模型架构:
Seq2Seq(Sequence-to-Sequence)模型是一种在自然语言处理(NLP)中广泛应用的架构,其核心思想是将一个序列作为输入,并输出另一个序列。这种模型特别适用于机器翻译、聊天机器人、自动文摘等场景,其中输入和输出的长度都是可变的。
下载: https://download.pytorch.org/tutorial/data.zip
🍸️步骤:
基于GRU的seq2seq模型架构实现翻译的过程:
- # 从io工具包导入open方法
- from io import open
- # 用于字符规范化
- import unicodedata
- # 用于正则表达式
- import re
- # 用于随机生成数据
- import random
- # 用于构建网络结构和函数的torch工具包
- import torch
- import torch.nn as nn
- import torch.nn.functional as F
- # torch中预定义的优化方法工具包
- from torch import optim
- # 设备选择, 我们可以选择在cuda或者cpu上运行你的代码
- device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
将指定语言中的词汇映射成数值💫
- # 起始标志
- SOS_token = 0
- # 结束标志
- EOS_token = 1
-
- class Lang:
- def __init__(self, name):
- self.name = name
- self.word2index = {}
- self.index2word = {0: "SOS", 1: "EOS"}
- self.n_words = 2
-
- def addSentence(self, sentence):
- for word in sentence.split(' '):
- self.addWord(word)
-
-
- def addWord(self, word):
-
- if word not in self.word2index:
- self.word2index[word] = self.n_words
- self.index2word[self.n_words] = words
- self.n_words += 1
- name = "eng"
- sentence = "hello I am Jay"
-
- engl = Lang(name)
- engl.addSentence(sentence)
- print("word2index:", engl.word2index)
- print("index2word:", engl.index2word)
- print("n_words:", engl.n_words)
-
- # 输出
- word2index: {'hello': 2, 'I': 3, 'am': 4, 'Jay': 5}
- index2word: {0: 'SOS', 1: 'EOS', 2: 'hello', 3: 'I', 4: 'am', 5: 'Jay'}
- n_words: 6
字符规范化💫
-
- def unicodeToAscii(s):
- return ''.join(
- c for c in unicodedata.normalize('NFD', s)
- if unicodedata.category(c) != 'Mn'
- )
-
-
- def normalizeString(s):
- s = unicodeToAscii(s.lower().strip())
- s = re.sub(r"([.!?])", r" \1", s)
- s = re.sub(r"[^a-zA-Z.!?]+", r" ", s)
- return s
将文件中的数据加载到内存,实例化类Lang💫
- data_path = 'eng-fra.txt'
-
- def readLangs(lang1, lang2):
- """读取语言函数, 参数lang1是源语言的名字, 参数lang2是目标语言的名字
- 返回对应的class Lang对象, 以及语言对列表"""
- # 从文件中读取语言对并以/n划分存到列表lines中
- lines = open(data_path, encoding='utf-8').read().strip().split('\n')
- # 对lines列表中的句子进行标准化处理,并以\t进行再次划分, 形成子列表, 也就是语言对
- pairs = [[normalizeString(s) for s in l.split('\t')] for l in lines]
- # 然后分别将语言名字传入Lang类中, 获得对应的语言对象, 返回结果
- input_lang = Lang(lang1)
- output_lang = Lang(lang2)
- return input_lang, output_lang, pairs
- lang1 = "eng"
- lang2 = "fra"
-
- input_lang, output_lang, pairs = readLangs(lang1, lang2)
- print("pairs中的前五个:", pairs[:5])
-
- # 输出
- pairs中的前五个: [['go .', 'va !'], ['run !', 'cours !'], ['run !', 'courez !'], ['wow !', 'ca alors !'], ['fire !', 'au feu !']]
