Merge pull request #251 from Iwontbecreative/active_loss_tok_classif

Only keep the active part mof the loss for token classification

pytorch_pretrained_bert/modeling.py

@@ -1025,6 +1025,13 @@ class BertForTokenClassification(PreTrainedBertModel):
 
         if labels is not None:
             loss_fct = CrossEntropyLoss()
+            # Only keep active parts of the loss
+            if attention_mask is not None:
+                active_loss = attention_mask.view(-1) == 1
+                active_logits = logits.view(-1, self.num_labels)[active_loss]
+                active_labels = labels.view(-1)[active_loss]
+                loss = loss_fct(active_logits, active_labels)
+            else:
                 loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
             return loss
         else: