')
parser.add_argument('--save-model', action='store_true', default=True,
help='For Saving the current Model')
args = parser.parse_args()
use_cuda = not args.no_cuda and torch.cuda.is_available()
torch.manual_seed(args.seed)
device = torch.device("cuda" if use_cuda else "cpu")
kwargs = {'num_workers': 1, 'pin_memory': True} if use_cuda else {}
train_loader = torch.utils.data.DataLoader(
datasets.FashionMNIST('./fashionmnist_data/', train=True, download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.batch_size, shuffle=True, **kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.FashionMNIST('./fashionmnist_data/', train=False, transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.test_batch_size, shuffle=True, **kwargs)
model = Net().to(device)
# optimizer存储了所有parameters的引用，每个parameter都包含gradient
optimizer = optim.SGD(model.parameters(), lr=args.lr,
momentum=args.momentum)
scheduler = optim.lr_scheduler.MultiStepLR(
optimizer, milestones=[12, 24], gamma=0.1) # 学习率按区间更新
for epoch in range(1, args.epochs + 1):
train(args, model, device, train_loader, optimizer, epoch)
test(args, model, device, test_loader)
if (args.save_model):
torch.save(model.state_dict(), "mnist_fc.pt")
# 当.py文件直接运行时，该语句及以下的代码被执行，当.py被调用时，该语句及以下的代码不被执行
if __name__ == '__main__':
main()
结果
Using downloaded and verified file: ./fashionmnist_data/FashionMNIST/raw/train-images-idx3-ubyte.gz
Extracting ./fashionmnist_data/FashionMNIST/raw/train-images-idx3-ubyte.gz to ./fashionmnist_data/FashionMNIST/raw
Using downloaded and verified file: ./fashionmnist_data/FashionMNIST/raw/train-labels-idx1-ubyte.gz
Extracting ./fashionmnist_data/FashionMNIST/raw/train-labels-idx1-ubyte.gz to ./fashionmnist_data/FashionMNIST/raw
Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz to ./fashionmnist_data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz
4423680it [00:05, 746242.74it/s]
Extracting ./fashionmnist_data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz to ./fashionmnist_data/FashionMNIST/raw
Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz to ./fashionmnist_data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz
8192it [00:00, 16499.39it/s]
Extracting ./fashionmnist_data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz to ./fashionmnist_data/FashionMNIST/raw
Processing...
/Users/xuqian/Library/Python/3.7/lib/python/site-packages/torchvision/datasets/mnist.py:469: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at ../torch/csrc/utils/tensor_numpy.cpp:141.)
return torch.from_numpy(parsed.astype(m[2], copy=False)).view(*s)
Done!
Train Epoch: 1 [0/60000 (0%)] Loss: 2.293269
Train Epoch: 1 [640/60000 (1%)] Loss: 2.262661
Train Epoch: 1 [1280/60000 (2%)] Loss: 2.207895
Train Epoch: 1 [1920/60000 (3%)] Loss: 2.168133
…………
Train Epoch: 10 [57600/60000 (96%)] Loss: 0.288563
Train Epoch: 10 [58240/60000 (97%)] Loss: 0.152729
Train Epoch: 10 [58880/60000 (98%)] Loss: 0.212459
Train Epoch: 10 [59520/60000 (99%)] Loss: 0.240333
Test set: Average loss: 0.3357, Accuracy: 8802/10000 (88%)
即训练结果top1准确率为88%，模型大小为5.3M。
CNN网络
FC网络参数量太大，而CNN网络考虑到图像的局部关联特性，使用卷积网络，参数量大大减小，设计如下CNN：
conv(1, 20, 5) + Relu + conv(20, 50, 5) + flatten + Relu + FC(10) + log_softmax
from __future__ import print_function # 从future版本导入print函数功能
import argparse # 加载处理命令行参数的库
import torch # 引入相关的包
import torch.nn as nn # 指定torch.nn别名nn
import torch.nn.functional as F # 引用神经网络常用函数包，不具有可学习的参数
import torch.optim as optim
from torchvision import datasets, transforms # 加载pytorch官方提供的dataset
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
# 1表示输入通道，20表示输出通道，5表示conv核大小，1表示conv步长
self.conv1 = nn.Conv2d(1, 20, 5, 1)
self.conv2 = nn.Conv2d(20, 50, 5, 1)
self.fc1 = nn.Linear(4 * 4 * 50, 500)
self.fc2 = nn.Linear(500, 10)
def forward(self, x):
x = F.relu(self.conv1(x))
x = F.max_pool2d(x, 2, 2)
x = F.relu(self.conv2(x))
x = F.max_pool2d(x, 2, 2)
x = x.view(-1, 4 * 4 * 50)
x = F.relu(self.fc1(x))
x = self.fc2(x)
return F.log_softmax(x, dim=1)
