动手深度学习(Pytorch)之路---第三次打卡
引言
今天分享两个深度学习实战任务：图像分类和文本分类。分别使用卷积神经网络和循环神经网络完成这次的实战任务。
图像分类
这里使用的数据集是CIFAR-10，比赛数据分为训练集和测试集。训练集包含 50,000 图片。测试集包含 300,000 图片。两个数据集中的图像格式均为PNG，高度和宽度均为32像素，并具有三个颜色通道（RGB）。图像涵盖10个类别：飞机，汽车，鸟类，猫，鹿，狗，青蛙，马，船和卡车。
import numpy as np
import torch
import torch.nn as nn
import torch.optim as optim
import torchvision
import torchvision.transforms as transforms
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
import os
import time
# 数据预处理
# 图像增强
transform_train = transforms.Compose([
transforms.RandomCrop(32, padding=4), #先四周填充0，再把图像随机裁剪成32*32
transforms.RandomHorizontalFlip(), #图像一半的概率翻转，一半的概率不翻转
transforms.ToTensor(),
transforms.Normalize((0.4731, 0.4822, 0.4465), (0.2212, 0.1994, 0.2010)), #R,G,B每层的归一化用到的均值和方差
])
transform_test = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4731, 0.4822, 0.4465), (0.2212, 0.1994, 0.2010)),
])
train_dir = '/cifar-10/train'
test_dir = '/cifar-10/test'
trainset = torchvision.datasets.ImageFolder(root=train_dir, transform=transform_train)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=256, shuffle=True)
testset = torchvision.datasets.ImageFolder(root=test_dir, transform=transform_test)
testloader = torch.utils.data.DataLoader(testset, batch_size=256, shuffle=False)
classes = ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'forg', 'horse', 'ship', 'truck']
这里使用了数据增强的思想。随机改变训练样本，如加入噪声，翻转等操作，可以降低模型对某些属性的依赖，从而提高模型的泛化能力。例如，我们可以对图像进行不同方式的裁剪，使感兴趣的物体出现在不同位置，从而减轻模型对物体出现位置的依赖性。我们也可以调整亮度、色彩等因素来降低模型对色彩的敏感度。
class ResidualBlock(nn.Module): # 我们定义网络时一般是继承的torch.nn.Module创建新的子类
def __init__(self, inchannel, outchannel, stride=1):
super(ResidualBlock, self).__init__()
#torch.nn.Sequential是一个Sequential容器，模块将按照构造函数中传递的顺序添加到模块中。
self.left = nn.Sequential(
nn.Conv2d(inchannel, outchannel, kernel_size=3, stride=stride, padding=1, bias=False),
# 添加第一个卷积层,调用了nn里面的Conv2d（）
nn.BatchNorm2d(outchannel), # 进行数据的归一化处理
nn.ReLU(inplace=True), # 修正线性单元，是一种人工神经网络中常用的激活函数
nn.Conv2d(outchannel, outchannel, kernel_size=3, stride=1, padding=1, bias=False),
nn.BatchNorm2d(outchannel)
)
self.shortcut = nn.Sequential()
if stride != 1 or inchannel != outchannel:
self.shortcut = nn.Sequential(
nn.Conv2d(inchannel, outchannel, kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(outchannel)
)
# 便于之后的联合,要判断Y = self.left(X)的形状是否与X相同
def forward(self, x): # 将两个模块的特征进行结合，并使用ReLU激活函数得到最终的特征。
out = self.left(x)
out += self.shortcut(x)
out = F.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, ResidualBlock, num_classes=10):
super(ResNet, self).__init__()
self.inchannel = 64
self.conv1 = nn.Sequential( # 用3个3x3的卷积核代替7x7的卷积核，减少模型参数
nn.Conv2d(3, 64, kernel_size=3, stride=1, padding=1, bias=False),
nn.BatchNorm2d(64),
nn.ReLU(),
)
self.layer1 = self.make_layer(ResidualBlock, 64, 2, stride=1)
self.layer2 = self.make_layer(ResidualBlock, 128, 2, stride=2)
self.layer3 = self.make_layer(ResidualBlock, 256, 2, stride=2)
self.layer4 = self.make_layer(ResidualBlock, 512, 2, stride=2)
self.fc = nn.Linear(512, num_classes)
