Transformer代码实现【pytorch】

参考文献：
博客：【手撕Transformer】Transformer输入输出细节以及代码实现（pytorch）
博客：Transformer 的 PyTorch 实现

代码+注释

import math
import torch
import numpy as np
import torch.nn as nn
import torch.optim as optim
import torch.utils.data as Data

# 自制数据集
# S:开始符号 E：结束符号 P：占位符
            # Encoder_input  Decoder_input       Decoder_output
sentences = [['我 是 学 生 P' , 'S I am a student'   , 'I am a student E'],
             ['我 喜 欢 学 习', 'S I like learning P', 'I like learning P E'],
             ['我 是 男 生 P' , 'S I am a boy'       , 'I am a boy E']]

#词源词典
src_vocab = {'P':0, '我':1, '是':2, '学':3, '生':4, '喜':5, '欢':6, '习':7, '男':8}
src_idx2word = {src_vocab[key]: key for key in src_vocab} #把字典转换成idx:字形式
src_vocab_size = len(src_vocab)

#目标字典
tgt_vocab = {'S':0, 'E':1, 'P':2, 'I':3, 'am':4, 'a':5, 'student':6, 'like':7, 'learning':8, 'boy':9}
idx2word = {tgt_vocab[key]: key for key in tgt_vocab}
tgt_vocab_size = len(tgt_vocab)

src_len = len(sentences[0][0].split(" ")) # Encoder的最大输入长度
tgt_len = len(sentences[0][1].split(" ")) # Decoder的最大输入长度
#print(src_len, tgt_len)  5,5

#把sentences转换成字典索引
def make_data(sentences):
    enc_inputs, dec_inputs, dec_outputs = [], [], []
    for i in range(len(sentences)):
        enc_input = [[src_vocab[n] for n in sentences[i][0].split()]]
        dec_input = [[tgt_vocab[n] for n in sentences[i][1].split()]]
        dec_output = [[tgt_vocab[n] for n in sentences[i][2].split()]]
        enc_inputs.extend(enc_input)
        dec_inputs.extend(dec_input)
        dec_outputs.extend(dec_output)
    return torch.LongTensor(enc_inputs), torch.LongTensor(dec_inputs), torch.LongTensor(dec_outputs)

enc_inputs, dec_inputs, dec_outputs = make_data(sentences)
# print(enc_inputs)
# print(dec_inputs)
# print(dec_inputs)

#自定义数据集
class MyDataSet(Data.Dataset):
    def __init__(self, enc_inputs, dec_inputs, dec_outputs):
        super(MyDataSet, self).__init__()
        self.enc_inputs = enc_inputs
        self.dec_inputs = dec_inputs
        self.dec_outputs = dec_outputs

    def __len__(self):
        return self.enc_inputs.shape[0]

    def __getitem__(self, idx):
        return self.enc_inputs[idx], self.dec_inputs[idx], self.dec_outputs[idx]

loader = Data.DataLoader(MyDataSet(enc_inputs, dec_inputs, dec_outputs), 2, True)

#参数设置
d_model = 512 #字embedding的维度
d_ff = 2048 #前向传播的隐藏层维度
d_k = d_v = 64 #K,Q,V的维度
n_layers = 6 #有多少个Encoder和decoder
n_heads = 8

#定义位置信息
class PositionalEncoding(nn.Module):
    def __init__(self, d_model, dropout=0.1, max_len=5000):
        super(PositionalEncoding, self).__init__()
        self.dropout = nn.Dropout(p=dropout)
        pos_table = np.array([
            [pos / np.power(10000, 2 * i / d_model) for i in range(d_model)]
            if pos!=0 else np.zeros(d_model) for pos in range(max_len)
        ])
        pos_table[1:, 0::2] = np.sin(pos_table[1:, 0::2])  # 字嵌入维度为偶数时
        pos_table[1:, 1::2] = np.cos(pos_table[1:, 1::2])  # 字嵌入维度为奇数时
        self.pos_table = torch.FloatTensor(pos_table).cuda()
    def forward(self, enc_inputs):
        enc_inputs += self.pos_table[:enc_inputs.size(1),:]
        return self.dropout(enc_inputs.cuda())

