Correct @is_flaky test decoration (#31480)

* Correct @is_flaky decorator

tests/models/gemma/test_modeling_gemma.py

@@ -493,7 +493,7 @@ class GemmaModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMixi
     @require_flash_attn
     @require_torch_gpu
     @pytest.mark.flash_attn_test
-    @is_flaky
+    @is_flaky()
     @slow
     def test_flash_attn_2_equivalence(self):
         for model_class in self.all_model_classes:

tests/test_modeling_common.py

@@ -3407,7 +3407,7 @@ class ModelTesterMixin:
     @require_torch_gpu
     @mark.flash_attn_test
     @slow
-    @is_flaky
+    @is_flaky()
     def test_flash_attn_2_inference_equivalence(self):
         for model_class in self.all_model_classes:
             if not model_class._supports_flash_attn_2:
@@ -3501,7 +3501,7 @@ class ModelTesterMixin:
     @require_torch_gpu
     @mark.flash_attn_test
     @slow
-    @is_flaky
+    @is_flaky()
     def test_flash_attn_2_inference_equivalence_right_padding(self):
         for model_class in self.all_model_classes:
             if not model_class._supports_flash_attn_2:
@@ -3591,7 +3591,7 @@ class ModelTesterMixin:
     @require_torch_gpu
     @mark.flash_attn_test
     @slow
-    @is_flaky
+    @is_flaky()
     def test_flash_attn_2_generate_left_padding(self):
         for model_class in self.all_generative_model_classes:
             if not model_class._supports_flash_attn_2:
@@ -3635,7 +3635,7 @@ class ModelTesterMixin:
     @require_flash_attn
     @require_torch_gpu
     @mark.flash_attn_test
-    @is_flaky
+    @is_flaky()
     @slow
     def test_flash_attn_2_generate_padding_right(self):
         for model_class in self.all_generative_model_classes: