[自然语言处理|NLP]NLP在小样本学习与元学习的应用：从原理到实践

# 示例代码：小样本学习在情感分析中的应用
import torch
from transformers import BertTokenizer, BertForSequenceClassification, AdamW
from torch.utils.data import DataLoader, TensorDataset

# 加载预训练的BERT模型和tokenizer
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
model = BertForSequenceClassification.from_pretrained('bert-base-uncased', num_labels=3)

# 构建小样本数据集
comments = ["I love this product!", "Not satisfied with the quality.", "Amazing service!"]
labels = [1, 0, 1]

# 对评论进行tokenize和padding
inputs = tokenizer(comments, padding=True, truncation=True, return_tensors="pt")
labels = torch.tensor(labels)

# 构建数据集和数据加载器
dataset = TensorDataset(inputs['input_ids'], inputs['attention_mask'], labels)
dataloader = DataLoader(dataset, batch_size=2, shuffle=True)

# 定义优化器和损失函数
optimizer = AdamW(model.parameters(), lr=1e-5)
loss_fn = torch.nn.CrossEntropyLoss()

# 训练模型
epochs = 5
for epoch in range(epochs):
    for batch in dataloader:
        optimizer.zero_grad()
        input_ids, attention_mask, label = batch
        outputs = model(input_ids, attention_mask=attention_mask, labels=label)
        loss = outputs.loss


        loss.backward()
        optimizer.step()

# 模型训练完成，可以在新评论上进行情感分析
new_comment = "The customer service was excellent!"
input_tensor = tokenizer(new_comment, return_tensors="pt")
output = model(**input_tensor)
predicted_label = torch.argmax(output.logits).item()

print("Predicted Label:", predicted_label)