过滤出符合我们要求的语言对💫
- # 设置组成句子中单词或标点的最多个数
- MAX_LENGTH = 10
-
- eng_prefixes = (
- "i am ", "i m ",
- "he is", "he s ",
- "she is", "she s ",
- "you are", "you re ",
- "we are", "we re ",
- "they are", "they re "
- )
-
-
- def filterPair(p):
- return len(p[0].split(' ')) < MAX_LENGTH and \
- p[0].startswith(eng_prefixes) and \
- len(p[1].split(' ')) < MAX_LENGTH
-
-
- def filterPairs(pairs):
- return [pair for pair in pairs if filterPair(pair)]
对以上数据准备函数进行整合💫
- def prepareData(lang1, lang2):
-
- input_lang, output_lang, pairs = readLangs(lang1, lang2)
-
- pairs = filterPairs(pairs)
- for pair in pairs:
- input_lang.addSentence(pair[0])
- output_lang.addSentence(pair[1])
- return input_lang, output_lang, pairs
将语言对转化为模型输入需要的张量💫
- def tensorFromSentence(lang, sentence):
- indexes = [lang.word2index[word] for word in sentence.split(' ')]
-
- indexes.append(EOS_token)
- return torch.tensor(indexes, dtype=torch.long, device=device).view(-1, 1)
-
-
- def tensorsFromPair(pair):
-
- input_tensor = tensorFromSentence(input_lang, pair[0])
- target_tensor = tensorFromSentence(output_lang, pair[1])
- return (input_tensor, target_tensor)
- # 取pairs的第一条
- pair = pairs[0]
- pair_tensor = tensorsFromPair(pair)
- print(pair_tensor)
-
- # 输出
- (tensor([[2],
- [3],
- [4],
- [1]]),
- tensor([[2],
- [3],
- [4],
- [5],
- [1]]))
构建基于GRU的编码器
- class EncoderRNN(nn.Module):
- def __init__(self, input_size, hidden_size):
- super(EncoderRNN, self).__init__()
- self.hidden_size = hidden_size
-
- self.embedding = nn.Embedding(input_size, hidden_size)
- self.gru = nn.GRU(hidden_size, hidden_size)
-
- def forward(self, input, hidden):
- output = self.embedding(input).view(1, 1, -1)
- output, hidden = self.gru(output, hidden)
- return output, hidden
-
- def initHidden(self):
- return torch.zeros(1, 1, self.hidden_size, device=device)
- hidden_size = 25
- input_size = 20
-
- # pair_tensor[0]代表源语言即英文的句子,pair_tensor[0][0]代表句子中
- 的第一个词
- input = pair_tensor[0][0]
- # 初始化第一个隐层张量,1x1xhidden_size的0张量
- hidden = torch.zeros(1, 1, hidden_size)
-
- encoder = EncoderRNN(input_size, hidden_size)
- encoder_output, hidden = encoder(input, hidden)
- print(encoder_output)
-
- # 输出
- tensor([[[ 1.9149e-01, -2.0070e-01, -8.3882e-02, -3.3037e-02, -1.3491e-01,
- -8.8831e-02, -1.6626e-01, -1.9346e-01, -4.3996e-01, 1.8020e-02,
- 2.8854e-02, 2.2310e-01, 3.5153e-01, 2.9635e-01, 1.5030e-01,
- -8.5266e-02, -1.4909e-01, 2.4336e-04, -2.3522e-01, 1.1359e-01,
- 1.6439e-01, 1.4872e-01, -6.1619e-02, -1.0807e-02, 1.1216e-02]]],
- grad_fn=
)
构建基于GRU的解码器
- class DecoderRNN(nn.Module):
- def __init__(self, hidden_size, output_size):
-
- super(DecoderRNN, self).__init__()
-
- self.hidden_size = hidden_size
- self.embedding = nn.Embedding(output_size, hidden_size)
-
- self.gru = nn.GRU(hidden_size, hidden_size)
- self.out = nn.Linear(hidden_size, output_size)
- self.softmax = nn.LogSoftmax(dim=1)
-
-
- def forward(self, input, hidden):
-
- output = self.embedding(input).view(1, 1, -1)
- output = F.relu(output)
- output, hidden = self.gru(output, hidden)
-
- output = self.softmax(self.out(output[0]))
- return output, hidden
-
- def initHidden(self):
- return torch.zeros(1, 1, self.hidden_size, device=device)
构建基于GRU和Attention的解码器💥
💥三个输入:
input
是当前时间步解码器的输入。在解码的开始阶段,它可能是一个特殊的起始符号。在随后的解码步骤中,input
通常是上一个时间步解码器输出的词(或对应的词向量)。encoder_outputs
:是编码器处理输入序列后生成的一系列输出向量,在基于Attention的解码器中,这些输出向量将作为注意力机制的候选记忆单元,用于计算当前解码步与输入序列中不同位置的相关性。- class AttnDecoderRNN(nn.Module):
- def __init__(self, hidden_size, output_size, dropout_p=0.1, max_length=MAX_LENGTH):
-
- super(AttnDecoderRNN, self).__init__()
- self.hidden_size = hidden_size
- self.output_size = output_size
- self.dropout_p = dropout_p
- self.max_length = max_length
-
- self.embedding = nn.Embedding(self.output_size, self.hidden_size)
-
- self.attn = nn.Linear(self.hidden_size * 2, self.max_length)
- self.attn_combine = nn.Linear(self.hidden_size * 2, self.hidden_size)
- self.dropout = nn.Dropout(self.dropout_p)
-
- self.gru = nn.GRU(self.hidden_size, self.hidden_size)
- self.out = nn.Linear(self.hidden_size, self.output_size)
-
-
- def forward(self, input, hidden, encoder_outputs):
-
- embedded = self.embedding(input).view(1, 1, -1)
-
- embedded = self.dropout(embedded)
-
- attn_weights = F.softmax(
- self.attn(torch.cat((embedded[0], hidden[0]), 1)), dim=1)
-
- attn_applied = torch.bmm(attn_weights.unsqueeze(0),
- encoder_outputs.unsqueeze(0))
-
-
- output = torch.cat((embedded[0], attn_applied[0]), 1)
-
- output = self.attn_combine(output).unsqueeze(0)
-
- output = F.relu(output)
-
- output, hidden = self.gru(output, hidden)
-
-
- output = F.log_softmax(self.out(output[0]), dim=1)
-
- return output, hidden, attn_weights
-
- def initHidden(self):
-
- return torch.zeros(1, 1, self.hidden_size, device=device)
teacher_forcing介绍
Teacher Forcing是一种在训练序列生成模型,特别是循环神经网络(RNN)和序列到序列(seq2seq)模型时常用的技术。在seq2seq架构中,根据循环神经网络理论,解码器每次应该使用上一步的结果作为输入的一部分, 但是训练过程中,一旦上一步的结果是错误的,就会导致这种错误被累积,无法达到训练效果,我们需要一种机制改变上一步出错的情况,因为训练时我们是已知正确的输出应该是什么,因此可以强制将上一步结果设置成正确的输出, 这种方式就叫做teacher_forcing。
teacher_forcing的作用
- # 设置teacher_forcing比率为0.5
- teacher_forcing_ratio = 0.5
-
-
- def train(input_tensor, target_tensor, encoder, decoder, encoder_optimizer, decoder_optimizer, criterion, max_length=MAX_LENGTH):
-
- encoder_hidden = encoder.initHidden()
-
- encoder_optimizer.zero_grad()
- decoder_optimizer.zero_grad()
-
- input_length = input_tensor.size(0)
- target_length = target_tensor.size(0)
-
- encoder_outputs = torch.zeros(max_length, encoder.hidden_size, device=device)
-
- loss = 0
-
- for ei in range(input_length):
-
- encoder_output, encoder_hidden = encoder(
- input_tensor[ei], encoder_hidden)
-
- encoder_outputs[ei] = encoder_output[0, 0]
-
-
- decoder_input = torch.tensor([[SOS_token]], device=device)
-
- decoder_hidden = encoder_hidden
-
- use_teacher_forcing = True if random.random() < teacher_forcing_ratio else False
-
-
- if use_teacher_forcing:
-
- for di in range(target_length):
-
- decoder_output, decoder_hidden, decoder_attention = decoder(
- decoder_input, decoder_hidden, encoder_outputs)
-
- loss += criterion(decoder_output, target_tensor[di])
-
- decoder_input = target_tensor[di]
-
- else:
-
- for di in range(target_length):
- decoder_output, decoder_hidden, decoder_attention = decoder(
- decoder_input, decoder_hidden, encoder_outputs)
-
- topv, topi = decoder_output.topk(1)
-
- loss += criterion(decoder_output, target_tensor[di])
-
- if topi.squeeze().item() == EOS_token:
- break
-
- decoder_input = topi.squeeze().detach()
-
-
- # 误差进行反向传播
- loss.backward()
- # 编码器和解码器进行优化即参数更新
- encoder_optimizer.step()
- decoder_optimizer.step()
-
- # 返回平均损失
- return loss.item() / target_length
构建时间计算函数
- import time
- import math
-
- def timeSince(since):
- now = time.time()
- # 获得时间差
- s = now - since
- # 将秒转化为分钟
- m = math.floor(s / 60)
- s -= m * 60
- return '%dm %ds' % (m, s)
调用训练函数并打印日志和制图
- import matplotlib.pyplot as plt
-
- def trainIters(encoder, decoder, n_iters, print_every=1000, plot_every=100, learning_rate=0.01):
-
- start = time.time()
-
- plot_losses = []
-
- print_loss_total = 0
-
- plot_loss_total = 0
-
- encoder_optimizer = optim.SGD(encoder.parameters(), lr=learning_rate)
- decoder_optimizer = optim.SGD(decoder.parameters(), lr=learning_rate)
-
-
- criterion = nn.NLLLoss()
-
- for iter in range(1, n_iters + 1):
-
- training_pair = tensorsFromPair(random.choice(pairs))
-
- input_tensor = training_pair[0]
