我的NLP实践之旅06 经历了这些日子的组队学习，这次的NLP实践之路也逐渐接近尾声了，在以后的日子里，我也要努力学习，多多加油啊，扯远了，下面进入正题
Bert
通过这次的组队学习，我终于接触到了目前最热门的自然语言处理模型之一的Bert，起初我听到Bert的概念时是完全懵逼的，不过后来在看了李宏毅在YouTube上的视频后自己逐渐理解了其中的具体含义，下面对Bert进行介绍。
BERT的全称是Bidirectional Encoder Representation from Transformers，即双向Transformer的Encoder，其实它是TransFormer模型的一部分，TransFormer模型的示意图如下：

文章图片

简单来说，Bert模型为TransFormer模型的左半部分，TransFromer模型训练时需要提供输入和输出数据，其中要输入的输出数据在预测下一段文本的情况下可以理解为下一个词，而Transformer的参数十分的庞大，故大多数情况只运用模型的左边来进行数据的预处理。
从上图中大家应该注意到了模型有一个结构叫Attention，这个结构可以理解为一个特殊的神经网络，在文本预测的场景中，我们通常需要基于上一段文本预测下一段话，因此需要模型有一定的记忆能力；已知的模型如LSTM等可以一定程度上解决这个问题，然而由于在神经网络中会对之前的信息进行削减，而且，关键的信息也可能隔字符出现，以此Attention机制应运而生。
Attention主要通过在神经网络隐层中将k、v、q矩阵进行相乘转换后输出的结果，其中k为键矩阵，v为值矩阵，q为问题矩阵。
【我的NLP实践之旅06】bert代码如下：

# build module import torch.nn as nn import torch.nn.functional as Fclass Attention(nn.Module): def __init__(self, hidden_size): super(Attention, self).__init__() self.weight = nn.Parameter(torch.Tensor(hidden_size, hidden_size)) self.weight.data.normal_(mean=0.0, std=0.05)self.bias = nn.Parameter(torch.Tensor(hidden_size)) b = np.zeros(hidden_size, dtype=np.float32) self.bias.data.copy_(torch.from_numpy(b))self.query = nn.Parameter(torch.Tensor(hidden_size)) self.query.data.normal_(mean=0.0, std=0.05)def forward(self, batch_hidden, batch_masks): # batch_hidden: b x len x hidden_size (2 * hidden_size of lstm) # batch_masks:b x len# linear key = torch.matmul(batch_hidden, self.weight) + self.bias# b x len x hidden# compute attention outputs = torch.matmul(key, self.query)# b x lenmasked_outputs = outputs.masked_fill((1 - batch_masks).bool(), float(-1e32))attn_scores = F.softmax(masked_outputs, dim=1)# b x len# 对于全零向量，-1e32的结果为 1/len, -inf为nan, 额外补0 masked_attn_scores = attn_scores.masked_fill((1 - batch_masks).bool(), 0.0)# sum weighted sources batch_outputs = torch.bmm(masked_attn_scores.unsqueeze(1), key).squeeze(1)# b x hiddenreturn batch_outputs, attn_scores# build word encoder bert_path = '../bert-mini/' dropout = 0.15from transformers import BertModelclass WordBertEncoder(nn.Module): def __init__(self): super(WordBertEncoder, self).__init__() self.dropout = nn.Dropout(dropout)self.tokenizer = WhitespaceTokenizer() self.bert = BertModel.from_pretrained(bert_path)self.pooled = False logging.info('Build Bert encoder with pooled {}.'.format(self.pooled))def encode(self, tokens): tokens = self.tokenizer.tokenize(tokens) return tokensdef get_bert_parameters(self): no_decay = ['bias', 'LayerNorm.weight'] optimizer_parameters = [ {'params': [p for n, p in self.bert.named_parameters() if not any(nd in n for nd in no_decay)], 'weight_decay': 0.01}, {'params': [p for n, p in self.bert.named_parameters() if any(nd in n for nd in no_decay)], 'weight_decay': 0.0} ] return optimizer_parametersdef forward(self, input_ids, token_type_ids): # input_ids: sen_num x bert_len # token_type_ids: sen_numx bert_len# sen_num x bert_len x 256, sen_num x 256 sequence_output, pooled_output = self.bert(input_ids=input_ids, token_type_ids=token_type_ids)if self.pooled: reps = pooled_output else: reps = sequence_output[:, 0, :]# sen_num x 256if self.training: reps = self.dropout(reps)return repsclass WhitespaceTokenizer(): """WhitespaceTokenizer with vocab."""def __init__(self): vocab_file = bert_path + 'vocab.txt' self._token2id = self.load_vocab(vocab_file) self._id2token = {v: k for k, v in self._token2id.items()} self.max_len = 256 self.unk = 1logging.info("Build Bert vocab with size %d." % (self.vocab_size))def load_vocab(self, vocab_file): f = open(vocab_file, 'r') lines = f.readlines() lines = list(map(lambda x: x.strip(), lines)) vocab = dict(zip(lines, range(len(lines)))) return vocabdef tokenize(self, tokens): assert len(tokens) <= self.max_len - 2 tokens = ["[CLS]"] + tokens + ["[SEP]"] output_tokens = self.token2id(tokens) return output_tokensdef token2id(self, xs): if isinstance(xs, list): return [self._token2id.get(x, self.unk) for x in xs] return self._token2id.get(xs, self.unk)@property def vocab_size(self): return len(self._id2token)# build sent encoder sent_hidden_size = 256 sent_num_layers = 2class SentEncoder(nn.Module): def __init__(self, sent_rep_size): super(SentEncoder, self).__init__() self.dropout = nn.Dropout(dropout)self.sent_lstm = nn.LSTM( input_size=sent_rep_size, hidden_size=sent_hidden_size, num_layers=sent_num_layers, batch_first=True, bidirectional=True )def forward(self, sent_reps, sent_masks): # sent_reps:b x doc_len x sent_rep_size # sent_masks: b x doc_lensent_hiddens, _ = self.sent_lstm(sent_reps)# b x doc_len x hidden*2 sent_hiddens = sent_hiddens * sent_masks.unsqueeze(2)if self.training: sent_hiddens = self.dropout(sent_hiddens)return sent_hiddens