# Mask掉停用词 P对应句子没有实际意义，所以需要Mask
# seq_k.data.eq(0)，这句的作用是返回一个大小和 seq_k 一样的 tensor，
# 只不过里面的值只有 True 和 False。
# 如果 seq_k 某个位置的值等于 0，那么对应位置就是 True，否则即为 False
# encoder和decoder都会用到，如果实在Encoder调用 seq_len=src_len,如果在decoder调用seq_len有可能调用seq-len也有可能等于src_len
def get_attn_pad_mask(seq_q, seq_k):
    batch_size, len_q = seq_q.size()
    batch_size, len_k = seq_k.size()
    pad_attn_mask = seq_k.data.eq(0).unsqueeze(1) #[batch_size, 1, len_k] unsqueeze()函数起升维的作用,参数表示在哪个地方加一个维度
    return pad_attn_mask.expand(batch_size, len_q, len_k) #[batch_size, len_q, len_k]

# Decoder输入  mask未来输入信息
# 对于decoder，我们每次输入是在前一次基础上加一个词，所以输入是一个下三角矩阵
# 比如输入"S I am a student"
# 在T0时刻先输入"S"预测，预测第一个词"I"；在下一个T1时刻，
# 同时输入"S"和"I"到Decoder预测下一个单词"am"；
# 然后在T2时刻把"S,I,am"同时输入到Decoder预测下一个单词"a"
# np.triu(data, k) k=0表示正常的上三角矩阵
# S
# S I
# S I am

def get_attn_subsequence_mask(seq):                       # seq: [batch_size, tgt_len]
    attn_shape = [seq.size(0), seq.size(1), seq.size(1)]  # 生成上三角矩阵,[batch_size, tgt_len, tgt_len]
    subsequence_mask = np.triu(np.ones(attn_shape), k=1)
    subsequence_mask = torch.from_numpy(subsequence_mask).byte()
    return subsequence_mask

# 计算注意力 残差和归一化
class ScaledDotProductAttention(nn.Module):
    def __init__(self):
        super(ScaledDotProductAttention, self).__init__()

    def forward(self, Q, K, V, attn_mask):
        # Q: [batch_size, n_heads, len_q, d_k]
        # K: [batch_size, n_heads, len_k, d_k]
        # V: [batch_size, n_heads, len_k, d_v]
        scores = torch.matmul(Q, K.transpose(-1, -2)) / np.sqrt(d_k)
        scores.masked_fill_(attn_mask, -1e9)
        attn = nn.Softmax(dim=-1)(scores)
        context = torch.matmul(attn, V)
        return context, attn

class MultiHeadAttention(nn.Module):
    def __init__(self):
        super(MultiHeadAttention, self).__init__()
        self.W_Q = nn.Linear(d_model, d_k * n_heads, bias=False)
        self.W_K = nn.Linear(d_model, d_k * n_heads, bias=False)
        self.W_V = nn.Linear(d_model, d_v * n_heads, bias=False)
        self.fc = nn.Linear(d_v * n_heads, d_model,  bias=False)

    def forward(self, input_Q, input_K, input_V, attn_mask):
        '''
        input_Q: [batch_size, len_q, d_model]
        input_K: [batch_size, len_k, d_model]
        input_V: [batch_size, len_v(=len_k), d_model]
        attn_mask: [batch_size, seq_len, seq_len]
        '''
        residual, batch_size = input_Q, input_Q.size(0)
        Q = self.W_Q(input_Q).view(batch_size, -1, n_heads, d_k).transpose(1, 2)
        K = self.W_K(input_K).view(batch_size, -1, n_heads, d_k).transpose(1, 2)
        V = self.W_V(input_V).view(batch_size, -1, n_heads, d_v).transpose(1, 2)
        attn_mask = attn_mask.unsqueeze(1).repeat(1, n_heads, 1, 1)
        context, attn = ScaledDotProductAttention()(Q, K, V, attn_mask)
        context = context.transpose(1, 2).reshape(batch_size, -1, n_heads*d_v)
        output = self.fc(context)
        return nn.LayerNorm(d_model).cuda()(output + residual), attn

#前馈神经网络
class PoswizeFeedForwardNet(nn.Module):
    def __init__(self):
        super(PoswizeFeedForwardNet, self).__init__()
        self.fc = nn.Sequential(
            nn.Linear(d_model, d_ff, bias=False),
            nn.ReLU(),
            nn.Linear(d_ff, d_model, bias=False)
        )
    def forward(self, inputs):
        residual = inputs
        output = self.fc(inputs)
        return nn.LayerNorm(d_model).cuda()(output + residual)

class EncoderLayer(nn.Module):
    def __init__(self):
        super(EncoderLayer, self).__init__()
        self.enc_self_attn = MultiHeadAttention()
        self.pos_ffn = PoswizeFeedForwardNet()

    def forward(self, enc_inputs, enc_self_attn_mask):
        #输入三个enc_inputs分别与W_q,Q_k,W_v相乘得到Q,K,V
        enc_outputs, attn = self.enc_self_attn(enc_inputs, enc_inputs, enc_inputs, enc_self_attn_mask)
        enc_outputs = self.pos_ffn(enc_inputs)
        return  enc_outputs, attn


# Encoder
# 首先 将中文embedding
# 将embedding加上位置信息
# Mask掉句子中的占位符
# 通过N个Encoder， N=6
class Encoder(nn.Module):
    def __init__(self):
        super(Encoder, self).__init__()
        self.src_emb = nn.Embedding(src_vocab_size, d_model)
        self.pos_emb = PositionalEncoding(d_model)
        self.layers = nn.ModuleList([EncoderLayer() for _ in range(n_layers)])

    def forward(self, enc_inputs):
        '''
        :param enc_inputs: [batch_size, src_len]
        :return:
        '''
        enc_outputs = self.src_emb(enc_inputs) # [batch_size, src_len d_model]
        enc_outputs = self.pos_emb(enc_outputs.transpose(0, 1)).transpose(0, 1)
        enc_self_attn_mask = get_attn_pad_mask(enc_inputs, enc_inputs)
        enc_self_attns = []
        for layer in self.layers:
            # enc_outputs: [batch_size, src_len, d_model], enc_self_attn: [batch_size, n_heads, src_len, src_len]
            enc_outputs, enc_self_attn = layer(enc_outputs, enc_self_attn_mask)
            enc_self_attns.append(enc_self_attn)
        return enc_outputs, enc_self_attns

class DecoderLayer(nn.Module):
    def __init__(self):
        super(DecoderLayer, self).__init__()
        self.dec_self_attn = MultiHeadAttention()
        self.dec_enc_attn = MultiHeadAttention()
        self.pos_ffn = PoswizeFeedForwardNet()

    def forward(self, dec_inputs, enc_outputs, dec_self_attn_mask, dec_enc_attn_mask):
        dec_outputs, dec_self_attn = self.dec_self_attn(dec_inputs, dec_inputs, dec_inputs, dec_self_attn_mask)
        dec_outputs, dec_enc_attn = self.dec_enc_attn(dec_outputs, dec_inputs, dec_inputs, dec_enc_attn_mask)
        dec_outputs = self.pos_ffn(dec_outputs)
        return dec_outputs, dec_self_attn, dec_enc_attn

# Decoder
class Decoder(nn.Module):
    def __init__(self):
        super(Decoder, self).__init__()
        self.tgt_emb = nn.Embedding(tgt_vocab_size, d_model)
        self.pos_emb = PositionalEncoding(d_model)
        self.layers = nn.ModuleList([DecoderLayer() for _ in range(n_layers)])

    def forward(self, dec_inputs, enc_inputs, enc_outputs):
        '''
        :param dec_inputs: [batch_size, tgt_leb]
        :param enc_inputs: [batch_size, src_len]
        :param enc_outputs: [batch_size, src_len, d_model]
        :return:
        '''
        dec_outputs = self.tgt_emb(dec_inputs) # [batch_size, tgt_leb, d_model]
        dec_outputs = self.pos_emb(dec_outputs.transpose(0,1)).transpose(0,1)
        dec_self_attn_pad_mask = get_attn_pad_mask(dec_inputs, dec_inputs).cuda()
        dec_self_attn_subsequence_mask = get_attn_subsequence_mask(dec_inputs).cuda()
        dec_self_attn_mask = torch.gt((dec_self_attn_pad_mask + dec_self_attn_subsequence_mask), 0).cuda()

        # 这个mask主要是enc_inputs的pad mask矩阵
        dec_enc_attn_mask = get_attn_pad_mask(dec_inputs, enc_inputs)
        dec_self_attns, dec_enc_attns = [], []
        for layer in self.layers:
            dec_outputs, dec_self_attn, dec_enc_attn = layer(dec_outputs, enc_outputs, dec_self_attn_mask, dec_enc_attn_mask)
            dec_self_attns.append(dec_self_attn)
            dec_enc_attns.append(dec_enc_attn)
        return dec_outputs, dec_self_attns, dec_enc_attns

# Transformer
class Transformer(nn.Module):
    def __init__(self):
        super(Transformer, self).__init__()
        self.Encoder = Encoder().cuda()
        self.Decoder = Decoder().cuda()
        self.projection = nn.Linear(d_model, tgt_vocab_size, bias=False).cuda()
    def forward(self, enc_inputs, dec_inputs):
        enc_outputs, enc_self_attns = self.Encoder(enc_inputs)
        dec_outputs, dec_self_attns, dec_enc_attns = self.Decoder(
            dec_inputs, enc_inputs, enc_outputs)
        dec_logits = self.projection(dec_outputs)
        return dec_logits.view(-1, dec_logits.size(-1)), enc_self_attns, dec_self_attns, dec_enc_attns

# 定义网络
model = Transformer().cuda()
criterion = nn.CrossEntropyLoss(ignore_index=0)
optimizer = optim.SGD(model.parameters(), lr=1e-3, momentum=0.99)

# 训练
# for epoch in range(1000):
#     for enc_inputs, dec_inputs, dec_outputs in loader:
#         enc_inputs, dec_inputs, dec_outputs = enc_inputs.cuda(), dec_inputs.cuda(), dec_outputs.cuda()
#         outputs, enc_self_attns, dec_self_attns, dec_enc_attns = model(enc_inputs, dec_inputs)
#
#         loss = criterion(outputs, dec_outputs.view(-1))
#         print('Epoch:', '%04d' % (epoch + 1), 'loss =', '{:.6f}'.format(loss))
#         optimizer.zero_grad()
#         loss.backward()
#         optimizer.step()


# 测试
def test(model, enc_input, start_symbol):
    enc_outputs, enc_self_attns = model.Encoder(enc_input)
    dec_input = torch.zeros(1,tgt_len).type_as(enc_input.data)
    next_symbol = start_symbol
    for i in range(0,tgt_len):
        dec_input[0][i] = next_symbol
        dec_outputs, _, _ = model.Decoder(dec_input,enc_input,enc_outputs)
        projected = model.projection(dec_outputs)
        prob = projected.squeeze(0).max(dim=-1, keepdim=False)[1]
        next_word = prob.data[i]
        next_symbol = next_word.item()
    return dec_input
enc_inputs, _, _ = next(iter(loader))
predict_dec_input = test(model, enc_inputs[1].view(1, -1).cuda(), start_symbol=tgt_vocab["S"])
predict, _, _, _ = model(enc_inputs[1].view(1, -1).cuda(), predict_dec_input)
predict = predict.data.max(1, keepdim=True)[1]
print([src_idx2word[int(i)] for i in enc_inputs[1]], '->',
[idx2word[n.item()] for n in predict.squeeze()])