Big TF test cleanup (#24282)

* Fix one BLIP arg not being optional, remove misspelled arg

* Remove the lxmert test overrides and just use the base test_saved_model_creation

* saved_model_creation fixes and re-enabling tests across the board

* Remove unnecessary skip

* Stop caching sinusoidal embeddings in speech_to_text

* Fix transfo_xl compilation

* Fix transfo_xl compilation

* Fix the conditionals in xglm

* Set the save spec only when building

* Clarify comment

* Move comment correctly

* Correct embeddings generation for speech2text

* Mark RAG generation tests as @slow

* Remove redundant else:

* Add comment to clarify the save_spec line in build()

* Fix size tests for XGLM at last!

* make fixup

* Remove one band_part operation

* Mark test_keras_fit as @slow

tests/models/led/test_modeling_tf_led.py

@@ -19,7 +19,7 @@ from __future__ import annotations
 import unittest
 
 from transformers import LEDConfig, is_tf_available
-from transformers.testing_utils import require_tf, slow, tooslow
+from transformers.testing_utils import require_tf, slow
 
 from ...test_configuration_common import ConfigTester
 from ...test_modeling_tf_common import TFModelTesterMixin, ids_tensor
@@ -292,11 +292,7 @@ class TFLEDModelTest(TFModelTesterMixin, PipelineTesterMixin, unittest.TestCase)
             self.assertEqual(model.config.output_hidden_states, True)
             check_encoder_attentions_output(outputs)
 
-    def test_xla_mode(self):
-        # TODO JP: Make LED XLA compliant
-        pass
-
-    @tooslow
+    @unittest.skip("LED keeps using potentially symbolic tensors in conditionals and breaks tracing.")
     def test_saved_model_creation(self):
         pass