keras-hub-nightly 0.15.0.dev20240823171555__py3-none-any.whl → 0.15.0.dev20240911134614__py3-none-any.whl

keras_hub/src/models/gemma/gemma_causal_lm_preprocessor.py

@@ -12,19 +12,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import keras
-from absl import logging
-
 from keras_hub.src.api_export import keras_hub_export
-from keras_hub.src.models.gemma.gemma_preprocessor import GemmaPreprocessor
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
-)
-from keras_hub.src.utils.tensor_utils import strip_to_ragged
+from keras_hub.src.models.causal_lm_preprocessor import CausalLMPreprocessor
+from keras_hub.src.models.gemma.gemma_backbone import GemmaBackbone
+from keras_hub.src.models.gemma.gemma_tokenizer import GemmaTokenizer
 
 
 @keras_hub_export("keras_hub.models.GemmaCausalLMPreprocessor")
-class GemmaCausalLMPreprocessor(GemmaPreprocessor):
+class GemmaCausalLMPreprocessor(CausalLMPreprocessor):
     """Gemma Causal LM preprocessor.
 
     This preprocessing layer is meant for use with
@@ -84,84 +79,5 @@ class GemmaCausalLMPreprocessor(GemmaPreprocessor):
     ```
     """
 
-    def call(
-        self,
-        x,
-        y=None,
-        sample_weight=None,
-        sequence_length=None,
-    ):
-        if y is not None or sample_weight is not None:
-            logging.warning(
-                "`GemmaCausalLMPreprocessor` generates `y` and `sample_weight` "
-                "based on your input data, but your data already contains `y` "
-                "or `sample_weight`. Your `y` and `sample_weight` will be "
-                "ignored."
-            )
-        sequence_length = sequence_length or self.sequence_length
-
-        x = convert_inputs_to_list_of_tensor_segments(x)[0]
-        x = self.tokenizer(x)
-        # Pad with one extra token to account for the truncation below.
-        token_ids, padding_mask = self.packer(
-            x,
-            sequence_length=sequence_length + 1,
-            add_start_value=self.add_start_token,
-            add_end_value=self.add_end_token,
-        )
-        # The last token does not have a next token, so we truncate it out.
-        x = {
-            "token_ids": token_ids[..., :-1],
-            "padding_mask": padding_mask[..., :-1],
-        }
-        # Target `y` will be the next token.
-        y, sample_weight = token_ids[..., 1:], padding_mask[..., 1:]
-        return keras.utils.pack_x_y_sample_weight(x, y, sample_weight)
-
-    def generate_preprocess(
-        self,
-        x,
-        sequence_length=None,
-    ):
-        """Convert strings to integer token input for generation.
-
-        Similar to calling the layer for training, this method takes in strings
-        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 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).
-        """
-        if not self.built:
-            self.build(None)
-
-        x = convert_inputs_to_list_of_tensor_segments(x)[0]
-        x = self.tokenizer(x)
-        token_ids, padding_mask = self.packer(
-            x, sequence_length=sequence_length, add_end_value=False
-        )
-        return {
-            "token_ids": token_ids,
-            "padding_mask": padding_mask,
-        }
-
-    def generate_postprocess(self, x):
-        """Convert 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 integer sequence
-        back to a string.
-        """
-        if not self.built:
-            self.build(None)
-
-        token_ids, padding_mask = x["token_ids"], x["padding_mask"]
-        ids_to_strip = (
-            self.tokenizer.start_token_id,
-            self.tokenizer.end_token_id,
-            self.tokenizer.pad_token_id,
-        )
-        token_ids = strip_to_ragged(token_ids, padding_mask, ids_to_strip)
-        return self.tokenizer.detokenize(token_ids)
+    backbone_cls = GemmaBackbone
+    tokenizer_cls = GemmaTokenizer

keras_hub/src/models/gemma/gemma_preprocessor.py

@@ -17,11 +17,10 @@ import keras
 
 from keras_hub.src.api_export import keras_hub_export
 from keras_hub.src.layers.preprocessing.start_end_packer import StartEndPacker
+from keras_hub.src.models.gemma.gemma_backbone import GemmaBackbone
 from keras_hub.src.models.gemma.gemma_tokenizer import GemmaTokenizer
 from keras_hub.src.models.preprocessor import Preprocessor
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
-)
+from keras_hub.src.utils.tensor_utils import preprocessing_function
 
 
 @keras_hub_export("keras_hub.models.GemmaPreprocessor")
@@ -122,6 +121,7 @@ class GemmaPreprocessor(Preprocessor):
     ```
     """
 
+    backbone_cls = GemmaBackbone
     tokenizer_cls = GemmaTokenizer
 
     def __init__(
@@ -151,6 +151,7 @@ class GemmaPreprocessor(Preprocessor):
         )
         self.built = True
 
+    @preprocessing_function
     def call(
         self,
         x,
@@ -158,17 +159,9 @@ class GemmaPreprocessor(Preprocessor):
         sample_weight=None,
         sequence_length=None,
     ):
-        x = convert_inputs_to_list_of_tensor_segments(x)
-        if len(x) != 1:
-            raise ValueError(
-                "GemmaPreprocessor requires each input to contain only "
-                f"one segment, but received {len(x)}. If you are using Gemma "
-                "for a multi-segment classification task, please combine your "
-                "input into a single string."
-            )
         sequence_length = sequence_length or self.sequence_length
         token_ids, padding_mask = self.packer(
-            self.tokenizer(x[0]),
+            self.tokenizer(x),
             sequence_length=sequence_length,
             add_start_value=self.add_start_token,
             add_end_value=self.add_end_token,

keras_hub/src/models/gemma/gemma_tokenizer.py

@@ -13,12 +13,18 @@
 # limitations under the License.
 
 from keras_hub.src.api_export import keras_hub_export
+from keras_hub.src.models.gemma.gemma_backbone import GemmaBackbone
 from keras_hub.src.tokenizers.sentence_piece_tokenizer import (
     SentencePieceTokenizer,
 )
 
 
-@keras_hub_export("keras_hub.models.GemmaTokenizer")
+@keras_hub_export(
+    [
+        "keras_hub.tokenizers.GemmaTokenizer",
+        "keras_hub.models.GemmaTokenizer",
+    ]
+)
 class GemmaTokenizer(SentencePieceTokenizer):
     """Gemma tokenizer layer based on SentencePiece.
 
@@ -77,27 +83,10 @@ class GemmaTokenizer(SentencePieceTokenizer):
     ```
     """
 
-    def __init__(self, proto, **kwargs):
-        self.start_token = "<bos>"
-        self.end_token = "<eos>"
-        self.pad_token = "<pad>"
+    backbone_cls = GemmaBackbone
 
+    def __init__(self, proto, **kwargs):
+        self._add_special_token("<bos>", "start_token")
+        self._add_special_token("<eos>", "end_token")
+        self._add_special_token("<pad>", "pad_token")
         super().__init__(proto=proto, **kwargs)
-
-    def set_proto(self, proto):
-        super().set_proto(proto)
-        if proto is not None:
-            for token in [self.end_token, self.pad_token]:
-                if token not in self.get_vocabulary():
-                    raise ValueError(
-                        f"Cannot find token `'{token}'` in the provided "
-                        f"`vocabulary`. Please provide `'{token}'` in your "
-                        "`vocabulary` or use a pretrained `vocabulary` name."
-                    )
-            self.start_token_id = self.token_to_id(self.start_token)
-            self.end_token_id = self.token_to_id(self.end_token)
-            self.pad_token_id = self.token_to_id(self.pad_token)
-        else:
-            self.start_token_id = None
-            self.end_token_id = None
-            self.pad_token_id = None

keras_hub/src/models/gpt2/__init__.py

@@ -14,7 +14,6 @@
 
 from keras_hub.src.models.gpt2.gpt2_backbone import GPT2Backbone
 from keras_hub.src.models.gpt2.gpt2_presets import backbone_presets
-from keras_hub.src.models.gpt2.gpt2_tokenizer import GPT2Tokenizer
 from keras_hub.src.utils.preset_utils import register_presets
 
-register_presets(backbone_presets, (GPT2Backbone, GPT2Tokenizer))
+register_presets(backbone_presets, GPT2Backbone)

keras_hub/src/models/gpt2/gpt2_causal_lm_preprocessor.py

@@ -12,19 +12,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import keras
-from absl import logging
-
 from keras_hub.src.api_export import keras_hub_export
-from keras_hub.src.models.gpt2.gpt2_preprocessor import GPT2Preprocessor
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
-)
-from keras_hub.src.utils.tensor_utils import strip_to_ragged
+from keras_hub.src.models.causal_lm_preprocessor import CausalLMPreprocessor
+from keras_hub.src.models.gpt2.gpt2_backbone import GPT2Backbone
+from keras_hub.src.models.gpt2.gpt2_tokenizer import GPT2Tokenizer
 
 
 @keras_hub_export("keras_hub.models.GPT2CausalLMPreprocessor")
-class GPT2CausalLMPreprocessor(GPT2Preprocessor):
+class GPT2CausalLMPreprocessor(CausalLMPreprocessor):
     """GPT2 Causal LM preprocessor.
 
