深度学习训练营之文本分类识别
- 原文链接
- 环境介绍
- 前置工作
- 设置环境
- 设置GPU
- 加载数据
- 构建词典
- 生成数据批次和迭代器
- 模型定义
- 定义实例
- 定义训练函数和评估函数
- 模型训练
- 模型评估
原文链接
- 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
- 🍦 参考文章:[365天深度学习训练营-第N1周:pytorch文本分类识别]https://www.yuque.com/mingtian-fkmxf/hv4lcq/mscin5fy03p1q6xq)
- 🍖 原作者:K同学啊|接辅导、项目定制
环境介绍
- 语言环境:Python3.9.12
- 编译器:jupyter notebook
- 深度学习环境:pytorch
前置工作
文本分类大致过程如下:
设置环境
在工作开始之前,先保证下载了需要使用到的两个包torchtext
和portalocker
,我是在anaconda prompt当中下载的,下载 命令如下
pip install torchtext
pip install portalocker
设置GPU
选择运行的设备
import torch
import torch.nn as nn
import torchvision
from torchvision import transforms, datasets
import os,PIL,pathlib,warnings
warnings.filterwarnings("ignore") #忽略警告信息
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(device)
cpu
加载数据
from torchtext.datasets import AG_NEWS
train_iter = AG_NEWS(split='train') # 加载 AG News 数据集
torchtext.datasets.AG_NEWS()
是用于加载AG News数据集的TorchText数据集类,主要包含有世界,科技,体育和商业等新闻文章
构建词典
import torch
import torch.nn as nn
import torchvision
import os,PIL,pathlib,warnings
import time
from torchvision import transforms, datasets
from torchtext.datasets import AG_NEWS
from torchtext.data.utils import get_tokenizer
from torchtext.vocab import build_vocab_from_iterator
from torch import nn
from torch.utils.data.dataset import random_split
from torchtext.data.functional import to_map_style_dataset
warnings.filterwarnings("ignore") #忽略警告信息
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(device)
#数据集
train_iter = AG_NEWS(split='train') # 加载 AG News 数据集
#构建词典
tokenizer = get_tokenizer('basic_english') # 返回分词器函数
def yield_tokens(data_iter):
for _, text in data_iter:
yield tokenizer(text)
vocab = build_vocab_from_iterator(yield_tokens(train_iter), specials=["<unk>"])
vocab.set_default_index(vocab["<unk>"]) # 设置默认索引,如果找不到单词,则会选择默认索引
print(vocab(['here', 'is', 'an', 'example']))
[475, 21, 30, 5297]
[475, 21, 2, 30, 5297]
print(label_pipeline('10'))
9
生成数据批次和迭代器
from torch.utils.data import DataLoader
def collate_batch(batch):
label_list, text_list, offsets = [], [], [0]
for (_label, _text) in batch:
# 标签列表
label_list.append(label_pipeline(_label))
# 文本列表
processed_text = torch.tensor(text_pipeline(_text), dtype=torch.int64)
text_list.append(processed_text)
# 偏移量,即语句的总词汇量
offsets.append(processed_text.size(0))
label_list = torch.tensor(label_list, dtype=torch.int64)
text_list = torch.cat(text_list)
offsets = torch.tensor(offsets[:-1]).cumsum(dim=0) #返回维度dim中输入元素的累计和
return label_list.to(device), text_list.to(device), offsets.to(device)
dataloader = DataLoader(train_iter,
batch_size=8,
shuffle =False,
collate_fn=collate_batch)
模型定义
首先先定义我们进行分类用到的模型,然后嵌入文本,然后对句子嵌入之后的结果均值聚合
from torch import nn
class TextClassificationModel(nn.Module):
def __init__(self, vocab_size, embed_dim, num_class):
super(TextClassificationModel, self).__init__()
self.embedding = nn.EmbeddingBag(vocab_size, # 词典大小
embed_dim, # 嵌入的维度
sparse=False) #
self.fc = nn.Linear(embed_dim, num_class)
self.init_weights()
def init_weights(self):
initrange = 0.5
self.embedding.weight.data.uniform_(-initrange, initrange)
self.fc.weight.data.uniform_(-initrange, initrange)
self.fc.bias.data.zero_()
def forward(self, text, offsets):
embedded = self.embedding(text, offsets)
return self.fc(embedded)
self.embedding.weight.data.uniform_(-initrange, initrange)
这段代码在PyTorch框架下用于初始化神经网络的词嵌入层权重的一种方法,这样使得模型在训练时具有一定的随机性,避免了梯度消失或者梯度爆炸等问题
定义实例
在设置好的模型当中进行一个命名为model
num_class = len(set([label for (label, text) in train_iter]))
vocab_size = len(vocab)
em_size = 64
model = TextClassificationModel(vocab_size, em_size, num_class).to(device)
定义训练函数和评估函数
def train(dataloader):
model.train() # 切换为训练模式
total_acc, train_loss, total_count = 0, 0, 0
log_interval = 500
start_time = time.time()
for idx, (label, text, offsets) in enumerate(dataloader):
predicted_label = model(text, offsets)
optimizer.zero_grad() # grad属性归零
loss = criterion(predicted_label, label) # 计算网络输出和真实值之间的差距,label为真实值