- target_tensor = training_pair[1]
-
-
- loss = train(input_tensor, target_tensor, encoder,
- decoder, encoder_optimizer, decoder_optimizer, criterion)
-
- print_loss_total += loss
- plot_loss_total += loss
-
-
- if iter % print_every == 0:
-
- print_loss_avg = print_loss_total / print_every
- print_loss_total = 0
- print('%s (%d %d%%) %.4f' % (timeSince(start),
- iter, iter / n_iters * 100, print_loss_avg))
-
- if iter % plot_every == 0:
- plot_loss_avg = plot_loss_total / plot_every
- plot_losses.append(plot_loss_avg)
- plot_loss_total = 0
-
-
- plt.figure()
- plt.plot(plot_losses)
- plt.savefig("loss.png")
💥训练模型:
- # 设置隐层大小为256 ,也是词嵌入维度
- hidden_size = 256
- # 通过input_lang.n_words获取输入词汇总数,与hidden_size一同传入EncoderRNN类中
- # 得到编码器对象encoder1
- encoder1 = EncoderRNN(input_lang.n_words, hidden_size).to(device)
-
- # 通过output_lang.n_words获取目标词汇总数,与hidden_size和dropout_p一同传入AttnDecoderRNN类中
- # 得到解码器对象attn_decoder1
- attn_decoder1 = AttnDecoderRNN(hidden_size, output_lang.n_words, dropout_p=0.1).to(device)
-
- # 设置迭代步数
- n_iters = 80000
- # 设置日志打印间隔
- print_every = 5000
-
- trainIters(encoder1, attn_decoder1, n_iters, print_every=print_every)
模型会不断打印loss损失值并且绘制图像
- def evaluate(encoder, decoder, sentence, max_length=MAX_LENGTH):
- with torch.no_grad():
- # 对输入的句子进行张量表示
- input_tensor = tensorFromSentence(input_lang, sentence)
- # 获得输入的句子长度
- input_length = input_tensor.size()[0]
- encoder_hidden = encoder.initHidden()
-
- encoder_outputs = torch.zeros(max_length, encoder.hidden_size, device=device)
-
- for ei in range(input_length):
-
- encoder_output, encoder_hidden = encoder(input_tensor[ei],
- encoder_hidden)
-
- encoder_outputs[ei] += encoder_output[0, 0]
-
- decoder_input = torch.tensor([[SOS_token]], device=device)
-
- decoder_hidden = encoder_hidden
-
- decoded_words = []
- # 初始化attention张量
- decoder_attentions = torch.zeros(max_length, max_length)
- # 开始循环解码
- for di in range(max_length):
-
- decoder_output, decoder_hidden, decoder_attention = decoder(
- decoder_input, decoder_hidden, encoder_outputs)
-
-
- decoder_attentions[di] = decoder_attention.data
- topv, topi = decoder_output.data.topk(1)
- if topi.item() == EOS_token:
- decoded_words.append('
' ) - break
-
- else:
-
- decoded_words.append(output_lang.index2word[topi.item()])
-
-
- decoder_input = topi.squeeze().detach()
- return decoded_words, decoder_attentions[:di + 1]
随机选择指定数量的数据进行评估
- def evaluateRandomly(encoder, decoder, n=6):
- for i in range(n):
- pair = random.choice(pairs)
- # > 代表输入
- print('>', pair[0])
- # = 代表正确的输出
- print('=', pair[1])
- # 调用evaluate进行预测
- output_words, attentions = evaluate(encoder, decoder, pair[0])
- # 将结果连成句子
- output_sentence = ' '.join(output_words)
- # < 代表模型的输出
- print('<', output_sentence)
- print('')
-
- evaluateRandomly(encoder1, attn_decoder1)
效果:
- > i m impressed with your french .
- = je suis impressionne par votre francais .
- < je suis impressionnee par votre francais .
-
- > i m more than a friend .
- = je suis plus qu une amie .
- < je suis plus qu une amie .
-
- > she is beautiful like her mother .
- = elle est belle comme sa mere .
- < elle est sa sa mere .
-
- > you re winning aren t you ?
- = vous gagnez n est ce pas ?
- < tu restez n est ce pas ?
-
- > he is angry with you .
- = il est en colere apres toi .
- < il est en colere apres toi .
-
- > you re very timid .
- = vous etes tres craintifs .
- < tu es tres craintive .
Attention张量制图
- sentence = "we re both teachers ."
- # 调用评估函数
- output_words, attentions = evaluate(
- encoder1, attn_decoder1, sentence)
- print(output_words)
- # 将attention张量转化成numpy, 使用matshow绘制
- plt.matshow(attentions.numpy())
- plt.savefig("attn.png")
如果迭代次数过少,训练不充分,那么注意力就不会很好:
💯迭代次数变大: