用pytorch写一个NLP Classifier

以下是一个使用PyTorch实现NLP分类器的简单示例：

import torch
import torch.nn as nn
import torch.optim as optim
from torchtext.datasets import IMDB
from torchtext.data import Field, LabelField, BucketIterator

# 设置随机种子以确保结果可重复
SEED = 1234
torch.manual_seed(SEED)
torch.backends.cudnn.deterministic = True

# 定义字段
TEXT = Field(tokenize='spacy', tokenizer_language='en_core_web_sm')
LABEL = LabelField(dtype=torch.float)

# 加载IMDB数据集并分割为训练和测试集
train_data, test_data = IMDB.splits(TEXT, LABEL)
train_data, valid_data = train_data.split(random_state=random.seed(SEED))

# 构建词汇表
TEXT.build_vocab(train_data, max_size=25000, vectors="glove.6B.100d")
LABEL.build_vocab(train_data)

# 设定设备
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

# 定义模型
class NLPClassifier(nn.Module):
    def __init__(self, input_dim, embedding_dim, hidden_dim, output_dim):
        super().__init__()
        self.embedding = nn.Embedding(input_dim, embedding_dim)
        self.rnn = nn.LSTM(embedding_dim, hidden_dim)
        self.fc = nn.Linear(hidden_dim, output_dim)
        
    def forward(self, text):
        embedded = self.embedding(text)
        output, (hidden, cell) = self.rnn(embedded)
        return self.fc(hidden.squeeze(0))

# 定义超参数
INPUT_DIM = len(TEXT.vocab)
EMBEDDING_DIM = 100
HIDDEN_DIM = 256
OUTPUT_DIM = 1

# 初始化模型、损失函数和优化器
model = NLPClassifier(INPUT_DIM, EMBEDDING_DIM, HIDDEN_DIM, OUTPUT_DIM)
criterion = nn.BCEWithLogitsLoss()
optimizer = optim.Adam(model.parameters())

# 将数据集分批
BATCH_SIZE = 64
train_iterator, valid_iterator, test_iterator = BucketIterator.splits(
    (train_data, valid_data, test_data),
    batch_size=BATCH_SIZE,
    device=device)

# 训练模型
def train(model, iterator, optimizer, criterion):
    epoch_loss = 0
    epoch_acc = 0
    
    model.train()
    
    for batch in iterator:
        optimizer.zero_grad()
        predictions = model(batch.text).squeeze(1)
        loss = criterion(predictions, batch.label)
        acc = binary_accuracy(predictions, batch.label)
        loss.backward()
        optimizer.step()
        
        epoch_loss += loss.item()
        epoch_acc += acc.item()
        
    return epoch_loss / len(iterator), epoch_acc / len(iterator)

# 评估模型
def evaluate(model, iterator, criterion):
    epoch_loss = 0
    epoch_acc = 0
    
    model.eval()
    
    with torch.no_grad():
        for batch in iterator:
            predictions = model(batch.text).squeeze(1)
            loss = criterion(predictions, batch.label)
            acc = binary_accuracy(predictions, batch.label)

            epoch_loss += loss.item()
            epoch_acc += acc.item()
        
    return epoch_loss / len(iterator), epoch_acc / len(iterator)

# 评估在二进制分类中的准确性
def binary_accuracy(predictions, y):
    rounded_preds = torch.round(torch.sigmoid(predictions))
    correct = (rounded_preds == y).float()
    acc = correct.sum() / len(correct)
    return acc

# 训练和评估模型
N_EPOCHS = 5
best_valid_loss = float('inf')

for epoch in range(N_EPOCHS):

    train_loss, train_acc = train(model, train_iterator, optimizer, criterion)
    valid_loss, valid_acc = evaluate(model, valid_iterator, criterion)
    
    if valid_loss < best_valid_loss:
        best_valid_loss = valid_loss
        torch.save(model.state_dict(), 'nlp_classifier.pt')

    print(f'Epoch: {epoch+1:02}\nTrain Loss: {train_loss:.3f} | Train Acc: {train_acc*100:.2f}%')
    print(f'Val. Loss: {valid_loss:.3f} | Val. Acc: {valid_acc*100:.2f}%')

# 加载并评估模型
model.load_state_dict(torch.load('nlp_classifier.pt'))
test_loss, test_acc = evaluate(model, test_iterator, criterion)
print(f'Test Loss: {test_loss:.3f} | Test Acc: {test_acc*100:.2f}%')

该模型使用了一个简单的LSTM层和一个全连接层，将文本嵌入到一个低维空间中，并输出二进制分类结果。模型使用了BCEWithLogitsLoss损失函数和Adam优化器来最小化损失。在数据集上训练5个时期后，模型的测试准确率为约85％。