Add tf_keras imports to prepare for Keras 3 (#28588)

* Port core files + ESM (because ESM code is odd)

* Search-replace in modelling code

* Fix up transfo_xl as well

* Fix other core files + tests (still need to add correct import to tests)

* Fix cookiecutter

* make fixup, fix imports in some more core files

* Auto-add imports to tests

* Cleanup, add imports to sagemaker tests

* Use correct exception for importing tf_keras

* Fixes in modeling_tf_utils

* make fixup

* Correct version parsing code

* Ensure the pipeline tests correctly revert to float32 after each test

* Ensure the pipeline tests correctly revert to float32 after each test

* More tf.keras -> keras

* Add dtype cast

* Better imports of tf_keras

* Add a cast for tf.assign, just in case

* Fix callback imports

tests/models/encoder_decoder/test_modeling_tf_encoder_decoder.py

@@ -509,7 +509,7 @@ class TFEncoderDecoderMixin:
         tf_outputs = tf_model(tf_inputs_dict)
 
         # tf models returned loss is usually a tensor rather than a scalar.
-        # (see `hf_compute_loss`: it uses `tf.keras.losses.Reduction.NONE`)
+        # (see `hf_compute_loss`: it uses `keras.losses.Reduction.NONE`)
         # Change it here to a scalar to match PyTorch models' loss
         tf_loss = getattr(tf_outputs, "loss", None)
         if tf_loss is not None: