多层感知机与自注意力机制使用代码

以下是使用Python和PyTorch实现多层感知机和自注意力机制的代码示例：

多层感知机：

import torch
import torch.nn as nn

class MLP(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(MLP, self).__init__()
        self.fc1 = nn.Linear(input_size, hidden_size)
        self.relu = nn.ReLU()
        self.fc2 = nn.Linear(hidden_size, output_size)
        
    def forward(self, x):
        out = self.fc1(x)
        out = self.relu(out)
        out = self.fc2(out)
        return out

自注意力机制：

import torch
import torch.nn as nn

class SelfAttention(nn.Module):
    def __init__(self, input_size, heads):
        super(SelfAttention, self).__init__()
        self.input_size = input_size
        self.heads = heads
        self.head_size = input_size // heads
        
        self.query = nn.Linear(input_size, input_size)
        self.key = nn.Linear(input_size, input_size)
        self.value = nn.Linear(input_size, input_size)
        self.softmax = nn.Softmax(dim=-1)
        self.output = nn.Linear(input_size, input_size)
        
    def forward(self, x):
        batch_size, seq_len, input_size = x.size()
        queries = self.query(x).view(batch_size, seq_len, self.heads, self.head_size).transpose(1,2)
        keys = self.key(x).view(batch_size, seq_len, self.heads, self.head_size).transpose(1,2)
        values = self.value(x).view(batch_size, seq_len, self.heads, self.head_size).transpose(1,2)
        
        attention_scores = torch.matmul(queries, keys.transpose(-2, -1)) / (self.head_size ** 0.5)
        attention_probs = self.softmax(attention_scores)
        context = torch.matmul(attention_probs, values).transpose(1,2).contiguous().view(batch_size, seq_len, input_size)
        output = self.output(context)
        return output

这里的`input_size`指输入数据的特征数，`hidden_size`指隐藏层的大小，`output_size`指输出层的大小，`heads`指自注意力机制中注意力头的数量。在`forward`函数中，`x`表示输入数据，`out`表示输出数据。