Update forward signature test for vision models (#27681)

* Update forward signature

* Empty-Commit

tests/models/vit/test_modeling_vit.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch ViT model. """
 
 
-import inspect
 import unittest
 
 from transformers import ViTConfig
@@ -224,18 +223,6 @@ class ViTModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
             x = model.get_output_embeddings()
             self.assertTrue(x is None or isinstance(x, nn.Linear))
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/models/vit_hybrid/test_modeling_vit_hybrid.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch ViT Hybrid model. """
 
 
-import inspect
 import unittest
 
 from transformers import ViTHybridConfig
@@ -185,18 +184,6 @@ class ViTHybridModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCas
             x = model.get_output_embeddings()
             self.assertTrue(x is None or isinstance(x, nn.Linear))
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/models/vit_mae/test_modeling_vit_mae.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch ViTMAE model. """
 
 
-import inspect
 import math
 import tempfile
 import unittest
@@ -192,18 +191,6 @@ class ViTMAEModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
             x = model.get_output_embeddings()
             self.assertTrue(x is None or isinstance(x, nn.Linear))
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/models/vit_msn/test_modeling_vit_msn.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch ViTMSN model. """
 
 
-import inspect
 import unittest
 
 from transformers import ViTMSNConfig
@@ -183,18 +182,6 @@ class ViTMSNModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
             x = model.get_output_embeddings()
             self.assertTrue(x is None or isinstance(x, nn.Linear))
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/models/vitdet/test_modeling_vitdet.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch ViTDet model. """
 
 
-import inspect
 import unittest
 
 from transformers import VitDetConfig
@@ -210,18 +209,6 @@ class VitDetModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
             x = model.get_output_embeddings()
             self.assertTrue(x is None or isinstance(x, nn.Linear))
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/models/vitmatte/test_modeling_vitmatte.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch VitMatte model. """
 
 
-import inspect
 import unittest
 
 from huggingface_hub import hf_hub_download
@@ -189,18 +188,6 @@ class VitMatteModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase
     def test_model_common_attributes(self):
         pass
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/models/yolos/test_modeling_yolos.py

@@ -15,7 +15,6 @@
 """ Testing suite for the PyTorch YOLOS model. """
 
 
-import inspect
 import unittest
 
 from transformers import YolosConfig
@@ -217,18 +216,6 @@ class YolosModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
             x = model.get_output_embeddings()
             self.assertTrue(x is None or isinstance(x, nn.Linear))
 
-    def test_forward_signature(self):
-        config, _ = self.model_tester.prepare_config_and_inputs_for_common()
-
-        for model_class in self.all_model_classes:
-            model = model_class(config)
-            signature = inspect.signature(model.forward)
-            # signature.parameters is an OrderedDict => so arg_names order is deterministic
-            arg_names = [*signature.parameters.keys()]
-
-            expected_arg_names = ["pixel_values"]
-            self.assertListEqual(arg_names[:1], expected_arg_names)
-
     def test_model(self):
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)

tests/test_modeling_common.py

@@ -543,7 +543,7 @@ class ModelTesterMixin:
                 )
                 self.assertListEqual(arg_names[: len(expected_arg_names)], expected_arg_names)
             else:
-                expected_arg_names = ["input_ids"]
+                expected_arg_names = [model.main_input_name]
                 self.assertListEqual(arg_names[:1], expected_arg_names)
 
     def check_training_gradient_checkpointing(self, gradient_checkpointing_kwargs=None):