def train(args, model, device, train_loader, optimizer, epoch):
model.train()
for batch_idx, (data, target) in enumerate(train_loader):
data, target = data.to(device), target.to(device)
optimizer.zero_grad()
output = model(data)
# negative log likelihood loss(nll_loss), sum up batch cross entropy
loss = F.nll_loss(output, target)
loss.backward()
optimizer.step() # 根据parameter的梯度更新parameter的值
#print(epoch, batch_idx, type(batch_idx))
if batch_idx % args.log_interval == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
def test(args, model, device, test_loader):
model.eval()
test_loss = 0
correct = 0
with torch.no_grad(): # 无需计算梯度
for data, target in test_loader:
data, target = data.to(device), target.to(device)
output = model(data)
# sum up batch loss
test_loss += F.nll_loss(output, target, reduction='sum').item()
# get the index of the max log-probability
pred = output.argmax(dim=1, keepdim=True)
correct += pred.eq(target.view_as(pred)).sum().item()
test_loss /= len(test_loader.dataset)
print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
test_loss, correct, len(test_loader.dataset),
100. * correct / len(test_loader.dataset)))
def main():
# Training settings
parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
parser.add_argument('--batch-size', type=int, default=64, metavar='N',
help='input batch size for training (default: 64)')
parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N',
help='input batch size for testing (default: 1000)')
parser.add_argument('--epochs', type=int, default=10, metavar='N',
help='number of epochs to train (default: 10)')
parser.add_argument('--lr', type=float, default=0.01, metavar='LR',
help='learning rate (default: 0.01)')
parser.add_argument('--momentum', type=float, default=0.5, metavar='M',
help='SGD momentum (default: 0.5)')
parser.add_argument('--no-cuda', action='store_true', default=False,
help='disables CUDA training')
parser.add_argument('--seed', type=int, default=1, metavar='S',
help='random seed (default: 1)')
parser.add_argument('--log-interval', type=int, default=10, metavar='N',
help='how many batches to wait before logging training status')
parser.add_argument('--save-model', action='store_true', default=True,
help='For Saving the current Model')
args = parser.parse_args()
use_cuda = not args.no_cuda and torch.cuda.is_available()
torch.manual_seed(args.seed)
device = torch.device("cuda" if use_cuda else "cpu")
kwargs = {'num_workers': 1, 'pin_memory': True} if use_cuda else {}
train_loader = torch.utils.data.DataLoader(
datasets.FashionMNIST('./fashionmnist_data/', train=True, download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.batch_size, shuffle=True, **kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.FashionMNIST('./fashionmnist_data/', train=False, transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.test_batch_size, shuffle=True, **kwargs)
model = Net().to(device)
# optimizer存储了所有parameters的引用，每个parameter都包含gradient
optimizer = optim.SGD(model.parameters(), lr=args.lr,
momentum=args.momentum)
scheduler = optim.lr_scheduler.MultiStepLR(
optimizer, milestones=[12, 24], gamma=0.1) # 学习率按区间更新
for epoch in range(1, args.epochs + 1):
train(args, model, device, train_loader, optimizer, epoch)
test(args, model, device, test_loader)
if (args.save_model):
torch.save(model.state_dict(), "mnist_cnn.pt")
# 当.py文件直接运行时，该语句及以下的代码被执行，当.py被调用时，该语句及以下的代码不被执行
if __name__ == '__main__':
main()
结果
Train Epoch: 1 [0/60000 (0%)] Loss: 2.293269
Train Epoch: 1 [640/60000 (1%)] Loss: 2.262661
Train Epoch: 1 [1280/60000 (2%)] Loss: 2.207895
Train Epoch: 1 [1920/60000 (3%)] Loss: 2.168133
…………
Train Epoch: 10 [56960/60000 (95%)] Loss: 0.345022
Train Epoch: 10 [57600/60000 (96%)] Loss: 0.293099
Train Epoch: 10 [58240/60000 (97%)] Loss: 0.098339
Train Epoch: 10 [58880/60000 (98%)] Loss: 0.137070
Train Epoch: 10 [59520/60000 (99%)] Loss: 0.154875
Test set: Average loss: 0.2843, Accuracy: 8996/10000 (90%)
即训练结果top1准确率为90%，模型大小为1.7M。