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

keras_hub/src/models/resnet/resnet_presets.py

@@ -0,0 +1,95 @@
+# Copyright 2024 The KerasHub Authors
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     https://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""ResNet preset configurations."""
+
+backbone_presets = {
+    "resnet_18": {
+        "metadata": {
+            "description": (
+                "18-layer ResNet model pre-trained on the ImageNet 1k dataset "
+                "at a 224x224 resolution."
+            ),
+            "params": 11186112,
+            "official_name": "ResNet",
+            "path": "resnet",
+            "model_card": "https://arxiv.org/abs/2110.00476",
+        },
+        "kaggle_handle": "kaggle://kerashub/resnetv1/keras/resnet_18/1",
+    },
+    "resnet_50": {
+        "metadata": {
+            "description": (
+                "50-layer ResNet model pre-trained on the ImageNet 1k dataset "
+                "at a 224x224 resolution."
+            ),
+            "params": 23561152,
+            "official_name": "ResNet",
+            "path": "resnet",
+            "model_card": "https://arxiv.org/abs/2110.00476",
+        },
+        "kaggle_handle": "kaggle://kerashub/resnetv1/keras/resnet_50/1",
+    },
+    "resnet_101": {
+        "metadata": {
+            "description": (
+                "101-layer ResNet model pre-trained on the ImageNet 1k dataset "
+                "at a 224x224 resolution."
+            ),
+            "params": 42605504,
+            "official_name": "ResNet",
+            "path": "resnet",
+            "model_card": "https://arxiv.org/abs/2110.00476",
+        },
+        "kaggle_handle": "kaggle://kerashub/resnetv1/keras/resnet_101/1",
+    },
+    "resnet_152": {
+        "metadata": {
+            "description": (
+                "152-layer ResNet model pre-trained on the ImageNet 1k dataset "
+                "at a 224x224 resolution."
+            ),
+            "params": 58295232,
+            "official_name": "ResNet",
+            "path": "resnet",
+            "model_card": "https://arxiv.org/abs/2110.00476",
+        },
+        "kaggle_handle": "kaggle://kerashub/resnetv1/keras/resnet_152/1",
+    },
+    "resnet_v2_50": {
+        "metadata": {
+            "description": (
+                "50-layer ResNetV2 model pre-trained on the ImageNet 1k "
+                "dataset at a 224x224 resolution."
+            ),
+            "params": 23561152,
+            "official_name": "ResNet",
+            "path": "resnet",
+            "model_card": "https://arxiv.org/abs/2110.00476",
+        },
+        "kaggle_handle": "kaggle://kerashub/resnetv2/keras/resnet_v2_50/1",
+    },
+    "resnet_v2_101": {
+        "metadata": {
+            "description": (
+                "101-layer ResNetV2 model pre-trained on the ImageNet 1k "
+                "dataset at a 224x224 resolution."
+            ),
+            "params": 42605504,
+            "official_name": "ResNet",
+            "path": "resnet",
+            "model_card": "https://arxiv.org/abs/2110.00476",
+        },
+        "kaggle_handle": "kaggle://kerashub/resnetv2/keras/resnet_v2_101/1",
+    },
+}

keras_hub/src/models/roberta/__init__.py

@@ -14,7 +14,6 @@
 
 from keras_hub.src.models.roberta.roberta_backbone import RobertaBackbone
 from keras_hub.src.models.roberta.roberta_presets import backbone_presets
-from keras_hub.src.models.roberta.roberta_tokenizer import RobertaTokenizer
 from keras_hub.src.utils.preset_utils import register_presets
 
-register_presets(backbone_presets, (RobertaBackbone, RobertaTokenizer))
+register_presets(backbone_presets, RobertaBackbone)

keras_hub/src/models/roberta/roberta_masked_lm_preprocessor.py

@@ -13,19 +13,19 @@
 # limitations under the License.
 
 import keras
-from absl import logging
 
 from keras_hub.src.api_export import keras_hub_export
-from keras_hub.src.layers.preprocessing.masked_lm_mask_generator import (
-    MaskedLMMaskGenerator,
-)
-from keras_hub.src.models.roberta.roberta_preprocessor import (
-    RobertaPreprocessor,
+from keras_hub.src.layers.preprocessing.multi_segment_packer import (
+    MultiSegmentPacker,
 )
+from keras_hub.src.models.masked_lm_preprocessor import MaskedLMPreprocessor
+from keras_hub.src.models.roberta.roberta_backbone import RobertaBackbone
+from keras_hub.src.models.roberta.roberta_tokenizer import RobertaTokenizer
+from keras_hub.src.utils.tensor_utils import preprocessing_function
 
 
 @keras_hub_export("keras_hub.models.RobertaMaskedLMPreprocessor")
-class RobertaMaskedLMPreprocessor(RobertaPreprocessor):
+class RobertaMaskedLMPreprocessor(MaskedLMPreprocessor):
     """RoBERTa preprocessing for the masked language modeling task.
 
