Rework pipeline tests (#19366)

* Rework pipeline tests

* Try to fix Flax tests

* Try to put it before

* Use a new decorator instead

* Remove ignore marker since it doesn't work

* Filter pipeline tests

* Woopsie

* Use the fitlered list

* Clean up and fake modif

* Remove init

* Revert fake modif

tests/pipelines/test_pipelines_common.py

@@ -48,13 +48,13 @@ from transformers.testing_utils import (
     USER,
     CaptureLogger,
     RequestCounter,
-    is_pipeline_test,
     is_staging_test,
     nested_simplify,
     require_scatter,
     require_tensorflow_probability,
     require_tf,
     require_torch,
+    require_torch_or_tf,
     slow,
 )
 from transformers.utils import is_tf_available, is_torch_available
@@ -307,7 +307,6 @@ class PipelineTestCaseMeta(type):
         return type.__new__(mcs, name, bases, dct)
 
 
-@is_pipeline_test
 class CommonPipelineTest(unittest.TestCase):
     @require_torch
     def test_pipeline_iteration(self):
@@ -416,7 +415,6 @@ class CommonPipelineTest(unittest.TestCase):
         self.assertEqual(len(outputs), 20)
 
 
-@is_pipeline_test
 class PipelinePadTest(unittest.TestCase):
     @require_torch
     def test_pipeline_padding(self):
@@ -498,7 +496,6 @@ class PipelinePadTest(unittest.TestCase):
         )
 
 
-@is_pipeline_test
 class PipelineUtilsTest(unittest.TestCase):
     @require_torch
     def test_pipeline_dataset(self):
@@ -795,7 +792,6 @@ class CustomPipeline(Pipeline):
         return model_outputs["logits"].softmax(-1).numpy()
 
 
-@is_pipeline_test
 class CustomPipelineTest(unittest.TestCase):
     def test_warning_logs(self):
         transformers_logging.set_verbosity_debug()
@@ -835,6 +831,7 @@ class CustomPipelineTest(unittest.TestCase):
         # Clean registry for next tests.
         del PIPELINE_REGISTRY.supported_tasks["custom-text-classification"]
 
+    @require_torch_or_tf
     def test_dynamic_pipeline(self):
         PIPELINE_REGISTRY.register_pipeline(
             "pair-classification",
@@ -886,6 +883,7 @@ class CustomPipelineTest(unittest.TestCase):
             [{"label": "LABEL_0", "score": 0.505}],
         )
 
+    @require_torch_or_tf
     def test_cached_pipeline_has_minimum_calls_to_head(self):
         # Make sure we have cached the pipeline.
         _ = pipeline("text-classification", model="hf-internal-testing/tiny-random-bert")