Pytorch变压器编码器解释了

import torch
import torch.nn as nn

class TransformerEncoderLayer(nn.Module):
    def __init__(self, d_model, nhead, dropout=0.1):
        super().__init__()
        self.multihead_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
        self.linear1 = nn.Linear(d_model, d_model * 4)
        self.linear2 = nn.Linear(d_model * 4, d_model)
        self.dropout = nn.Dropout(dropout)
        self.activation = nn.ReLU()
        self.layer_norm1 = nn.LayerNorm(d_model)
        self.layer_norm2 = nn.LayerNorm(d_model)

    def forward(self, x, mask=None):
        x = self.layer_norm1(x)
        x = self.multihead_attn(x, x, x, key_padding_mask=mask)[0] #Simplified multihead_attn call
        x = self.dropout(x)
        x = x + x
        x = self.layer_norm2(x)
        temp = self.linear2(self.activation(self.linear1(x)))
        x = self.dropout(temp)
        x = x + x
        return x

class TransformerEncoder(nn.Module):
    def __init__(self, d_model, nhead, num_layers, dropout=0.1):
        super().__init__()
        self.layers = nn.ModuleList(
            [TransformerEncoderLayer(d_model, nhead, dropout) for _ in range(num_layers)]
        )
        self.num_layers = num_layers

    def forward(self, x, mask=None):
        for i in range(self.num_layers):
            x = self.layers[i](x, mask)
        return x