def make_layer(self, block, channels, num_blocks, stride):
strides = [stride] + [1] * (num_blocks - 1) #第一个ResidualBlock的步幅由make_layer的函数参数stride指定
# ，后续的num_blocks-1个ResidualBlock步幅是1
layers = []
for stride in strides:
layers.append(block(self.inchannel, channels, stride))
self.inchannel = channels
return nn.Sequential(*layers)
def forward(self, x):
out = self.conv1(x)
out = self.layer1(out)
out = self.layer2(out)
out = self.layer3(out)
out = self.layer4(out)
out = F.avg_pool2d(out, 4)
out = out.view(out.size(0), -1)
out = self.fc(out)
return out
def ResNet18():
return ResNet(ResidualBlock)
ResNet-18网络结构：ResNet全名Residual Network残差网络。Kaiming He 的《Deep Residual Learning for Image Recognition》获得了CVPR最佳论文。他提出的深度残差网络在2015年可以说是洗刷了图像方面的各大比赛，以绝对优势取得了多个比赛的冠军。而且它在保证网络精度的前提下，将网络的深度达到了152层，后来又进一步加到1000的深度。
# 定义是否使用GPU
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
# 超参数设置
EPOCH = 20 #遍历数据集次数
pre_epoch = 0 # 定义已经遍历数据集的次数
LR = 0.1 #学习率
# 模型定义-ResNet
net = ResNet18().to(device)
# 定义损失函数和优化方式
criterion = nn.CrossEntropyLoss() #损失函数为交叉熵，多用于多分类问题
optimizer = optim.SGD(net.parameters(), lr=LR, momentum=0.9, weight_decay=5e-4)
#优化方式为mini-batch momentum-SGD，并采用L2正则化（权重衰减）
# 训练
if __name__ == "__main__":
print("Start Training, Resnet-18!")
num_iters = 0
for epoch in range(pre_epoch, EPOCH):
print('\nEpoch: %d' % (epoch + 1))
net.train()
sum_loss = 0.0
correct = 0.0
total = 0
for i, data in enumerate(trainloader, 0):
#用于将一个可遍历的数据对象(如列表、元组或字符串)组合为一个索引序列，同时列出数据和数据下标，
#下标起始位置为0，返回 enumerate(枚举) 对象。
num_iters += 1
inputs, labels = data
inputs, labels = inputs.to(device), labels.to(device)
optimizer.zero_grad() # 清空梯度
# forward + backward
outputs = net(inputs)
loss = criterion(outputs, labels)
loss.backward()
optimizer.step()
sum_loss += loss.item() * labels.size(0)
_, predicted = torch.max(outputs, 1) #选出每一列中最大的值作为预测结果
total += labels.size(0)
correct += (predicted == labels).sum().item()
# 每20个batch打印一次loss和准确率
if (i + 1) % 20 == 0:
print('[epoch:%d, iter:%d] Loss: %.03f | Acc: %.3f%% '
% (epoch + 1, num_iters, sum_loss / (i + 1), 100. * correct / total))
print("Training Finished, TotalEPOCH=%d" % EPOCH)
文本分类
文本分类是自然语言处理的一个常见任务，它把一段不定长的文本序列变换为文本的类别。这次的任务是根据文本判断作者的情绪，也就是一个二分类任务，使用LSTM模型。
使用斯坦福的IMDb数据集（Stanford’s Large Movie Review Dataset）作为文本情感分类的数据集。
import collections
import os
import random
import time
from tqdm import tqdm
import torch
from torch import nn
import torchtext.vocab as Vocab
import torch.utils.data as Data
import torch.nn.functional as F
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
def read_imdb(folder='train', data_root="/aclImdb_v1/aclImdb"):
data = []
for label in ['pos', 'neg']:
folder_name = os.path.join(data_root, folder, label)
for file in tqdm(os.listdir(folder_name)):
with open(os.path.join(folder_name, file), 'rb') as f:
review = f.read().decode('utf-8').replace('\n', '').lower()
data.append([review, 1 if label == 'pos' else 0])
random.shuffle(data)
return data
DATA_ROOT = "/aclImdb_v1/"
data_root = os.path.join(DATA_ROOT, "aclImdb")
train_data, test_data = read_imdb('train', data_root), read_imdb('test', data_root)
def get_tokenized_imdb(data):
'''
@params:
data: 数据的列表，列表中的每个元素为 [文本字符串，0/1标签] 二元组
@return: 切分词后的文本的列表，列表中的每个元素为切分后的词序列
'''
def tokenizer(text):
return [tok.lower() for tok in text.split(' ')]
return [tokenizer(review) for review, _ in data]
def get_vocab_imdb(data):
'''
@params:
data: 同上
@return: 数据集上的词典，Vocab 的实例（freqs, stoi, itos）
'''
tokenized_data = get_tokenized_imdb(data)
counter = collections.Counter([tk for st in tokenized_data for tk in st])
return Vocab.Vocab(counter, min_freq=5)
vocab = get_vocab_imdb(train_data)
def preprocess_imdb(data, vocab):
'''
@params:
data: 同上，原始的读入数据
vocab: 训练集上生成的词典
@return:
features: 单词下标序列，形状为 (n, max_l) 的整数张量
labels: 情感标签，形状为 (n,) 的0/1整数张量
'''
max_l = 500 # 将每条评论通过截断或者补0，使得长度变成500
def pad(x):
return x[:max_l] if len(x) > max_l else x + [0] * (max_l - len(x))
tokenized_data = get_tokenized_imdb(data)
features = torch.tensor([pad([vocab.stoi[word] for word in words]) for words in tokenized_data])
labels = torch.tensor([score for _, score in data])
return features, labels
train_set = Data.TensorDataset(*preprocess_imdb(train_data, vocab))
test_set = Data.TensorDataset(*preprocess_imdb(test_data, vocab))
# 上面的代码等价于下面的注释代码
# train_features, train_labels = preprocess_imdb(train_data, vocab)
# test_features, test_labels = preprocess_imdb(test_data, vocab)
# train_set = Data.TensorDataset(train_features, train_labels)
# test_set = Data.TensorDataset(test_features, test_labels)
# len(train_set) = features.shape[0] or labels.shape[0]
# train_set[index] = (features[index], labels[index])
batch_size = 64
train_iter = Data.DataLoader(train_set, batch_size, shuffle=True)
test_iter = Data.DataLoader(test_set, batch_size)
相较于图像来说，文本的数据预处理更加复杂。主要有以下几个步骤：
- 读入文本
- 分词。对每个句子进行分词，也就是将一个句子划分成若干个词（token），转换为一个词的序列。
- 建立字典，将每个词映射到一个唯一的索引（index）。为了方便模型处理，我们需要将字符串转换为数字。因此我们需要先构建一个字典（vocabulary），将每个词映射到一个唯一的索引编号。
- 将文本从词的序列转换为索引的序列，方便输入模型。使用字典，可以将原文本中的句子从单词序列转换为索引序列。
下面开始构建模型
class BiRNN(nn.Module):
def __init__(self, vocab, embed_size, num_hiddens, num_layers):
'''
@params:
vocab: 在数据集上创建的词典，用于获取词典大小
embed_size: 嵌入维度大小
num_hiddens: 隐藏状态维度大小
num_layers: 隐藏层个数
'''
super(BiRNN, self).__init__()
self.embedding = nn.Embedding(len(vocab), embed_size)
# encoder-decoder framework
# bidirectional设为True即得到双向循环神经网络
self.encoder = nn.LSTM(input_size=embed_size,
hidden_size=num_hiddens,
num_layers=num_layers,
bidirectional=True)
self.decoder = nn.Linear(4*num_hiddens, 2) # 初始时间步和最终时间步的隐藏状态作为全连接层输入
def forward(self, inputs):
'''
@params:
inputs: 词语下标序列，形状为 (batch_size, seq_len) 的整数张量
@return:
outs: 对文本情感的预测，形状为 (batch_size, 2) 的张量
'''
# 因为LSTM需要将序列长度(seq_len)作为第一维，所以需要将输入转置
embeddings = self.embedding(inputs.permute(1, 0)) # (seq_len, batch_size, d)
# rnn.LSTM 返回输出、隐藏状态和记忆单元，格式如 outputs, (h, c)
outputs, _ = self.encoder(embeddings) # (seq_len, batch_size, 2*h)
encoding = torch.cat((outputs[0], outputs[-1]), -1) # (batch_size, 4*h)
outs = self.decoder(encoding) # (batch_size, 2)
return outs
