Skip tests properly (#31308)

* Skip tests properly

* [test_all]

* Add 'reason' as kwarg for skipTest

* [test_all] Fix up

* [test_all]

tests/models/oneformer/test_modeling_oneformer.py

@@ -377,7 +377,7 @@ class OneFormerModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCas
 
     def test_training(self):
         if not self.model_tester.is_training:
-            return
+            self.skipTest(reason="model_tester.is_training is set to False")
         # only OneFormerForUniversalSegmentation has the loss
         model_class = self.all_model_classes[1]
         (

tests/models/oneformer/test_processor_oneformer.py

@@ -209,6 +209,7 @@ class OneFormerProcessingTest(unittest.TestCase):
         self.assertTrue(hasattr(processor, "max_seq_length"))
         self.assertTrue(hasattr(processor, "task_seq_length"))
 
+    @unittest.skip
     def test_batch_feature(self):
         pass
 
@@ -397,6 +398,7 @@ class OneFormerProcessingTest(unittest.TestCase):
 
         return inputs
 
+    @unittest.skip
     def test_init_without_params(self):
         pass
 

tests/models/openai/test_tokenization_openai.py

@@ -131,7 +131,7 @@ class OpenAIGPTTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
                     padding="max_length",
                 )
 
-    # tokenizer has no padding token
+    @unittest.skip(reason="tokenizer has no padding token")
     def test_padding_different_model_input_name(self):
         pass
 

tests/models/opt/test_modeling_opt.py

@@ -322,7 +322,7 @@ class OPTModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMixin,
         result = model(input_ids, attention_mask=attention_mask, labels=sequence_labels)
         self.assertEqual(result.logits.shape, (self.model_tester.batch_size, self.model_tester.num_labels))
 
-    @unittest.skip("Does not work on the tiny model as we keep hitting edge cases.")
+    @unittest.skip(reason="Does not work on the tiny model as we keep hitting edge cases.")
     def test_model_parallelism(self):
         super().test_model_parallelism()
 

tests/models/owlv2/test_image_processing_owlv2.py

@@ -168,6 +168,6 @@ class Owlv2ImageProcessingTest(ImageProcessingTestMixin, unittest.TestCase):
                 f"Batch image bounding boxes fail. Expected {expected_boxes}, got {boxes}",
             )
 
-    @unittest.skip("OWLv2 doesn't treat 4 channel PIL and numpy consistently yet")  # FIXME Amy
+    @unittest.skip(reason="OWLv2 doesn't treat 4 channel PIL and numpy consistently yet")  # FIXME Amy
     def test_call_numpy_4_channels(self):
         pass

tests/models/owlv2/test_modeling_owlv2.py

@@ -494,7 +494,7 @@ class Owlv2ModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
 
     def _create_and_check_torchscript(self, config, inputs_dict):
         if not self.test_torchscript:
-            return
+            self.skipTest(reason="test_torchscript is set to False")
 
         configs_no_init = _config_zero_init(config)  # To be sure we have no Nan
         configs_no_init.torchscript = True
@@ -708,7 +708,7 @@ class Owlv2ForObjectDetectionTest(ModelTesterMixin, unittest.TestCase):
 
     def _create_and_check_torchscript(self, config, inputs_dict):
         if not self.test_torchscript:
-            return
+            self.skipTest(reason="test_torchscript is set to False")
 
         configs_no_init = _config_zero_init(config)  # To be sure we have no Nan
         configs_no_init.torchscript = True

tests/models/owlvit/test_modeling_owlvit.py

@@ -489,7 +489,7 @@ class OwlViTModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCase):
 
     def _create_and_check_torchscript(self, config, inputs_dict):
         if not self.test_torchscript:
-            return
+            self.skipTest(reason="test_torchscript is set to False")
 
         configs_no_init = _config_zero_init(config)  # To be sure we have no Nan
         configs_no_init.torchscript = True
@@ -701,7 +701,7 @@ class OwlViTForObjectDetectionTest(ModelTesterMixin, unittest.TestCase):
 
     def _create_and_check_torchscript(self, config, inputs_dict):
         if not self.test_torchscript:
-            return
+            self.skipTest(reason="test_torchscript is set to False")
 
         configs_no_init = _config_zero_init(config)  # To be sure we have no Nan
         configs_no_init.torchscript = True

tests/models/paligemma/test_modeling_paligemma.py

@@ -234,7 +234,7 @@ class PaliGemmaForConditionalGenerationModelTest(ModelTesterMixin, unittest.Test
         pass
 
     # TODO extend valid outputs to include this test @Molbap
-    @unittest.skip("PaliGemma has currently one output format.")
+    @unittest.skip(reason="PaliGemma has currently one output format.")
     def test_model_outputs_equivalence(self):
         pass
 

tests/models/patchtsmixer/test_modeling_patchtsmixer.py

@@ -317,7 +317,7 @@ class PatchTSMixerModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.Test
         for model_class in self.all_model_classes:
             check_hidden_states_output(inputs_dict, config, model_class)
 