     This preprocessing layer will prepare inputs for a masked language modeling
@@ -122,77 +122,25 @@ class RobertaMaskedLMPreprocessor(RobertaPreprocessor):
     ```
     """
 
-    def __init__(
-        self,
-        tokenizer,
-        sequence_length=512,
-        truncate="round_robin",
-        mask_selection_rate=0.15,
-        mask_selection_length=96,
-        mask_token_rate=0.8,
-        random_token_rate=0.1,
-        **kwargs,
-    ):
-        super().__init__(
-            tokenizer,
-            sequence_length=sequence_length,
-            truncate=truncate,
-            **kwargs,
-        )
-        self.mask_selection_rate = mask_selection_rate
-        self.mask_selection_length = mask_selection_length
-        self.mask_token_rate = mask_token_rate
-        self.random_token_rate = random_token_rate
-        self.masker = None
+    backbone_cls = RobertaBackbone
+    tokenizer_cls = RobertaTokenizer
 
     def build(self, input_shape):
         super().build(input_shape)
-        # Defer packer creation to `build()` so that we can be sure tokenizer
-        # assets have loaded when restoring a saved model.
-        self.masker = MaskedLMMaskGenerator(
-            mask_selection_rate=self.mask_selection_rate,
-            mask_selection_length=self.mask_selection_length,
-            mask_token_rate=self.mask_token_rate,
-            random_token_rate=self.random_token_rate,
-            vocabulary_size=self.tokenizer.vocabulary_size(),
-            mask_token_id=self.tokenizer.mask_token_id,
-            unselectable_token_ids=[
-                self.tokenizer.start_token_id,
-                self.tokenizer.end_token_id,
-                self.tokenizer.pad_token_id,
-            ],
+        # Roberta is doubles up the sep token, so we override build.
+        self.packer = MultiSegmentPacker(
+            start_value=self.tokenizer.start_token_id,
+            end_value=self.tokenizer.end_token_id,
+            sep_value=[self.tokenizer.end_token_id] * 2,
+            pad_value=self.tokenizer.pad_token_id,
+            truncate=self.truncate,
+            sequence_length=self.sequence_length,
         )
-        self.built = True
 
+    @preprocessing_function
     def call(self, x, y=None, sample_weight=None):
-        if y is not None or sample_weight is not None:
-            logging.warning(
-                f"{self.__class__.__name__} 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."
-            )
-
-        x = super().call(x)
-        token_ids, padding_mask = x["token_ids"], x["padding_mask"]
-        masker_outputs = self.masker(token_ids)
-        x = {
-            "token_ids": masker_outputs["token_ids"],
-            "padding_mask": padding_mask,
-            "mask_positions": masker_outputs["mask_positions"],
-        }
-        y = masker_outputs["mask_ids"]
-        sample_weight = masker_outputs["mask_weights"]
+        output = super().call(x, y=y, sample_weight=sample_weight)
+        x, y, sample_weight = keras.utils.unpack_x_y_sample_weight(output)
+        # Backbone has no segment ID input.
+        del x["segment_ids"]
         return keras.utils.pack_x_y_sample_weight(x, y, sample_weight)
-
-    def get_config(self):
-        config = super().get_config()
-        config.update(
-            {
-                "mask_selection_rate": self.mask_selection_rate,
-                "mask_selection_length": self.mask_selection_length,
-                "mask_token_rate": self.mask_token_rate,
-                "random_token_rate": self.random_token_rate,
-            }
-        )
-        return config

keras_hub/src/models/roberta/{roberta_classifier.py → roberta_text_classifier.py}

@@ -16,18 +16,18 @@
 import keras
 
 from keras_hub.src.api_export import keras_hub_export
-from keras_hub.src.models.classifier import Classifier
 from keras_hub.src.models.roberta.roberta_backbone import RobertaBackbone
 from keras_hub.src.models.roberta.roberta_backbone import (
     roberta_kernel_initializer,
 )
-from keras_hub.src.models.roberta.roberta_preprocessor import (
-    RobertaPreprocessor,
+from keras_hub.src.models.roberta.roberta_text_classifier_preprocessor import (
+    RobertaTextClassifierPreprocessor,
 )
+from keras_hub.src.models.text_classifier import TextClassifier
 
 
-@keras_hub_export("keras_hub.models.RobertaClassifier")
-class RobertaClassifier(Classifier):
+@keras_hub_export("keras_hub.models.RobertaTextClassifier")
+class RobertaTextClassifier(TextClassifier):
     """An end-to-end RoBERTa model for classification tasks.
 