# build model class Model(nn.Module): def __init__(self, vocab): super(Model, self).__init__() self.sent_rep_size = 256 self.doc_rep_size = sent_hidden_size * 2 self.all_parameters = {} parameters = [] self.word_encoder = WordBertEncoder() bert_parameters = self.word_encoder.get_bert_parameters()self.sent_encoder = SentEncoder(self.sent_rep_size) self.sent_attention = Attention(self.doc_rep_size) parameters.extend(list(filter(lambda p: p.requires_grad, self.sent_encoder.parameters()))) parameters.extend(list(filter(lambda p: p.requires_grad, self.sent_attention.parameters())))self.out = nn.Linear(self.doc_rep_size, vocab.label_size, bias=True) parameters.extend(list(filter(lambda p: p.requires_grad, self.out.parameters())))if use_cuda: self.to(device)if len(parameters) > 0: self.all_parameters["basic_parameters"] = parameters self.all_parameters["bert_parameters"] = bert_parameterslogging.info('Build model with bert word encoder, lstm sent encoder.')para_num = sum([np.prod(list(p.size())) for p in self.parameters()]) logging.info('Model param num: %.2f M.' % (para_num / 1e6))def forward(self, batch_inputs): # batch_inputs(batch_inputs1, batch_inputs2): b x doc_len x sent_len # batch_masks : b x doc_len x sent_len batch_inputs1, batch_inputs2, batch_masks = batch_inputs batch_size, max_doc_len, max_sent_len = batch_inputs1.shape[0], batch_inputs1.shape[1], batch_inputs1.shape[2] batch_inputs1 = batch_inputs1.view(batch_size * max_doc_len, max_sent_len)# sen_num x sent_len batch_inputs2 = batch_inputs2.view(batch_size * max_doc_len, max_sent_len)# sen_num x sent_len batch_masks = batch_masks.view(batch_size * max_doc_len, max_sent_len)# sen_num x sent_lensent_reps = self.word_encoder(batch_inputs1, batch_inputs2)# sen_num x sent_rep_sizesent_reps = sent_reps.view(batch_size, max_doc_len, self.sent_rep_size)# b x doc_len x sent_rep_size batch_masks = batch_masks.view(batch_size, max_doc_len, max_sent_len)# b x doc_len x max_sent_len sent_masks = batch_masks.bool().any(2).float()# b x doc_lensent_hiddens = self.sent_encoder(sent_reps, sent_masks)# b x doc_len x doc_rep_size doc_reps, atten_scores = self.sent_attention(sent_hiddens, sent_masks)# b x doc_rep_sizebatch_outputs = self.out(doc_reps)# b x num_labelsreturn batch_outputsmodel = Model(vocab)

# build optimizer learning_rate = 2e-4 bert_lr = 5e-5 decay = .75 decay_step = 1000 from transformers import AdamW, get_linear_schedule_with_warmupclass Optimizer: def __init__(self, model_parameters, steps): self.all_params = [] self.optims = [] self.schedulers = []for name, parameters in model_parameters.items(): if name.startswith("basic"): optim = torch.optim.Adam(parameters, lr=learning_rate) self.optims.append(optim)l = lambda step: decay ** (step // decay_step) scheduler = torch.optim.lr_scheduler.LambdaLR(optim, lr_lambda=l) self.schedulers.append(scheduler) self.all_params.extend(parameters) elif name.startswith("bert"): optim_bert = AdamW(parameters, bert_lr, eps=1e-8) self.optims.append(optim_bert)scheduler_bert = get_linear_schedule_with_warmup(optim_bert, 0, steps) self.schedulers.append(scheduler_bert)for group in parameters: for p in group['params']: self.all_params.append(p) else: Exception("no nameed parameters.")self.num = len(self.optims)def step(self): for optim, scheduler in zip(self.optims, self.schedulers): optim.step() scheduler.step() optim.zero_grad()def zero_grad(self): for optim in self.optims: optim.zero_grad()def get_lr(self): lrs = tuple(map(lambda x: x.get_lr()[-1], self.schedulers)) lr = ' %.5f' * self.num res = lr % lrs return res

# build trainerimport time from sklearn.metrics import classification_reportclip = 5.0 epochs = 1 early_stops = 3 log_interval = 50test_batch_size = 16 train_batch_size = 16save_model = './bert.bin' save_test = './bert.csv'class Trainer(): def __init__(self, model, vocab): self.model = model self.report = Trueself.train_data = https://www.it610.com/article/get_examples(train_data, model.word_encoder, vocab) self.batch_num = int(np.ceil(len(self.train_data) / float(train_batch_size))) self.dev_data = get_examples(dev_data, model.word_encoder, vocab) self.test_data = get_examples(test_data, model.word_encoder, vocab)# criterion self.criterion = nn.CrossEntropyLoss()# label name self.target_names = vocab.target_names# optimizer self.optimizer = Optimizer(model.all_parameters, steps=self.batch_num * epochs)# count self.step = 0 self.early_stop = -1 self.best_train_f1, self.best_dev_f1 = 0, 0 self.last_epoch = epochsdef train(self): logging.info('Start training...') for epoch in range(1, epochs + 1): train_f1 = self._train(epoch)dev_f1 = self._eval(epoch)if self.best_dev_f1 <= dev_f1: logging.info( "Exceed history dev = %.2f, current dev = %.2f" % (self.best_dev_f1, dev_f1)) torch.save(self.model.state_dict(), save_model)self.best_train_f1 = train_f1 self.best_dev_f1 = dev_f1 self.early_stop = 0 else: self.early_stop += 1 if self.early_stop == early_stops: logging.info( "Eearly stop in epoch %d, best train: %.2f, dev: %.2f" % ( epoch - early_stops, self.best_train_f1, self.best_dev_f1)) self.last_epoch = epoch break def test(self): self.model.load_state_dict(torch.load(save_model)) self._eval(self.last_epoch + 1, test=True)def _train(self, epoch): self.optimizer.zero_grad() self.model.train()start_time = time.time() epoch_start_time = time.time() overall_losses = 0 losses = 0 batch_idx = 1 y_pred = [] y_true = [] for batch_data in data_iter(self.train_data, train_batch_size, shuffle=True): torch.cuda.empty_cache() batch_inputs, batch_labels = self.batch2tensor(batch_data) batch_outputs = self.model(batch_inputs) loss = self.criterion(batch_outputs, batch_labels) loss.backward()loss_value = https://www.it610.com/article/loss.detach().cpu().item() losses += loss_value overall_losses += loss_valuey_pred.extend(torch.max(batch_outputs, dim=1)[1].cpu().numpy().tolist()) y_true.extend(batch_labels.cpu().numpy().tolist())nn.utils.clip_grad_norm_(self.optimizer.all_params, max_norm=clip) for optimizer, scheduler in zip(self.optimizer.optims, self.optimizer.schedulers): optimizer.step() scheduler.step() self.optimizer.zero_grad()self.step += 1if batch_idx % log_interval == 0: elapsed = time.time() - start_timelrs = self.optimizer.get_lr() logging.info('| epoch {:3d} | step {:3d} | batch {:3d}/{:3d} | lr{} | loss {:.4f} | s/batch {:.2f}'.format( epoch, self.step, batch_idx, self.batch_num, lrs, losses / log_interval, elapsed / log_interval))losses = 0 start_time = time.time()batch_idx += 1overall_losses /= self.batch_num during_time = time.time() - epoch_start_time# reformat overall_losses = reformat(overall_losses, 4) score, f1 = get_score(y_true, y_pred)logging.info( '| epoch {:3d} | score {} | f1 {} | loss {:.4f} | time {:.2f}'.format(epoch, score, f1, overall_losses, during_time)) if set(y_true) == set(y_pred) and self.report: report = classification_report(y_true, y_pred, digits=4, target_names=self.target_names) logging.info('\n' + report)return f1def _eval(self, epoch, test=False): self.model.eval() start_time = time.time() data = https://www.it610.com/article/self.test_data if test else self.dev_data y_pred = [] y_true = [] with torch.no_grad(): for batch_data in data_iter(data, test_batch_size, shuffle=False): torch.cuda.empty_cache() batch_inputs, batch_labels = self.batch2tensor(batch_data) batch_outputs = self.model(batch_inputs) y_pred.extend(torch.max(batch_outputs, dim=1)[1].cpu().numpy().tolist()) y_true.extend(batch_labels.cpu().numpy().tolist())score, f1 = get_score(y_true, y_pred)during_time = time.time() - start_timeif test: df = pd.DataFrame({'label': y_pred}) df.to_csv(save_test, index=False, sep=',') else: logging.info( '| epoch {:3d} | dev | score {} | f1 {} | time {:.2f}'.format(epoch, score, f1, during_time)) if set(y_true) == set(y_pred) and self.report: report = classification_report(y_true, y_pred, digits=4, target_names=self.target_names) logging.info('\n' + report)return f1def batch2tensor(self, batch_data): ''' [[label, doc_len, [[sent_len, [sent_id0, ...], [sent_id1, ...]], ...]] ''' batch_size = len(batch_data) doc_labels = [] doc_lens = [] doc_max_sent_len = [] for doc_data in batch_data: doc_labels.append(doc_data[0]) doc_lens.append(doc_data[1]) sent_lens = [sent_data[0] for sent_data in doc_data[2]] max_sent_len = max(sent_lens) doc_max_sent_len.append(max_sent_len)max_doc_len = max(doc_lens) max_sent_len = max(doc_max_sent_len)batch_inputs1 = torch.zeros((batch_size, max_doc_len, max_sent_len), dtype=torch.int64) batch_inputs2 = torch.zeros((batch_size, max_doc_len, max_sent_len), dtype=torch.int64) batch_masks = torch.zeros((batch_size, max_doc_len, max_sent_len), dtype=torch.float32) batch_labels = torch.LongTensor(doc_labels)for b in range(batch_size): for sent_idx in range(doc_lens[b]): sent_data = https://www.it610.com/article/batch_data[b][2][sent_idx] for word_idx in range(sent_data[0]): batch_inputs1[b, sent_idx, word_idx] = sent_data[1][word_idx] batch_inputs2[b, sent_idx, word_idx] = sent_data[2][word_idx] batch_masks[b, sent_idx, word_idx] = 1if use_cuda: batch_inputs1 = batch_inputs1.to(device) batch_inputs2 = batch_inputs2.to(device) batch_masks = batch_masks.to(device) batch_labels = batch_labels.to(device)return (batch_inputs1, batch_inputs2, batch_masks), batch_labels

