Change Keras preprocessing APIs to stable in the Parameter server strategy tutorial

PiperOrigin-RevId: 405774290

Author: tensorflower-gardener

site/en/tutorials/distribute/parameter_server_training.ipynb

@@ -187,8 +187,7 @@
         "import os\n",
         "import random\n",
         "import portpicker\n",
-        "import tensorflow as tf\n",
-        "from tensorflow.keras.layers.experimental import preprocessing"
+        "import tensorflow as tf"
       ]
     },
     {
@@ -527,7 +526,7 @@
         "\n",
         "You will create these layers outside the `dataset_fn` but apply the transformation inside the `dataset_fn`, since you will wrap the `dataset_fn` into a `tf.function`, which doesn't allow variables to be created inside it.\n",
         "\n",
-        "Note: There is a known performance implication when using lookup table resources, which layers, such as `tf.keras.layers.experimental.preprocessing.StringLookup`, employ. Refer to the [Known limitations](#known_limitations) section for more information."
+        "Note: There is a known performance implication when using lookup table resources, which layers, such as `tf.keras.layers.StringLookup`, employ. Refer to the [Known limitations](#known_limitations) section for more information."
       ]
     },
     {
@@ -544,10 +543,10 @@
         "label_vocab = [\"yes\", \"no\"]\n",
         "\n",
         "with strategy.scope():\n",
-        "  feature_lookup_layer = preprocessing.StringLookup(\n",
+        "  feature_lookup_layer = tf.keras.layers.StringLookup(\n",
         "      vocabulary=feature_vocab,\n",
         "      mask_token=None)\n",
-        "  label_lookup_layer = preprocessing.StringLookup(\n",
+        "  label_lookup_layer = tf.keras.layers.StringLookup(\n",
         "      vocabulary=label_vocab,\n",
         "      num_oov_indices=0,\n",
         "      mask_token=None)\n",
@@ -1264,7 +1263,7 @@
         "- It is not supported to load a saved_model via `tf.saved_model.load` containing sharded variables. Note loading such a saved_model using TensorFlow Serving is expected to work.\n",
         "- It is not supported to load a checkpoint containing sharded optimizer slot variables into a different number of shards.\n",
         "- It is not supported to recover from parameter server failure without restarting the coordinator task.\n",
-        "- Usage of `tf.lookup.StaticHashTable` (which is commonly employed by some `tf.keras.layers.experimental.preprocessing` layers, such as `IntegerLookup`, `StringLookup`, and `TextVectorization`) results in resources placed on the coordinator at this time with parameter server training. This has performance implications for lookup RPCs from workers to the coordinator. This is a current high priority to address.\n",
+        "- Usage of `tf.lookup.StaticHashTable` (which is commonly employed by some Keras preprocessing layers, such as `tf.keras.layers.IntegerLookup`, `tf.keras.layers.StringLookup`, and `tf.keras.layers.TextVectorization`) results in resources placed on the coordinator at this time with parameter server training. This has performance implications for lookup RPCs from workers to the coordinator. This is a current high priority to address.\n",
         "\n",
         "### `Model.fit` specifics\n",
         "\n",