Fix typos

keras_nlp/layers/modeling/masked_lm_head.py

@@ -31,7 +31,7 @@ class MaskedLMHead(keras.layers.Layer):
        predict with shape `(batch_size, masks_per_sequence)`.
 
     The token encodings should usually be the last output of an encoder model,
-    and mask positions should be the interger positions you would like to
+    and mask positions should be the integer positions you would like to
     predict for the MaskedLM task.
 
     The layer will first gather the token encodings at the mask positions. These

keras_nlp/layers/modeling/rotary_embedding_test.py

@@ -60,7 +60,7 @@ def test_multi_dimension_layer_output_shape(self):
         inputs = keras.Input(shape=(None, seq_length, hidden_size))
         outputs = embedding_layer(inputs)
 
-        # When using muliple dimensions before sequence length, the output is
+        # When using multiple dimensions before sequence length, the output is
         # expected to be the same as the input shape in all dimensions.
         expected_output_shape = (None, None, seq_length, hidden_size)
         self.assertEqual(expected_output_shape, outputs.shape)

keras_nlp/layers/modeling/sine_position_encoding_test.py

@@ -64,7 +64,7 @@ def test_multi_dimension_layer_output_shape(self):
         inputs = keras.Input(shape=(None, seq_length, hidden_size))
         outputs = pos_encoding(inputs)
 
-        # When using muliple dimensions before sequence length, the output is
+        # When using multiple dimensions before sequence length, the output is
         # expected to be the same as the input shape in all dimensions.
         expected_output_shape = (None, None, seq_length, hidden_size)
         self.assertEqual(expected_output_shape, outputs.shape)

keras_nlp/layers/preprocessing/multi_segment_packer.py

@@ -32,7 +32,7 @@ class MultiSegmentPacker(PreprocessingLayer):
     """Packs multiple sequences into a single fixed width model input.
 
     This layer packs multiple input sequences into a single fixed width sequence
-    containing start and end delimeters, forming an dense input suitable for a
+    containing start and end delimeters, forming a dense input suitable for a
     classification task for BERT and BERT-like models.
 
     Takes as input a tuple of token segments. Each tuple element should contain

keras_nlp/models/albert/albert_classifier.py

@@ -76,7 +76,7 @@ class AlbertClassifier(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     classifier.backbone.trainable = False
     # Fit again.
     classifier.fit(x=features, y=labels, batch_size=2)

keras_nlp/models/albert/albert_masked_lm.py

@@ -71,7 +71,7 @@ class AlbertMaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/models/backbone.py

@@ -128,7 +128,7 @@ def __init_subclass__(cls, **kwargs):
         super().__init_subclass__(**kwargs)
 
         # If the subclass does not define from_preset, assign a wrapper so that
-        # each class can have an distinct docstring.
+        # each class can have a distinct docstring.
         if "from_preset" not in cls.__dict__:
 
             def from_preset(calling_cls, *args, **kwargs):

keras_nlp/models/bart/bart_seq_2_seq_lm_preprocessor.py

@@ -209,7 +209,7 @@ def generate_preprocess(
         strings for values, tokenizes and packs the input, and computes a
         padding mask masking all inputs not filled in with a padded value.
 
-        Unlike calling the the layer for training, this method does not compute
+        Unlike calling the layer for training, this method does not compute
         labels and will never append a tokenizer.end_token_id to the end of
         the decoder sequence (as generation is expected to continue at the end
         of the inputted decoder prompt).

keras_nlp/models/bert/bert_classifier.py

@@ -77,7 +77,7 @@ class BertClassifier(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     classifier.backbone.trainable = False
     # Fit again.
     classifier.fit(x=features, y=labels, batch_size=2)

keras_nlp/models/bert/bert_masked_lm.py

@@ -70,7 +70,7 @@ class BertMaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/models/deberta_v3/deberta_v3_classifier.py

@@ -86,7 +86,7 @@ class DebertaV3Classifier(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     classifier.backbone.trainable = False
     # Fit again.
     classifier.fit(x=features, y=labels, batch_size=2)

keras_nlp/models/deberta_v3/deberta_v3_masked_lm.py

@@ -74,7 +74,7 @@ class DebertaV3MaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/models/distil_bert/distil_bert_classifier.py

@@ -88,7 +88,7 @@ class DistilBertClassifier(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     classifier.backbone.trainable = False
     # Fit again.
     classifier.fit(x=features, y=labels, batch_size=2)

keras_nlp/models/distil_bert/distil_bert_masked_lm.py

@@ -74,7 +74,7 @@ class DistilBertMaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/models/f_net/f_net_classifier.py

@@ -78,7 +78,7 @@ class FNetClassifier(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     classifier.backbone.trainable = False
     # Fit again.
     classifier.fit(x=features, y=labels, batch_size=2)

keras_nlp/models/f_net/f_net_masked_lm.py

@@ -69,7 +69,7 @@ class FNetMaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/models/generative_task.py

@@ -232,7 +232,7 @@ def generate(
                 `preprocessor` is attached to the model, `inputs` should match
                 the structure expected by the `preprocessor` layer. If a
                 `preprocessor` is not attached, `inputs` should match the
-                structure expected the the `backbone` model.
+                structure expected the `backbone` model.
             max_length: Optional. int. The max length of the generated sequence.
                 Will default to the max configured `sequence_length` of the
                 `preprocessor`. If `preprocessor` is `None`, `inputs` should be

keras_nlp/models/gpt2/gpt2_causal_lm.py

@@ -29,7 +29,7 @@
 
 @keras_nlp_export("keras_nlp.models.GPT2CausalLM")
 class GPT2CausalLM(GenerativeTask):
-    """An end-to-end GPT2 model for causal langauge modeling.
+    """An end-to-end GPT2 model for causal language modeling.
 
     A causal language model (LM) predicts the next token based on previous
     tokens. This task setup can be used to train the model unsupervised on

keras_nlp/models/gpt2/gpt2_causal_lm_preprocessor.py

@@ -139,7 +139,7 @@ def generate_preprocess(
         or tensor strings, tokenizes and packs the input, and computes a padding
         mask masking all inputs not filled in with a padded value.
 
-        Unlike calling the the layer for training, this method does not compute
+        Unlike calling the layer for training, this method does not compute
         labels and will never append a `tokenizer.end_token_id` to the end of
         the sequence (as generation is expected to continue at the end of the
         inputted prompt).

keras_nlp/models/gpt_neo_x/gpt_neo_x_causal_lm.py

@@ -26,7 +26,7 @@
 
 @keras_nlp_export("keras_nlp.models.GPTNeoXCausalLM")
 class GPTNeoXCausalLM(GenerativeTask):
-    """An end-to-end GPTNeoX model for causal langauge modeling.
+    """An end-to-end GPTNeoX model for causal language modeling.
 
     A causal language model (LM) predicts the next token based on previous
     tokens. This task setup can be used to train the model unsupervised on

keras_nlp/models/gpt_neo_x/gpt_neo_x_causal_lm_preprocessor.py

@@ -107,7 +107,7 @@ def generate_preprocess(
         or tensor strings, tokenizes and packs the input, and computes a padding
         mask masking all inputs not filled in with a padded value.
 
-        Unlike calling the the layer for training, this method does not compute
+        Unlike calling the layer for training, this method does not compute
         labels and will never append a `tokenizer.end_token_id` to the end of
         the sequence (as generation is expected to continue at the end of the
         inputted prompt).

keras_nlp/models/opt/opt_causal_lm.py

@@ -29,7 +29,7 @@
 
 @keras_nlp_export("keras_nlp.models.OPTCausalLM")
 class OPTCausalLM(GenerativeTask):
-    """An end-to-end OPT model for causal langauge modeling.
+    """An end-to-end OPT model for causal language modeling.
 
