计算模型中的两个损失并反向传播两次

Question

我正在使用 BertModel 创建一个模型来识别答案跨度（不使用 BertForQA）。

我有一个独立的线性层来分别确定开始和结束标记。在 init():

self.start_linear = nn.Linear(h, output_dim)

self.end_linear = nn.Linear(h, output_dim)

在forward()中，我输出了一个预测的起始层和预测的结束层：

 def forward(self, input_ids, attention_mask):

 outputs = self.bert(input_ids, attention_mask) # input = bert tokenizer encoding

 lhs = outputs.last_hidden_state # (batch_size, sequence_length, hidden_size)

 out = lhs[:, -1, :] # (batch_size, hidden_dim)

 st = self.start_linear(out)

 end = self.end_linear(out) 



 predict_start = self.softmax(st)

 predict_end = self.softmax(end)

 return predict_start, predict_end

然后在train_epoch()中，我尝试分别反向传播损失：

def train_epoch(model, train_loader, optimizer):

 model.train()

 total = 0

 st_loss, st_correct, st_total_loss = 0, 0, 0

 end_loss, end_correct, end_total_loss = 0, 0, 0

 for batch in train_loader:

   optimizer.zero_grad()

   input_ids = batch['input_ids'].to(device)

   attention_mask = batch['attention_mask'].to(device)

   start_idx = batch['start'].to(device)

   end_idx = batch['end'].to(device)

   start, end = model(input_ids=input_ids, attention_mask=attention_mask)


   st_loss = model.compute_loss(start, start_idx)

   end_loss = model.compute_loss(end, end_idx)

   st_total_loss += st_loss.item()

   end_total_loss += end_loss.item()

 # perform backward propagation to compute the gradients

   st_loss.backward()

   end_loss.backward()

 # update the weights

   optimizer.step()

但后来我上线了 end_loss.backward():

Trying to backward through the graph a second time, but the saved intermediate results have already been freed. Specify retain_graph=True when calling backward the first time.

我应该单独做反向传球吗？或者我应该用另一种方式来做吗？谢谢！

Answer 1

标准程序只是对损失求和并在总和上进行反向传播。

确保您要求和的两个损失的值平均大约一样大，或者至少与您希望每个损失相对于彼此的重要性成正比（否则，模型是将针对较大的损失而不是较小的损失进行优化）。在跨度检测案例中，我猜这不是必需的，但是由于问题的明显对称性。

计算模型中的两个损失并反向传播两次

calculate two losses in a model and backpropagate twice

python

nlp

pytorch

bert-language-model

huggingface-transformers