-    @unittest.skip("No tokens embeddings")
+    @unittest.skip(reason="No tokens embeddings")
     def test_resize_tokens_embeddings(self):
         pass
 

tests/models/pegasus/test_modeling_pegasus.py

@@ -592,6 +592,6 @@ class PegasusStandaloneDecoderModelTest(ModelTesterMixin, GenerationTesterMixin,
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_decoder_model_attention_mask_past(*config_and_inputs)
 
+    @unittest.skip(reason="Decoder cannot keep gradients")
     def test_retain_grad_hidden_states_attentions(self):
-        # decoder cannot keep gradients
         return

tests/models/pegasus/test_tokenization_pegasus.py

@@ -128,10 +128,6 @@ class PegasusTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
             revision="ba85d0851d708441f91440d509690f1ab6353415",
         )
 
-    # @unittest.skip("We have to use from_slow")
-    # def test_added_tokens_serialization(self):
-    #     pass
-
 
 @require_sentencepiece
 @require_tokenizers
@@ -215,7 +211,3 @@ class BigBirdPegasusTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
             token_ids,
             [182, 117, 142, 587, 4211, 120, 117, 263, 112, 804, 109, 856, 25016, 3137, 464, 109, 26955, 3137, 1],
         )
-
-    # @unittest.skip("We have to use from_slow")
-    # def test_added_tokens_serialization(self):
-    #     pass

tests/models/pegasus_x/test_modeling_pegasus_x.py

@@ -872,6 +872,6 @@ class PegasusXStandaloneDecoderModelTest(ModelTesterMixin, GenerationTesterMixin
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_decoder_model_attention_mask_past(*config_and_inputs)
 
+    @unittest.skip(reason="Decoder cannot keep gradients")
     def test_retain_grad_hidden_states_attentions(self):
-        # decoder cannot keep gradients
         return

tests/models/perceiver/test_modeling_perceiver.py

@@ -387,7 +387,7 @@ class PerceiverModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCas
 
     def test_training(self):
         if not self.model_tester.is_training:
-            return
+            self.skipTest(reason="model_tester.is_training is set to False")
 
         for model_class in self.all_model_classes:
             if model_class.__name__ in [
@@ -732,7 +732,7 @@ class PerceiverModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCas
 
     def test_correct_missing_keys(self):
         if not self.test_missing_keys:
-            return
+            self.skipTest(reason="test_missing_keys is set to False")
         config, _ = self.model_tester.prepare_config_and_inputs_for_common()
 
         for model_class in self.all_model_classes:

tests/models/perceiver/test_tokenization_perceiver.py

@@ -270,15 +270,16 @@ class PerceiverTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
         tokenizer = self.perceiver_tokenizer
         self.assertEqual(tokenizer.decode([178]), "�")
 
-    # tokenizer does not have vocabulary
+    @unittest.skip(reason="tokenizer does not have vocabulary")
     def test_get_vocab(self):
         pass
 
-    # inputs cannot be pretokenized since ids depend on whole input string and not just on single characters
+    @unittest.skip(reason="inputs cannot be pretokenized")
     def test_pretokenized_inputs(self):
+        # inputs cannot be pretokenized since ids depend on whole input string and not just on single characters
         pass
 
-    # tests all ids in vocab => vocab doesn't exist so unnecessary to test
+    @unittest.skip(reason="vocab does not exist")
     def test_conversion_reversible(self):
         pass
 

tests/models/persimmon/test_modeling_persimmon.py

@@ -384,7 +384,7 @@ class PersimmonModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTester
             (self.model_tester.batch_size, self.model_tester.seq_length, self.model_tester.num_labels),
         )
 
-    @unittest.skip("Persimmon buffers include complex numbers, which breaks this test")
+    @unittest.skip(reason="Persimmon buffers include complex numbers, which breaks this test")
     # Copied from tests.models.llama.test_modeling_llama.LlamaModelTest.test_save_load_fast_init_from_base
     def test_save_load_fast_init_from_base(self):
         pass

tests/models/pix2struct/test_image_processing_pix2struct.py

@@ -335,14 +335,16 @@ class Pix2StructImageProcessingTestFourChannels(ImageProcessingTestMixin, unitte
                 (self.image_processor_tester.batch_size, max_patch, expected_hidden_dim),
             )
 
-    @unittest.skip("Pix2StructImageProcessor does not support 4 channels yet")  # FIXME Amy
+    @unittest.skip(reason="Pix2StructImageProcessor does not support 4 channels yet")  # FIXME Amy
     def test_call_numpy(self):
         return super().test_call_numpy()
 
-    @unittest.skip("Pix2StructImageProcessor does not support 4 channels yet")  # FIXME Amy
+    @unittest.skip(reason="Pix2StructImageProcessor does not support 4 channels yet")  # FIXME Amy
     def test_call_pytorch(self):
         return super().test_call_torch()
 