embed_size, num_hiddens, num_layers = 100, 100, 2
net = BiRNN(vocab, embed_size, num_hiddens, num_layers)
这里是加在预训练的词向量。它将每个词表示成一个定长的向量，并通过在语料库上的预训练使得这些向量能较好地表达不同词之间的相似和类比关系，以引入一定的语义信息，从而提升模型的学习能力。
cache_dir = "/home/kesci/input/GloVe6B5429"
glove_vocab = Vocab.GloVe(name='6B', dim=100, cache=cache_dir)
def load_pretrained_embedding(words, pretrained_vocab):
'''
@params:
words: 需要加载词向量的词语列表，以 itos (index to string) 的词典形式给出
pretrained_vocab: 预训练词向量
@return:
embed: 加载到的词向量
'''
embed = torch.zeros(len(words), pretrained_vocab.vectors[0].shape[0]) # 初始化为0
oov_count = 0 # out of vocabulary
for i, word in enumerate(words):
try:
idx = pretrained_vocab.stoi[word]
embed[i, :] = pretrained_vocab.vectors[idx]
except KeyError:
oov_count += 1
if oov_count > 0:
print("There are %d oov words." % oov_count)
return embed
net.embedding.weight.data.copy_(load_pretrained_embedding(vocab.itos, glove_vocab))
net.embedding.weight.requires_grad = False # 直接加载预训练好的, 所以不需要更新它
定义训练和验证模型函数
def evaluate_accuracy(data_iter, net, device=None):
if device is None and isinstance(net, torch.nn.Module):
device = list(net.parameters())[0].device
acc_sum, n = 0.0, 0
with torch.no_grad():
for X, y in data_iter:
if isinstance(net, torch.nn.Module):
net.eval()
acc_sum += (net(X.to(device)).argmax(dim=1) == y.to(device)).float().sum().cpu().item()
net.train()
else:
if('is_training' in net.__code__.co_varnames):
acc_sum += (net(X, is_training=False).argmax(dim=1) == y).float().sum().item()
else:
acc_sum += (net(X).argmax(dim=1) == y).float().sum().item()
n += y.shape[0]
return acc_sum / n
def train(train_iter, test_iter, net, loss, optimizer, device, num_epochs):
net = net.to(device)
print("training on ", device)
batch_count = 0
for epoch in range(num_epochs):
train_l_sum, train_acc_sum, n, start = 0.0, 0.0, 0, time.time()
for X, y in train_iter:
X = X.to(device)
y = y.to(device)
y_hat = net(X)
l = loss(y_hat, y)
optimizer.zero_grad()
l.backward()
optimizer.step()
train_l_sum += l.cpu().item()
train_acc_sum += (y_hat.argmax(dim=1) == y).sum().cpu().item()
n += y.shape[0]
batch_count += 1
test_acc = evaluate_accuracy(test_iter, net)
print('epoch %d, loss %.4f, train acc %.3f, test acc %.3f, time %.1f sec'
% (epoch + 1, train_l_sum / batch_count, train_acc_sum / n, test_acc, time.time() - start))
开始训练
lr, num_epochs = 0.01, 5
optimizer = torch.optim.Adam(filter(lambda p: p.requires_grad, net.parameters()), lr=lr)
loss = nn.CrossEntropyLoss()
train(train_iter, test_iter, net, loss, optimizer, device, num_epochs)
预测模型，Demo实现。
def predict_sentiment(net, vocab, sentence):
'''
@params：
net: 训练好的模型
vocab: 在该数据集上创建的词典，用于将给定的单词序转换为单词下标的序列，从而输入模型
sentence: 需要分析情感的文本，以单词序列的形式给出
@return: 预测的结果，positive 为正面情绪文本，negative 为负面情绪文本
'''
device = list(net.parameters())[0].device # 读取模型所在的环境
sentence = torch.tensor([vocab.stoi[word] for word in sentence], device=device)
label = torch.argmax(net(sentence.view((1, -1))), dim=1)
return 'positive' if label.item() == 1 else 'negative'
predict_sentiment(net, vocab, ['this', 'movie', 'is', 'so', 'great'])