loss.backward() # 反向传播
optimizer.step() # 每一步自动更新
# 记录acc与loss
total_acc += (predicted_label.argmax(1) == label).sum().item()
train_loss += loss.item()
total_count += label.size(0)
if idx % log_interval == 0 and idx > 0:
elapsed = time.time() - start_time
print('| epoch {:1d} | {:4d}/{:4d} batches '
'| train_acc {:4.3f} train_loss {:4.5f}'.format(epoch, idx, len(dataloader),
total_acc/total_count, train_loss/total_count))
total_acc, train_loss, total_count = 0, 0, 0
start_time = time.time()
def evaluate(dataloader):
model.eval() # 切换为测试模式
total_acc, train_loss, total_count = 0, 0, 0
with torch.no_grad():
for idx, (label, text, offsets) in enumerate(dataloader):
predicted_label = model(text, offsets)
loss = criterion(predicted_label, label) # 计算loss值
# 记录测试数据
total_acc += (predicted_label.argmax(1) == label).sum().item()
train_loss += loss.item()
total_count += label.size(0)
return total_acc/total_count, train_loss/total_count
模型训练
EPOCHS = 10 # epoch
LR = 5 # 学习率
BATCH_SIZE = 64 # batch size for training
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=LR)
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, 1.0, gamma=0.1)
total_accu = None
train_iter, test_iter = AG_NEWS() # 加载数据
train_dataset = to_map_style_dataset(train_iter)
test_dataset = to_map_style_dataset(test_iter)
num_train = int(len(train_dataset) * 0.95)
split_train_, split_valid_ = random_split(train_dataset,
[num_train, len(train_dataset)-num_train])
train_dataloader = DataLoader(split_train_, batch_size=BATCH_SIZE,
shuffle=True, collate_fn=collate_batch)
valid_dataloader = DataLoader(split_valid_, batch_size=BATCH_SIZE,
shuffle=True, collate_fn=collate_batch)
test_dataloader = DataLoader(test_dataset, batch_size=BATCH_SIZE,
shuffle=True, collate_fn=collate_batch)
for epoch in range(1, EPOCHS + 1):
epoch_start_time = time.time()
train(train_dataloader)
val_acc, val_loss = evaluate(valid_dataloader)
if total_accu is not None and total_accu > val_acc:
scheduler.step()
else:
total_accu = val_acc
print('| epoch {:1d} | time: {:4.2f}s | '
'valid_acc {:4.3f} valid_loss {:4.3f}'.format(epoch,
time.time() - epoch_start_time,
val_acc,val_loss))
| epoch 1 | 500/1782 batches | train_acc 0.716 train_loss 0.01121
| epoch 1 | 1000/1782 batches | train_acc 0.864 train_loss 0.00622
| epoch 1 | 1500/1782 batches | train_acc 0.881 train_loss 0.00552
| epoch 1 | time: 24.91s | valid_acc 0.862 valid_loss 0.006
| epoch 2 | 500/1782 batches | train_acc 0.903 train_loss 0.00452
| epoch 2 | 1000/1782 batches | train_acc 0.904 train_loss 0.00445
| epoch 2 | 1500/1782 batches | train_acc 0.904 train_loss 0.00444
| epoch 2 | time: 26.10s | valid_acc 0.860 valid_loss 0.006
| epoch 3 | 500/1782 batches | train_acc 0.928 train_loss 0.00344
| epoch 3 | 1000/1782 batches | train_acc 0.931 train_loss 0.00337
| epoch 3 | 1500/1782 batches | train_acc 0.928 train_loss 0.00347
| epoch 3 | time: 24.54s | valid_acc 0.912 valid_loss 0.004
| epoch 4 | 500/1782 batches | train_acc 0.931 train_loss 0.00336
| epoch 4 | 1000/1782 batches | train_acc 0.931 train_loss 0.00328
| epoch 4 | 1500/1782 batches | train_acc 0.931 train_loss 0.00329
| epoch 4 | time: 22.38s | valid_acc 0.914 valid_loss 0.004
| epoch 5 | 500/1782 batches | train_acc 0.933 train_loss 0.00329
| epoch 5 | 1000/1782 batches | train_acc 0.932 train_loss 0.00320
| epoch 5 | 1500/1782 batches | train_acc 0.932 train_loss 0.00329
| epoch 5 | time: 25.84s | valid_acc 0.913 valid_loss 0.004
| epoch 6 | 500/1782 batches | train_acc 0.936 train_loss 0.00307
…
| epoch 10 | 1000/1782 batches | train_acc 0.933 train_loss 0.00318
| epoch 10 | 1500/1782 batches | train_acc 0.936 train_loss 0.00310
| epoch 10 | time: 18.34s | valid_acc 0.915 valid_loss 0.004
TorchText是Torch的一个拓展库,专注于处理文本数据,这样我们可以通过索引直接访问数据集当中的特定样本,简化了模型的训练,验证和测试过程当中的数据处理
模型评估
# 评估模型
test_acc, test_loss = evaluate(test_dataloader)
print('test accuracy {:8.3f}'.format(test_acc))
test accuracy 0.909