-    @unittest.skip("Pix2StructImageProcessor does treat numpy and PIL 4 channel images consistently")  # FIXME Amy
+    @unittest.skip(
+        reason="Pix2StructImageProcessor does treat numpy and PIL 4 channel images consistently"
+    )  # FIXME Amy
     def test_call_numpy_4_channels(self):
         return super().test_call_torch()

tests/models/pix2struct/test_modeling_pix2struct.py

@@ -489,7 +489,7 @@ class Pix2StructModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
 
     def test_training(self):
         if not self.model_tester.is_training:
-            return
+            self.skipTest(reason="model_tester.is_training is set to False")
 
         for model_class in self.all_model_classes[:-1]:
             config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
@@ -508,7 +508,7 @@ class Pix2StructModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
 
     def test_training_gradient_checkpointing(self):
         if not self.model_tester.is_training:
-            return
+            self.skipTest(reason="model_tester.is_training is set to False")
 
         for model_class in self.all_model_classes[:-1]:
             config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
@@ -555,7 +555,7 @@ class Pix2StructModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
     def test_resize_tokens_embeddings(self):
         original_config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
         if not self.test_resize_embeddings:
-            return
+            self.skipTest(reason="test_resize_embeddings is set to False")
 
         for model_class in self.all_model_classes:
             config = copy.deepcopy(original_config)
@@ -602,13 +602,13 @@ class Pix2StructModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
     def test_resize_embeddings_untied(self):
         original_config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
         if not self.test_resize_embeddings:
-            return
+            self.skipTest(reason="test_resize_embeddings is set to False")
 
         original_config.tie_word_embeddings = False
 
         # if model cannot untied embeddings -> leave test
         if original_config.tie_word_embeddings:
-            return
+            self.skipTest(reason="Model cannot untie embeddings")
 
         for model_class in self.all_model_classes:
             config = copy.deepcopy(original_config)
@@ -652,7 +652,7 @@ class Pix2StructModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
 
     def _create_and_check_torchscript(self, config, inputs_dict):
         if not self.test_torchscript:
-            return
+            self.skipTest(reason="test_torchscript is set to False")
 
         configs_no_init = _config_zero_init(config)  # To be sure we have no Nan
         configs_no_init.torchscript = True

tests/models/plbart/test_modeling_plbart.py

@@ -319,7 +319,7 @@ class PLBartModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMix
         model.generate(input_ids, attention_mask=attention_mask)
         model.generate(num_beams=4, do_sample=True, early_stopping=False, num_return_sequences=3)
 
-    @unittest.skip("Failing since #26752")
+    @unittest.skip(reason="Failing since #26752")
     def test_sample_generate(self):
         pass
 
@@ -664,6 +664,6 @@ class PLBartStandaloneDecoderModelTest(ModelTesterMixin, GenerationTesterMixin,
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_decoder_model_attention_mask_past(*config_and_inputs)
 
+    @unittest.skip(reason="Decoder cannot keep gradients")
     def test_retain_grad_hidden_states_attentions(self):
-        # decoder cannot keep gradients
         return

tests/models/poolformer/test_modeling_poolformer.py

@@ -144,11 +144,11 @@ class PoolFormerModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
         config_and_inputs = self.model_tester.prepare_config_and_inputs()
         self.model_tester.create_and_check_model(*config_and_inputs)
 
-    @unittest.skip("PoolFormer does not use inputs_embeds")
+    @unittest.skip(reason="PoolFormer does not use inputs_embeds")
     def test_inputs_embeds(self):
         pass
 
-    @unittest.skip("PoolFormer does not have get_input_embeddings method and get_output_embeddings methods")
+    @unittest.skip(reason="PoolFormer does not have get_input_embeddings method and get_output_embeddings methods")
     def test_model_get_set_embeddings(self):
         pass
 
@@ -190,7 +190,7 @@ class PoolFormerModelTest(ModelTesterMixin, PipelineTesterMixin, unittest.TestCa
 
     def test_training(self):
         if not self.model_tester.is_training:
-            return
+            self.skipTest(reason="model_tester.is_training is set to False")
 
         config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
         config.return_dict = True

tests/models/pop2piano/test_feature_extraction_pop2piano.py

@@ -280,14 +280,14 @@ class Pop2PianoFeatureExtractionTest(SequenceFeatureExtractionTestMixin, unittes
     def test_padding_from_array(self):
         pass
 
-    @unittest.skip("Pop2PianoFeatureExtractor does not support truncation")
+    @unittest.skip(reason="Pop2PianoFeatureExtractor does not support truncation")
     def test_attention_mask_with_truncation(self):
         pass
 
-    @unittest.skip("Pop2PianoFeatureExtractor does not supports truncation")
+    @unittest.skip(reason="Pop2PianoFeatureExtractor does not supports truncation")
     def test_truncation_from_array(self):
         pass
 
-    @unittest.skip("Pop2PianoFeatureExtractor does not supports truncation")
+    @unittest.skip(reason="Pop2PianoFeatureExtractor does not supports truncation")
     def test_truncation_from_list(self):
         pass