[Tests] Add Common Test for Training + Fix a couple of bugs (#8415)

* add training tests

* correct longformer

* fix docs

* fix some tests

* fix some more train tests

* remove ipdb

* fix multiple edge case model training

* fix funnel and prophetnet

* clean gpt models

* undo renaming of albert

tests/test_modeling_mobilebert.py

@@ -27,6 +27,7 @@ if is_torch_available():
     import torch
 
     from transformers import (
+        MODEL_FOR_PRETRAINING_MAPPING,
         MobileBertConfig,
         MobileBertForMaskedLM,
         MobileBertForMultipleChoice,
@@ -220,7 +221,7 @@ class MobileBertModelTester:
             input_ids,
             attention_mask=input_mask,
             token_type_ids=token_type_ids,
-            next_sentence_label=sequence_labels,
+            labels=sequence_labels,
         )
         self.parent.assertEqual(result.logits.shape, (self.batch_size, 2))
 
@@ -327,6 +328,20 @@ class MobileBertModelTest(ModelTesterMixin, unittest.TestCase):
         else ()
     )
 
+    # special case for ForPreTraining model
+    def _prepare_for_class(self, inputs_dict, model_class, return_labels=False):
+        inputs_dict = super()._prepare_for_class(inputs_dict, model_class, return_labels=return_labels)
+
+        if return_labels:
+            if model_class in MODEL_FOR_PRETRAINING_MAPPING.values():
+                inputs_dict["labels"] = torch.zeros(
+                    (self.model_tester.batch_size, self.model_tester.seq_length), dtype=torch.long, device=torch_device
+                )
+                inputs_dict["next_sentence_label"] = torch.zeros(
+                    self.model_tester.batch_size, dtype=torch.long, device=torch_device
+                )
+        return inputs_dict
+
     def setUp(self):
         self.model_tester = MobileBertModelTester(self)
         self.config_tester = ConfigTester(self, config_class=MobileBertConfig, hidden_size=37)