     A causal language model (LM) predicts the next token based on previous
     tokens. This task setup can be used to train the model unsupervised on

keras_nlp/models/opt/opt_causal_lm_preprocessor.py

@@ -140,7 +140,7 @@ def generate_preprocess(
         or tensor strings, tokenizes and packs the input, and computes a padding
         mask masking all inputs not filled in with a padded value.
 
-        Unlike calling the the layer for training, this method does not compute
+        Unlike calling the layer for training, this method does not compute
         labels and will never append a `tokenizer.end_token_id` to the end of
         the sequence (as generation is expected to continue at the end of the
         inputted prompt).
@@ -162,7 +162,7 @@ def generate_postprocess(
         """Covert integer token output to strings for generation.
 
         This method reverses `generate_preprocess()`, by first removing all
-        padding and start/end tokens, and then converting the interger sequence
+        padding and start/end tokens, and then converting the integer sequence
         back to a string.
         """
         token_ids, padding_mask = x["token_ids"], x["padding_mask"]

keras_nlp/models/preprocessor.py

@@ -125,7 +125,7 @@ def __init_subclass__(cls, **kwargs):
         super().__init_subclass__(**kwargs)
 
         # If the subclass does not define from_preset, assign a wrapper so that
-        # each class can have an distinct docstring.
+        # each class can have a distinct docstring.
         if "from_preset" not in cls.__dict__:
 
             def from_preset(calling_cls, *args, **kwargs):

keras_nlp/models/roberta/roberta_classifier.py

@@ -78,7 +78,7 @@ class RobertaClassifier(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     classifier.backbone.trainable = False
     # Fit again.
     classifier.fit(x=features, y=labels, batch_size=2)

keras_nlp/models/roberta/roberta_masked_lm.py

@@ -72,7 +72,7 @@ class RobertaMaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/models/xlm_roberta/xlm_roberta_masked_lm.py

@@ -75,7 +75,7 @@ class XLMRobertaMaskedLM(Task):
         optimizer=keras.optimizers.Adam(5e-5),
         jit_compile=True,
     )
-    # Access backbone programatically (e.g., to change `trainable`).
+    # Access backbone programmatically (e.g., to change `trainable`).
     masked_lm.backbone.trainable = False
     # Fit again.
     masked_lm.fit(x=features, batch_size=2)

keras_nlp/samplers/contrastive_sampler.py

@@ -185,7 +185,7 @@ def body(prompt, cache, index, logits, hidden_states):
                 (1 - self.alpha) * next_token_probabilities
                 - self.alpha * max_similarity_scores
             )
-            # Unflatten varibles to shape [batch_size, self.k, ...] for
+            # Unflatten variables to shape [batch_size, self.k, ...] for
             # gather purpose.
             unflat_score = unflatten_beams(accumulated_scores)
             unflat_prompt = unflatten_beams(prompt_beams)

keras_nlp/tokenizers/byte_pair_tokenizer.py

@@ -660,7 +660,7 @@ def __init_subclass__(cls, **kwargs):
         super().__init_subclass__(**kwargs)
 
         # If the subclass does not define from_preset, assign a wrapper so that
-        # each class can have an distinct docstring.
+        # each class can have a distinct docstring.
         if "from_preset" not in cls.__dict__:
 
             def from_preset(calling_cls, *args, **kwargs):

keras_nlp/tokenizers/sentence_piece_tokenizer.py

@@ -282,7 +282,7 @@ def __init_subclass__(cls, **kwargs):
         super().__init_subclass__(**kwargs)
 
         # If the subclass does not define from_preset, assign a wrapper so that
-        # each class can have an distinct docstring.
+        # each class can have a distinct docstring.
         if "from_preset" not in cls.__dict__:
 
             def from_preset(calling_cls, *args, **kwargs):