Error out in lookup layers when invert is true and output_mode is not int

This output is meaningless for IntegerLookup, and will error out for StringLookup.
An error on init will be more friendly.

PiperOrigin-RevId: 361255701
Change-Id: I4e022938cd8cd696546dd275c83a97d64f944c21

tensorflow/python/keras/layers/preprocessing/index_lookup.py

@@ -134,6 +134,10 @@ class IndexLookup(base_preprocessing_layer.CombinerPreprocessingLayer):
         layer_name=self.__class__.__name__,
         arg_name="output_mode")
 
+    if invert and output_mode != INT:
+      raise ValueError("`output_mode` must be {} when `invert` is true. You "
+                       "passed {}".format(INT, output_mode))
+
     self.invert = invert
     self.max_tokens = max_tokens
     self.num_oov_indices = num_oov_indices

tensorflow/python/keras/layers/preprocessing/index_lookup_test.py

@@ -1431,6 +1431,17 @@ class IndexLookupInverseVocabularyTest(
           dtype=dtypes.string,
           invert=True)
 
+  def test_non_int_output_fails(self):
+    with self.assertRaisesRegex(ValueError, "`output_mode` must be int"):
+      _ = get_layer_class()(
+          max_tokens=None,
+          num_oov_indices=1,
+          mask_token="",
+          oov_token="[OOV]",
+          dtype=dtypes.string,
+          output_mode=index_lookup.COUNT,
+          invert=True)
+
   def test_vocab_with_repeated_element_fails(self):
     vocab_data = ["earth", "earth", "wind", "and", "fire"]
     layer = get_layer_class()(