From cecbeb296109972ce2978a1073f41c823eafed44 Mon Sep 17 00:00:00 2001
From: Matt Watson <1389937+mattdangerw@users.noreply.github.com>
Date: Thu, 21 Sep 2023 12:34:48 -0700
Subject: [PATCH] Small docstring fixes for the upcoming release (#1253)

keras-io renders `Call arguments:` as a special keyword, but not
`Call args:`.
---
 keras_nlp/layers/modeling/position_embedding.py     | 2 +-
 keras_nlp/layers/modeling/reversible_embedding.py   | 4 ++--
 keras_nlp/layers/modeling/rotary_embedding.py       | 2 +-
 keras_nlp/layers/modeling/sine_position_encoding.py | 2 +-
 keras_nlp/models/xlnet/xlnet_backbone.py            | 2 +-
 5 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/keras_nlp/layers/modeling/position_embedding.py b/keras_nlp/layers/modeling/position_embedding.py
index 13a6e00a3b..6f9a44c29f 100644
--- a/keras_nlp/layers/modeling/position_embedding.py
+++ b/keras_nlp/layers/modeling/position_embedding.py
@@ -34,7 +34,7 @@ class PositionEmbedding(keras.layers.Layer):
             to `"glorot_uniform"`.
         seq_axis: The axis of the input tensor where we add the embeddings.
 
-    Call args:
+    Call arguments:
         inputs: The tensor inputs to compute an embedding for, with shape
             `(batch_size, sequence_length, hidden_dim)`. Only the input shape
             will be used, as the position embedding does not depend on the
diff --git a/keras_nlp/layers/modeling/reversible_embedding.py b/keras_nlp/layers/modeling/reversible_embedding.py
index e01d66983c..baa5fb7027 100644
--- a/keras_nlp/layers/modeling/reversible_embedding.py
+++ b/keras_nlp/layers/modeling/reversible_embedding.py
@@ -53,7 +53,7 @@ class ReversibleEmbedding(keras.layers.Embedding):
             For stability, it is usually best to use full precision even when
             working with half or mixed precision training.
 
-    Call args:
+    Call arguments:
         inputs: The tensor inputs to the layer.
         reverse: Boolean. If `True` the layer will perform a linear projection
             from `output_dim` to `input_dim`, instead of a normal embedding
@@ -69,7 +69,7 @@ class ReversibleEmbedding(keras.layers.Embedding):
     # Generate random inputs.
     token_ids = np.random.randint(vocab_size, size=(batch_size, seq_length))
 
-    embedding = keras.layers.Embedding(vocab_size, hidden_dim)
+    embedding = keras_nlp.layers.ReversibleEmbedding(vocab_size, hidden_dim)
     # Embed tokens to shape `(batch_size, seq_length, hidden_dim)`.
     hidden_states = embedding(token_ids)
     # Project hidden states to shape `(batch_size, seq_length, vocab_size)`.
diff --git a/keras_nlp/layers/modeling/rotary_embedding.py b/keras_nlp/layers/modeling/rotary_embedding.py
index e5723b7cf3..b3402f7e21 100644
--- a/keras_nlp/layers/modeling/rotary_embedding.py
+++ b/keras_nlp/layers/modeling/rotary_embedding.py
@@ -39,7 +39,7 @@ class RotaryEmbedding(keras.layers.Layer):
         sequence_axis: int. Sequence axis in the input tensor.
         feature_axis: int. Feature axis in the input tensor.
 
-    Call args:
+    Call arguments:
         inputs: The tensor inputs to apply the embedding to. This can have
             any shape, but must contain both a sequence and feature axis. The
             rotary embedding will be applied to `inputs` and returned.
diff --git a/keras_nlp/layers/modeling/sine_position_encoding.py b/keras_nlp/layers/modeling/sine_position_encoding.py
index 84d042a7bb..6e96a77e2c 100644
--- a/keras_nlp/layers/modeling/sine_position_encoding.py
+++ b/keras_nlp/layers/modeling/sine_position_encoding.py
@@ -35,7 +35,7 @@ class SinePositionEncoding(keras.layers.Layer):
             curves, as described in Attention is All You Need. Defaults to
             `10000`.
 
-    Call args:
+    Call arguments:
         inputs: The tensor inputs to compute an embedding for, with shape
             `(batch_size, sequence_length, hidden_dim)`.
         start_index: An integer or integer tensor. The starting position to
diff --git a/keras_nlp/models/xlnet/xlnet_backbone.py b/keras_nlp/models/xlnet/xlnet_backbone.py
index 3f3a8b771e..1d1b4d2343 100644
--- a/keras_nlp/models/xlnet/xlnet_backbone.py
+++ b/keras_nlp/models/xlnet/xlnet_backbone.py
@@ -53,7 +53,7 @@ class XLNetBackbone(Backbone):
             defaults to "zeros". The bias initializer for
             the dense and multiheaded relative attention layers.
 
-    Call Args:
+    Call arguments:
         token_ids: Indices of input sequence tokens in the vocabulary of shape
             `[batch_size, sequence_length]`.
         segment_ids: Segment token indices to indicate first and second portions