利用pytorch写一个self-attention的二分类代码

好的，以下是一个利用 PyTorch 实现 self-attention 的二分类代码：

import torch
import torch.nn as nn

class SelfAttention(nn.Module):
    def __init__(self, hidden_size):
        super(SelfAttention, self).__init__()
        self.hidden_size = hidden_size
        self.projection = nn.Sequential(
            nn.Linear(hidden_size, 64),
            nn.ReLU(True),
            nn.Linear(64, 1)
        )

    def forward(self, encoder_outputs):
        # encoder_outputs: [batch_size, sequence_length, hidden_size]
        energy = self.projection(encoder_outputs)
        weights = nn.functional.softmax(energy.squeeze(-1), dim=1)
        outputs = (encoder_outputs * weights.unsqueeze(-1)).sum(dim=1)
        return outputs

class Classifier(nn.Module):
    def __init__(self, hidden_size):
        super(Classifier, self).__init__()
        self.attention = SelfAttention(hidden_size)
        self.fc = nn.Linear(hidden_size, 2)

    def forward(self, inputs):
        # inputs: [batch_size, sequence_length, hidden_size]
        outputs = self.attention(inputs)
        outputs = self.fc(outputs)
        return outputs

# Example usage
model = Classifier(hidden_size=256)
inputs = torch.randn(32, 10, 256)
outputs = model(inputs)
print(outputs.shape)  # torch.Size([32, 2])

这个代码实现了一个基于 self-attention 的二分类器，其中 `SelfAttention` 类实现了 self-attention 的计算，`Classifier` 类则将 self-attention 的输出通过一个全连接层映射到二分类的结果。