     This model attaches a classification head to a
@@ -48,7 +48,7 @@ class RobertaClassifier(Classifier):
     Args:
         backbone: A `keras_hub.models.RobertaBackbone` instance.
         num_classes: int. Number of classes to predict.
-        preprocessor: A `keras_hub.models.RobertaPreprocessor` or `None`. If
+        preprocessor: A `keras_hub.models.RobertaTextClassifierPreprocessor` or `None`. If
             `None`, this model will not apply preprocessing, and inputs should
             be preprocessed before calling the model.
         activation: Optional `str` or callable. The activation function to use
@@ -66,7 +66,7 @@ class RobertaClassifier(Classifier):
     labels = [0, 3]
 
     # Pretrained classifier.
-    classifier = keras_hub.models.RobertaClassifier.from_preset(
+    classifier = keras_hub.models.RobertaTextClassifier.from_preset(
         "roberta_base_en",
         num_classes=4,
     )
@@ -94,7 +94,7 @@ class RobertaClassifier(Classifier):
     labels = [0, 3]
 
     # Pretrained classifier without preprocessing.
-    classifier = keras_hub.models.RobertaClassifier.from_preset(
+    classifier = keras_hub.models.RobertaTextClassifier.from_preset(
         "roberta_base_en",
         num_classes=4,
         preprocessor=None,
@@ -115,7 +115,7 @@ class RobertaClassifier(Classifier):
         vocabulary=vocab,
         merges=merges
     )
-    preprocessor = keras_hub.models.RobertaPreprocessor(
+    preprocessor = keras_hub.models.RobertaTextClassifierPreprocessor(
         tokenizer=tokenizer,
         sequence_length=128,
     )
@@ -127,7 +127,7 @@ class RobertaClassifier(Classifier):
         intermediate_dim=512,
         max_sequence_length=128
     )
-    classifier = keras_hub.models.RobertaClassifier(
+    classifier = keras_hub.models.RobertaTextClassifier(
         backbone=backbone,
         preprocessor=preprocessor,
         num_classes=4,
@@ -137,7 +137,7 @@ class RobertaClassifier(Classifier):
     """
 
     backbone_cls = RobertaBackbone
-    preprocessor_cls = RobertaPreprocessor
+    preprocessor_cls = RobertaTextClassifierPreprocessor
 
     def __init__(
         self,

keras_hub/src/models/roberta/{roberta_preprocessor.py → roberta_text_classifier_preprocessor.py}

@@ -12,22 +12,27 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-
 import keras
 
 from keras_hub.src.api_export import keras_hub_export
 from keras_hub.src.layers.preprocessing.multi_segment_packer import (
     MultiSegmentPacker,
 )
-from keras_hub.src.models.preprocessor import Preprocessor
+from keras_hub.src.models.roberta.roberta_backbone import RobertaBackbone
 from keras_hub.src.models.roberta.roberta_tokenizer import RobertaTokenizer
-from keras_hub.src.utils.keras_utils import (
-    convert_inputs_to_list_of_tensor_segments,
+from keras_hub.src.models.text_classifier_preprocessor import (
+    TextClassifierPreprocessor,
 )
+from keras_hub.src.utils.tensor_utils import preprocessing_function
 
 
-@keras_hub_export("keras_hub.models.RobertaPreprocessor")
-class RobertaPreprocessor(Preprocessor):
+@keras_hub_export(
+    [
+        "keras_hub.models.RobertaTextClassifierPreprocessor",
+        "keras_hub.models.RobertaPreprocessor",
+    ]
+)
+class RobertaTextClassifierPreprocessor(TextClassifierPreprocessor):
     """A RoBERTa preprocessing layer which tokenizes and packs inputs.
 
     This preprocessing layer will do three things:
@@ -71,7 +76,7 @@ class RobertaPreprocessor(Preprocessor):
 
     Directly calling the layer on data.
     ```python
-    preprocessor = keras_hub.models.RobertaPreprocessor.from_preset(
+    preprocessor = keras_hub.models.TextClassifierPreprocessor.from_preset(
         "roberta_base_en"
     )
 
@@ -95,12 +100,12 @@ class RobertaPreprocessor(Preprocessor):
         vocabulary=vocab,
         merges=merges
     )
-    preprocessor = keras_hub.models.RobertaPreprocessor(tokenizer)
+    preprocessor = keras_hub.models.RobertaTextClassifierPreprocessor(tokenizer)
     preprocessor("a quick fox")
     ```
     Mapping with `tf.data.Dataset`.
     ```python
-    preprocessor = keras_hub.models.RobertaPreprocessor.from_preset(
+    preprocessor = keras_hub.models.TextClassifierPreprocessor.from_preset(
         "roberta_base_en"
     )
 
