fix: same assertions in other models

modeling_for_glue.py

@@ -143,16 +143,16 @@ class BertForQuestionAnswering(BertPreTrainedModel):
             return_dict if return_dict is not None else self.config.use_return_dict
         )
 
+        assert head_mask is None
+        assert inputs_embeds is None
+        assert output_attentions is None
+        assert output_hidden_states is None
+        assert return_dict is None
         outputs = self.bert(
             input_ids,
             attention_mask=attention_mask,
             token_type_ids=token_type_ids,
             position_ids=position_ids,
-            head_mask=head_mask,
-            inputs_embeds=inputs_embeds,
-            output_attentions=output_attentions,
-            output_hidden_states=output_hidden_states,
-            return_dict=return_dict,
         )
 
         sequence_output = outputs[0]
@@ -230,16 +230,16 @@ class BertForTokenClassification(BertPreTrainedModel):
             return_dict if return_dict is not None else self.config.use_return_dict
         )
 
+        assert head_mask is None
+        assert inputs_embeds is None
+        assert output_attentions is None
+        assert output_hidden_states is None
+        assert return_dict is None
         outputs = self.bert(
             input_ids,
             attention_mask=attention_mask,
             token_type_ids=token_type_ids,
             position_ids=position_ids,
-            head_mask=head_mask,
-            inputs_embeds=inputs_embeds,
-            output_attentions=output_attentions,
-            output_hidden_states=output_hidden_states,
-            return_dict=return_dict,
         )
 
         sequence_output = outputs[0]