     This preprocessing layer is meant for use with
@@ -91,83 +86,5 @@ class GPT2CausalLMPreprocessor(GPT2Preprocessor):
     ```
     """
 
-    def call(
-        self,
-        x,
-        y=None,
-        sample_weight=None,
-        sequence_length=None,
-    ):
-        if y is not None or sample_weight is not None:
-            logging.warning(
-                "`GPT2CausalLMPreprocessor` generates `y` and `sample_weight` "
-                "based on your input data, but your data already contains `y` "
-                "or `sample_weight`. Your `y` and `sample_weight` will be "
-                "ignored."
-            )
-        sequence_length = sequence_length or self.sequence_length
-
-        x = convert_inputs_to_list_of_tensor_segments(x)[0]
-        x = self.tokenizer(x)
-        # Pad with one extra token to account for the truncation below.
-        token_ids, padding_mask = self.packer(
-            x,
-            sequence_length=sequence_length + 1,
-            add_start_value=self.add_start_token,
-            add_end_value=self.add_end_token,
-        )
-        # The last token does not have a next token, so we truncate it out.
-        x = {
-            "token_ids": token_ids[..., :-1],
-            "padding_mask": padding_mask[..., :-1],
-        }
-        # Target `y` will be the next token.
-        y, sample_weight = token_ids[..., 1:], padding_mask[..., 1:]
-        return keras.utils.pack_x_y_sample_weight(x, y, sample_weight)
-
-    def generate_preprocess(
-        self,
-        x,
-        sequence_length=None,
-    ):
-        """Convert strings to integer token input for generation.
-
-        Similar to calling the layer for training, this method takes in strings
-        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 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).
-        """
-        if not self.built:
-            self.build(None)
-
-        x = convert_inputs_to_list_of_tensor_segments(x)[0]
-        x = self.tokenizer(x)
-        token_ids, padding_mask = self.packer(
-            x, sequence_length=sequence_length, add_end_value=False
-        )
-        return {
-            "token_ids": token_ids,
-            "padding_mask": padding_mask,
-        }
-
-    def generate_postprocess(
-        self,
-        x,
-    ):
-        """Convert 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 integer sequence
-        back to a string.
-        """
-        if not self.built:
-            self.build(None)
-
-        token_ids, padding_mask = x["token_ids"], x["padding_mask"]
-        ids_to_strip = (self.tokenizer.end_token_id,)
-        token_ids = strip_to_ragged(token_ids, padding_mask, ids_to_strip)
-        return self.tokenizer.detokenize(token_ids)
+    backbone_cls = GPT2Backbone
+    tokenizer_cls = GPT2Tokenizer

keras_hub/src/models/gpt2/gpt2_preprocessor.py

@@ -17,11 +17,10 @@ import keras
 
 from keras_hub.src.api_export import keras_hub_export
 from keras_hub.src.layers.preprocessing.start_end_packer import StartEndPacker
+from keras_hub.src.models.gpt2.gpt2_backbone import GPT2Backbone
 from keras_hub.src.models.gpt2.gpt2_tokenizer import GPT2Tokenizer
 from keras_hub.src.models.preprocessor import Preprocessor
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
-)
+from keras_hub.src.utils.tensor_utils import preprocessing_function
 
 
 @keras_hub_export("keras_hub.models.GPT2Preprocessor")
@@ -107,6 +106,7 @@ class GPT2Preprocessor(Preprocessor):
     ```
     """
 
+    backbone_cls = GPT2Backbone
     tokenizer_cls = GPT2Tokenizer
 
     def __init__(
@@ -136,6 +136,7 @@ class GPT2Preprocessor(Preprocessor):
         )
         self.built = True
 
+    @preprocessing_function
     def call(
         self,
         x,
@@ -143,17 +144,9 @@ class GPT2Preprocessor(Preprocessor):
         sample_weight=None,
         sequence_length=None,
     ):
-        x = convert_inputs_to_list_of_tensor_segments(x)
-        if len(x) != 1:
-            raise ValueError(
-                "GPT2 requires each input feature to contain only "
-                f"one segment, but received {len(x)}. If you are using GPT2 "
-                "for a multi-segment classification task, please refer to "
-                "classification models like BERT or RoBERTa."
-            )
         sequence_length = sequence_length or self.sequence_length
         token_ids, padding_mask = self.packer(
-            self.tokenizer(x[0]),
+            self.tokenizer(x),
             sequence_length=sequence_length,
             add_start_value=self.add_start_token,
             add_end_value=self.add_end_token,

keras_hub/src/models/gpt2/gpt2_tokenizer.py

@@ -14,10 +14,16 @@
 
 
 from keras_hub.src.api_export import keras_hub_export
+from keras_hub.src.models.gpt2.gpt2_backbone import GPT2Backbone
 from keras_hub.src.tokenizers.byte_pair_tokenizer import BytePairTokenizer
 
 
-@keras_hub_export("keras_hub.models.GPT2Tokenizer")
+@keras_hub_export(
+    [
+        "keras_hub.tokenizers.GPT2Tokenizer",
+        "keras_hub.models.GPT2Tokenizer",
+    ]
+)
 class GPT2Tokenizer(BytePairTokenizer):
     """A GPT-2 tokenizer using Byte-Pair Encoding subword segmentation.
 
@@ -27,8 +33,6 @@ class GPT2Tokenizer(BytePairTokenizer):
     models and provides a `from_preset()` method to automatically download
     a matching vocabulary for a GPT-2 preset.
 
-    This tokenizer does not provide truncation or padding of inputs.
-
     If input is a batch of strings (rank > 0), the layer will output a
     `tf.RaggedTensor` where the last dimension of the output is ragged.
 
@@ -65,6 +69,8 @@ class GPT2Tokenizer(BytePairTokenizer):
     ```
     """
 
+    backbone_cls = GPT2Backbone
+
     def __init__(
         self,
         vocabulary=None,
@@ -72,39 +78,11 @@ class GPT2Tokenizer(BytePairTokenizer):
         **kwargs,
     ):
         # GPT2 uses the same start as end token, i.e., "<|endoftext|>".
-        self.end_token = self.start_token = "<|endoftext|>"
-
+        self._add_special_token("<|endoftext|>", "end_token")
+        self._add_special_token("<|endoftext|>", "start_token")
+        self.pad_token_id = 0
         super().__init__(
             vocabulary=vocabulary,
             merges=merges,
-            unsplittable_tokens=[self.end_token],
             **kwargs,
         )
-
-    def set_vocabulary_and_merges(self, vocabulary, merges):
-        super().set_vocabulary_and_merges(vocabulary, merges)
-
-        if vocabulary is not None:
-            # Check for necessary special tokens.
-            if self.end_token not in self.get_vocabulary():
-                raise ValueError(
-                    f"Cannot find token `'{self.end_token}'` in the provided "
-                    f"`vocabulary`. Please provide `'{self.end_token}'` in "
-                    "your `vocabulary` or use a pretrained `vocabulary` name."
-                )
-
-            self.end_token_id = self.token_to_id(self.end_token)
-            self.start_token_id = self.end_token_id
-            self.pad_token_id = 0
-        else:
-            self.end_token_id = None
-            self.start_token_id = None
-            self.pad_token_id = None
-
-    def get_config(self):
-        config = super().get_config()
-        # In the constructor, we pass the list of special tokens to the
-        # `unsplittable_tokens` arg of the superclass' constructor. Hence, we
-        # delete it from the config here.
-        del config["unsplittable_tokens"]
-        return config

keras_hub/src/models/gpt_neo_x/gpt_neo_x_causal_lm_preprocessor.py

@@ -12,21 +12,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import keras
-from absl import logging
-
 from keras_hub.src.api_export import keras_hub_export
-from keras_hub.src.models.gpt_neo_x.gpt_neo_x_preprocessor import (
-    GPTNeoXPreprocessor,
-)
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
-)
-from keras_hub.src.utils.tensor_utils import strip_to_ragged
+from keras_hub.src.models.causal_lm_preprocessor import CausalLMPreprocessor
+from keras_hub.src.models.gpt_neo_x.gpt_neo_x_backbone import GPTNeoXBackbone
+from keras_hub.src.models.gpt_neo_x.gpt_neo_x_tokenizer import GPTNeoXTokenizer
 
 
 @keras_hub_export("keras_hub.models.GPTNeoXCausalLMPreprocessor")
-class GPTNeoXCausalLMPreprocessor(GPTNeoXPreprocessor):
+class GPTNeoXCausalLMPreprocessor(CausalLMPreprocessor):
     """GPT-NeoX Causal LM preprocessor.
 
     This preprocessing layer is meant for use with
@@ -59,83 +52,5 @@ class GPTNeoXCausalLMPreprocessor(GPTNeoXPreprocessor):
 
     """
 
-    def call(
-        self,
-        x,
-        y=None,
-        sample_weight=None,
-        sequence_length=None,
-    ):
-        if y is not None or sample_weight is not None:
-            logging.warning(
-                "`GPTNeoXCausalLMPreprocessor` generates `y` and `sample_weight` "
-                "based on your input data, but your data already contains `y` "
-                "or `sample_weight`. Your `y` and `sample_weight` will be "
-                "ignored."
-            )
-        sequence_length = sequence_length or self.sequence_length
-
-        x = convert_inputs_to_list_of_tensor_segments(x)[0]
-        x = self.tokenizer(x)
-        # Pad with one extra token to account for the truncation below.
-        token_ids, padding_mask = self.packer(
-            x,
-            sequence_length=sequence_length + 1,
-            add_start_value=self.add_start_token,
-            add_end_value=self.add_end_token,
-        )
-        # The last token does not have a next token, so we truncate it out.
-        x = {
-            "token_ids": token_ids[..., :-1],
-            "padding_mask": padding_mask[..., :-1],
-        }
-        # Target `y` will be the next token.
-        y, sample_weight = token_ids[..., 1:], padding_mask[..., 1:]
-        return keras.utils.pack_x_y_sample_weight(x, y, sample_weight)
-
-    def generate_preprocess(
-        self,
-        x,
-        sequence_length=None,
-    ):
-        """Convert strings to integer token input for generation.
-
-        Similar to calling the layer for training, this method takes in strings
-        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 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).
-        """
-        if not self.built:
-            self.build(None)
-
-        x = convert_inputs_to_list_of_tensor_segments(x)[0]
-        x = self.tokenizer(x)
-        token_ids, padding_mask = self.packer(
-            x, sequence_length=sequence_length, add_end_value=False
-        )
-        return {
-            "token_ids": token_ids,
-            "padding_mask": padding_mask,
-        }
-
-    def generate_postprocess(
-        self,
-        x,
-    ):
-        """Convert 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 integer sequence
-        back to a string.
-        """
-        if not self.built:
-            self.build(None)
-
-        token_ids, padding_mask = x["token_ids"], x["padding_mask"]
-        ids_to_strip = (self.tokenizer.end_token_id,)
-        token_ids = strip_to_ragged(token_ids, padding_mask, ids_to_strip)
-        return self.tokenizer.detokenize(token_ids)
+    backbone_cls = GPTNeoXBackbone
+    tokenizer_cls = GPTNeoXTokenizer

keras_hub/src/models/gpt_neo_x/gpt_neo_x_preprocessor.py

@@ -16,11 +16,10 @@ import keras
 
 from keras_hub.src.api_export import keras_hub_export
 from keras_hub.src.layers.preprocessing.start_end_packer import StartEndPacker
+from keras_hub.src.models.gpt_neo_x.gpt_neo_x_backbone import GPTNeoXBackbone
 from keras_hub.src.models.gpt_neo_x.gpt_neo_x_tokenizer import GPTNeoXTokenizer
 from keras_hub.src.models.preprocessor import Preprocessor
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
-)
+from keras_hub.src.utils.tensor_utils import preprocessing_function
 
 
 @keras_hub_export("keras_hub.models.GPTNeoXPreprocessor")
@@ -65,6 +64,7 @@ class GPTNeoXPreprocessor(Preprocessor):
             the layer.
     """
 
+    backbone_cls = GPTNeoXBackbone
     tokenizer_cls = GPTNeoXTokenizer
 
     def __init__(
@@ -94,6 +94,7 @@ class GPTNeoXPreprocessor(Preprocessor):
         )
         self.built = True
 
+    @preprocessing_function
     def call(
         self,
         x,
@@ -101,17 +102,9 @@ class GPTNeoXPreprocessor(Preprocessor):
         sample_weight=None,
         sequence_length=None,
     ):
-        x = convert_inputs_to_list_of_tensor_segments(x)
-        if len(x) != 1:
-            raise ValueError(
-                "GPTNeoX requires each input feature to contain only "
-                f"one segment, but received {len(x)}. If you are using GPTNeoX "
-                "for a multi-segment classification task, please refer to "
-                "classification models like BERT or RoBERTa."
-            )
         sequence_length = sequence_length or self.sequence_length
         token_ids, padding_mask = self.packer(
-            self.tokenizer(x[0]),
+            self.tokenizer(x),
             sequence_length=sequence_length,
             add_start_value=self.add_start_token,
             add_end_value=self.add_end_token,