@@ -131,25 +136,11 @@ class RobertaPreprocessor(Preprocessor):
     ```
     """
 
+    backbone_cls = RobertaBackbone
     tokenizer_cls = RobertaTokenizer
 
-    def __init__(
-        self,
-        tokenizer,
-        sequence_length=512,
-        truncate="round_robin",
-        **kwargs,
-    ):
-        super().__init__(**kwargs)
-
-        self.tokenizer = tokenizer
-        self.packer = None
-        self.truncate = truncate
-        self.sequence_length = sequence_length
-
     def build(self, input_shape):
-        # Defer packer creation to `build()` so that we can be sure tokenizer
-        # assets have loaded when restoring a saved model.
+        # Roberta is doubles up the sep token, so we override build.
         self.packer = MultiSegmentPacker(
             start_value=self.tokenizer.start_token_id,
             end_value=self.tokenizer.end_token_id,
@@ -160,33 +151,10 @@ class RobertaPreprocessor(Preprocessor):
         )
         self.built = True
 
+    @preprocessing_function
     def call(self, x, y=None, sample_weight=None):
-        x = convert_inputs_to_list_of_tensor_segments(x)
-        x = [self.tokenizer(segment) for segment in x]
-        token_ids, _ = self.packer(x)
-        x = {
-            "token_ids": token_ids,
-            "padding_mask": token_ids != self.tokenizer.pad_token_id,
-        }
+        output = super().call(x, y=y, sample_weight=sample_weight)
+        x, y, sample_weight = keras.utils.unpack_x_y_sample_weight(output)
+        # Backbone has no segment ID input.
+        del x["segment_ids"]
         return keras.utils.pack_x_y_sample_weight(x, y, sample_weight)
-
-    def get_config(self):
-        config = super().get_config()
-        config.update(
-            {
-                "sequence_length": self.sequence_length,
-                "truncate": self.truncate,
-            }
-        )
-        return config
-
-    @property
-    def sequence_length(self):
-        """The padded length of model input sequences."""
-        return self._sequence_length
-
-    @sequence_length.setter
-    def sequence_length(self, value):
-        self._sequence_length = value
-        if self.packer is not None:
-            self.packer.sequence_length = value

keras_hub/src/models/roberta/roberta_tokenizer.py

@@ -14,10 +14,16 @@
 
 
 from keras_hub.src.api_export import keras_hub_export
+from keras_hub.src.models.roberta.roberta_backbone import RobertaBackbone
 from keras_hub.src.tokenizers.byte_pair_tokenizer import BytePairTokenizer
 
 
-@keras_hub_export("keras_hub.models.RobertaTokenizer")
+@keras_hub_export(
+    [
+        "keras_hub.tokenizers.RobertaTokenizer",
+        "keras_hub.models.RobertaTokenizer",
+    ]
+)
 class RobertaTokenizer(BytePairTokenizer):
     """A RoBERTa tokenizer using Byte-Pair Encoding subword segmentation.
 
@@ -27,10 +33,6 @@ class RobertaTokenizer(BytePairTokenizer):
     models and provides a `from_preset()` method to automatically download
     a matching vocabulary for a RoBERTa preset.
 
-    This tokenizer does not provide truncation or padding of inputs. It can be
-    combined with a `keras_hub.models.RobertaPreprocessor` layer for input
-    packing.
-
     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.
 
@@ -72,61 +74,20 @@ class RobertaTokenizer(BytePairTokenizer):
     ```
     """
 
+    backbone_cls = RobertaBackbone
+
     def __init__(
         self,
         vocabulary=None,
         merges=None,
         **kwargs,
     ):
-        self.start_token = "<s>"
-        self.pad_token = "<pad>"
-        self.end_token = "</s>"
-        self.mask_token = "<mask>"
-
+        self._add_special_token("<s>", "start_token")
+        self._add_special_token("</s>", "end_token")
+        self._add_special_token("<pad>", "pad_token")
+        self._add_special_token("<mask>", "mask_token")
         super().__init__(
             vocabulary=vocabulary,
             merges=merges,
-            unsplittable_tokens=[
-                self.start_token,
-                self.pad_token,
-                self.end_token,
-                self.mask_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.
-            for token in [
-                self.start_token,
-                self.pad_token,
-                self.end_token,
-                self.mask_token,
-            ]:
-                if token not in self.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.pad_token_id = self.token_to_id(self.pad_token)
-            self.end_token_id = self.token_to_id(self.end_token)
-            self.mask_token_id = self.token_to_id(self.mask_token)
-        else:
-            self.start_token_id = None
-            self.pad_token_id = None
-            self.end_token_id = None
-            self.mask_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