Fix past CI by skipping `LevitModelTest.test_problem_types` (#20859)

* Fix past CI

* Fix past CI
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

tests/models/levit/test_modeling_levit.py

@@ -20,6 +20,8 @@ import unittest
 import warnings
 from math import ceil, floor
 
+from packaging import version
+
 from transformers import LevitConfig
 from transformers.file_utils import cached_property, is_torch_available, is_vision_available
 from transformers.models.auto import get_values
@@ -335,6 +337,11 @@ class LevitModelTest(ModelTesterMixin, unittest.TestCase):
             loss.backward()
 
     def test_problem_types(self):
+
+        parsed_torch_version_base = version.parse(version.parse(torch.__version__).base_version)
+        if parsed_torch_version_base.base_version.startswith("1.9"):
+            self.skipTest(reason="This test fails with PyTorch 1.9.x: some CUDA issue")
+
         config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common()
 
         problem_types = [