深度学习——LSTM做情感分类任务_google colab 情感分类

Google Colab环境下运行

在Colab环境下安装包

使用GPU

获得训练集与测试集

定义LSTM模型

class RNN(nn.Module):
  def __init__(self,vocab_size,embedding_dim,hidden_dim):
    super(RNN,self).__init__()
    # [vocab_size 10000] => [embedding_dim:100]
    self.embedding = nn.Embedding(vocab_size,embedding_dim)
    # [embedding_dim 100] => [hidden_dim 256] 两层 双向lstm dropout防止过拟合
    self.rnn = nn.LSTM(embedding_dim,hidden_dim,num_layers=2,bidirectional=True,dropout=0.5)
    # lstm输出的h层信息[256*2] => 1
    self.fc = nn.Linear(hidden_dim*2,1)
    self.dropout = nn.Dropout(0.5)
  def forward(self,x):
    # [seq,b,1] => [seq,b,100]
    embedding = self.dropout(self.embedding(x))
    # output:[seq,b,hidden*2]
    # hidden/h:[num_layers*2,b,hid_dim]
    # cell/c:[num_layers*2,b,hid_dim]
    output,(hidden,cell) = self.rnn(embedding)
    # [num_layers*2,b,hid_dim] => 拼接2个[b,hid_dim] => [b,hid_dim*2]
    hidden = torch.cat([hidden[-2],hidden[-1]],dim=1)
    hidden = self.dropout(hidden)
    # [b,hid_dim*2] => [b,1]
    out = self.fc(hidden)
    return out
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

定义训练和评估函数

def train(rnn,iterator,optimizer,criterion):
  avg_acc = []
  rnn.train()

  for i,batch in enumerate(iterator):
    # [seq,b] => [b,1] = [b]
    pred = rnn(batch.text).squeeze(1)
    loss = criterion(pred,batch.label)
    acc = binary_acc(pred,batch.label).item()
    avg_acc.append(acc)

    optimizer.zero_grad()
    loss.backward()
    optimizer.step()

  if i%10==0:
    print(i,acc)
  avg_acc = np.array(avg_acc).mean()
  print('avg_acc',avg_acc)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

def binary_acc(preds,y):
  preds = torch.round(torch.sigmoid(preds))
  correct = torch.eq(preds,y).float()
  acc = correct.sum()/len(correct)
  return acc

def eval(rnn,iterator,criterion):
  avg_acc = []
  rnn.eval()
  with torch.no_grad():
    for batch in iterator:
      pred = rnn(batch.text).squeeze(1)
      loss = criterion(pred,batch.label)
      acc = binary_acc(pred,batch.label).item()
      avg_acc.append(acc)
  avg_acc = np.array(avg_acc).mean()
  print('>>test:',avg_acc)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

开始训练

1. 加载GLoVe预训练模型
   

2. 初始化模型 用GLoVe编码的权重 替换 模型的embedding权重

3. 设置batch_size optimizer criterion
   

4. epoch 训练
   

**关于embedding层的权重替换

用Gensim/GLoVe/Bert预训练获得的词向量进行LSTM训练 会获得更好的性能
可以见这篇blogNLP 预训练