# train trainer = Trainer(model, vocab) trainer.train()

推荐阅读

• 

  饮水机哪个品牌质量好
• 

  硕思logo设计师怎么去水印
• 

  《国王游戏》即将开始！你准备好了吗（）
• 

  老年人年龄大 智力并不差
• 

  怎么教孩子学数学 如何教小朋友学简谱，如何教孩子学认字
• 

  拼多多店铺品牌资质怎么弄
• 

  爱普生清零软件打不开怎么回事儿-手里微微儿老是自己关闭？
• 

  博世壁挂炉33代码故障怎么解决 博世壁挂炉出现故障代码
• 

  威能锅炉显示屏不停的闪怎么解决,这要根据具体情况来
• 

  兄弟j2320废墨清零-兄弟打印机mfcj2330dw如何清零，请发一
• 

  联想r7000为啥越来越贵？
• 

  嬉戏的近义词是什么 嬉戏的意思
• 

  C语言数据的存储超详细讲解下篇浮点型在内存中的存取
• 

  桔梗导航怎么彻底删除
• 

  afw，有没人知道AFW出入金是以什么方式
• 

  1831什么意思
• 

  胃疼吃什么食物 胃疼有什么饮食忌口
• 

  去刀疤痕最好的方法是什么？
• 

  重婚罪可以起诉离婚吗女方 重婚罪可以起诉离婚吗？
• 

  java语言基础知识点思维导图，Java的重要知识点详细扼要就行

• 我的NLP实践之旅02
• 我的NLP实践之旅05
• 我的博客|Vue.js2.0 入门实例（三）（数据对接）