keras-hub-nightly 0.19.0.dev202412120352__py3-none-any.whl → 0.19.0.dev202412140350__py3-none-any.whl

keras_hub/src/models/densenet/densenet_backbone.py

@@ -29,7 +29,9 @@ class DenseNetBackbone(FeaturePyramidBackbone):
     input_data = np.ones(shape=(8, 224, 224, 3))
 
     # Pretrained backbone
-    model = keras_hub.models.DenseNetBackbone.from_preset("densenet_121_imagenet")
+    model = keras_hub.models.DenseNetBackbone.from_preset(
+        "densenet_121_imagenet"
+    )
     model(input_data)
 
     # Randomly initialized backbone with a custom config

keras_hub/src/models/densenet/densenet_image_classifier.py

@@ -1,6 +1,6 @@
 from keras_hub.src.api_export import keras_hub_export
 from keras_hub.src.models.densenet.densenet_backbone import DenseNetBackbone
-from keras_hub.src.models.densenet.densenet_image_classifier_preprocessor import (
+from keras_hub.src.models.densenet.densenet_image_classifier_preprocessor import (  # noqa: E501
     DenseNetImageClassifierPreprocessor,
 )
 from keras_hub.src.models.image_classifier import ImageClassifier

keras_hub/src/models/densenet/densenet_presets.py

@@ -4,8 +4,8 @@ backbone_presets = {
     "densenet_121_imagenet": {
         "metadata": {
             "description": (
-                "121-layer DenseNet model pre-trained on the ImageNet 1k dataset "
-                "at a 224x224 resolution."
+                "121-layer DenseNet model pre-trained on the ImageNet 1k "
+                "dataset at a 224x224 resolution."
             ),
             "params": 7037504,
             "path": "densenet",
@@ -15,8 +15,8 @@ backbone_presets = {
     "densenet_169_imagenet": {
         "metadata": {
             "description": (
-                "169-layer DenseNet model pre-trained on the ImageNet 1k dataset "
-                "at a 224x224 resolution."
+                "169-layer DenseNet model pre-trained on the ImageNet 1k "
+                "dataset at a 224x224 resolution."
             ),
             "params": 12642880,
             "path": "densenet",
@@ -26,8 +26,8 @@ backbone_presets = {
     "densenet_201_imagenet": {
         "metadata": {
             "description": (
-                "201-layer DenseNet model pre-trained on the ImageNet 1k dataset "
-                "at a 224x224 resolution."
+                "201-layer DenseNet model pre-trained on the ImageNet 1k "
+                "dataset at a 224x224 resolution."
             ),
             "params": 18321984,
             "path": "densenet",

keras_hub/src/models/distil_bert/distil_bert_masked_lm.py

@@ -8,7 +8,7 @@ from keras_hub.src.models.distil_bert.distil_bert_backbone import (
 from keras_hub.src.models.distil_bert.distil_bert_backbone import (
     distilbert_kernel_initializer,
 )
-from keras_hub.src.models.distil_bert.distil_bert_masked_lm_preprocessor import (
+from keras_hub.src.models.distil_bert.distil_bert_masked_lm_preprocessor import (  # noqa: E501
     DistilBertMaskedLMPreprocessor,
 )
 from keras_hub.src.models.masked_lm import MaskedLM

keras_hub/src/models/distil_bert/distil_bert_masked_lm_preprocessor.py

@@ -17,8 +17,8 @@ class DistilBertMaskedLMPreprocessor(MaskedLMPreprocessor):
 
     This preprocessing layer will prepare inputs for a masked language modeling
     task. It is primarily intended for use with the
-    `keras_hub.models.DistilBertMaskedLM` task model. Preprocessing will occur in
-    multiple steps.
+    `keras_hub.models.DistilBertMaskedLM` task model. Preprocessing will occur
+    in multiple steps.
 
     1. Tokenize any number of input segments using the `tokenizer`.
     2. Pack the inputs together using a `keras_hub.layers.MultiSegmentPacker`.

keras_hub/src/models/distil_bert/distil_bert_presets.py

@@ -28,7 +28,8 @@ backbone_presets = {
     "distil_bert_base_multi": {
         "metadata": {
             "description": (
-                "6-layer DistilBERT model where case is maintained. Trained on Wikipedias of 104 languages"
+                "6-layer DistilBERT model where case is maintained. Trained on "
+                "Wikipedias of 104 languages"
             ),
             "params": 134734080,
             "path": "distil_bert",

keras_hub/src/models/distil_bert/distil_bert_text_classifier.py

@@ -7,7 +7,7 @@ from keras_hub.src.models.distil_bert.distil_bert_backbone import (
 from keras_hub.src.models.distil_bert.distil_bert_backbone import (
     distilbert_kernel_initializer,
 )
-from keras_hub.src.models.distil_bert.distil_bert_text_classifier_preprocessor import (
+from keras_hub.src.models.distil_bert.distil_bert_text_classifier_preprocessor import (  # noqa: E501
     DistilBertTextClassifierPreprocessor,
 )
 from keras_hub.src.models.text_classifier import TextClassifier
@@ -40,9 +40,9 @@ class DistilBertTextClassifier(TextClassifier):
     Args:
         backbone: A `keras_hub.models.DistilBert` instance.
         num_classes: int. Number of classes to predict.
-        preprocessor: A `keras_hub.models.DistilBertTextClassifierPreprocessor` or `None`. If
-            `None`, this model will not apply preprocessing, and inputs should
-            be preprocessed before calling the model.
+        preprocessor: A `keras_hub.models.DistilBertTextClassifierPreprocessor`
+            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 on the model outputs. Set
             `activation="softmax"` to return output probabilities.
@@ -128,7 +128,7 @@ class DistilBertTextClassifier(TextClassifier):
     )
     classifier.fit(x=features, y=labels, batch_size=2)
     ```
-    """
+    """  # noqa: E501
 
     backbone_cls = DistilBertBackbone
     preprocessor_cls = DistilBertTextClassifierPreprocessor

keras_hub/src/models/distil_bert/distil_bert_tokenizer.py

@@ -16,9 +16,9 @@ class DistilBertTokenizer(WordPieceTokenizer):
 
     This tokenizer class will tokenize raw strings into integer sequences and
     is based on `keras_hub.tokenizers.WordPieceTokenizer`. Unlike the
-    underlying tokenizer, it will check for all special tokens needed by DistilBERT
-    models and provides a `from_preset()` method to automatically download
-    a matching vocabulary for a DistilBERT preset.
+    underlying tokenizer, it will check for all special tokens needed by
+    DistilBERT models and provides a `from_preset()` method to automatically
+    download a matching vocabulary for a DistilBERT preset.
 
     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.

keras_hub/src/models/efficientnet/cba.py

@@ -43,7 +43,7 @@ class CBABlock(keras.layers.Layer):
         activation="swish",
         dropout=0.2,
         nores=False,
-        **kwargs
+        **kwargs,
     ):
         super().__init__(**kwargs)
         self.input_filters = input_filters

keras_hub/src/models/efficientnet/efficientnet_backbone.py

@@ -361,8 +361,12 @@ class EfficientNetBackbone(FeaturePyramidBackbone):
         config = super().get_config()
         config.update(
             {
-                "stackwise_width_coefficients": self.stackwise_width_coefficients,
-                "stackwise_depth_coefficients": self.stackwise_depth_coefficients,
+                "stackwise_width_coefficients": (
+                    self.stackwise_width_coefficients
+                ),
+                "stackwise_depth_coefficients": (
+                    self.stackwise_depth_coefficients
+                ),
                 "dropout": self.dropout,
                 "depth_divisor": self.depth_divisor,
                 "min_depth": self.min_depth,
@@ -373,12 +377,18 @@ class EfficientNetBackbone(FeaturePyramidBackbone):
                 "stackwise_input_filters": self.stackwise_input_filters,
                 "stackwise_output_filters": self.stackwise_output_filters,
                 "stackwise_expansion_ratios": self.stackwise_expansion_ratios,
-                "stackwise_squeeze_and_excite_ratios": self.stackwise_squeeze_and_excite_ratios,
+                "stackwise_squeeze_and_excite_ratios": (
+                    self.stackwise_squeeze_and_excite_ratios
+                ),
                 "stackwise_strides": self.stackwise_strides,
                 "stackwise_block_types": self.stackwise_block_types,
-                "stackwise_force_input_filters": self.stackwise_force_input_filters,
+                "stackwise_force_input_filters": (
+                    self.stackwise_force_input_filters
+                ),
                 "include_stem_padding": self.include_stem_padding,
-                "use_depth_divisor_as_min_depth": self.use_depth_divisor_as_min_depth,
+                "use_depth_divisor_as_min_depth": (
+                    self.use_depth_divisor_as_min_depth
+                ),
                 "cap_round_filter_decrease": self.cap_round_filter_decrease,
                 "stem_conv_padding": self.stem_conv_padding,
                 "batch_norm_momentum": self.batch_norm_momentum,
@@ -389,7 +399,7 @@ class EfficientNetBackbone(FeaturePyramidBackbone):
         return config
 
     def _correct_pad_downsample(self, inputs, kernel_size):
-        """Returns a tuple for zero-padding for 2D convolution with downsampling.
+        """Returns a tuple for zero-padding a 2D convolution with downsampling.
 
         Args:
             inputs: Input tensor.
@@ -436,9 +446,11 @@ class EfficientNetBackbone(FeaturePyramidBackbone):
             filters_out: integer, the number of output filters.
             kernel_size: integer, the dimension of the convolution window.
             strides: integer, the stride of the convolution.
-            activation: activation function to use between each convolutional layer.
+            activation: activation function to use between each convolutional
+                layer.
             expand_ratio: integer, scaling coefficient for the input filters.
-            se_ratio: float between 0 and 1, fraction to squeeze the input filters.
+            se_ratio: float between 0 and 1, fraction to squeeze the input
+                filters.
             dropout: float between 0 and 1, fraction of the input units to drop.
             name: string, block label.
 

keras_hub/src/models/efficientnet/efficientnet_image_classifier.py

@@ -2,7 +2,7 @@ from keras_hub.src.api_export import keras_hub_export
 from keras_hub.src.models.efficientnet.efficientnet_backbone import (
     EfficientNetBackbone,
 )
-from keras_hub.src.models.efficientnet.efficientnet_image_classifier_preprocessor import (
+from keras_hub.src.models.efficientnet.efficientnet_image_classifier_preprocessor import (  # noqa: E501
     EfficientNetImageClassifierPreprocessor,
 )
 from keras_hub.src.models.image_classifier import ImageClassifier

keras_hub/src/models/efficientnet/efficientnet_presets.py

@@ -15,10 +15,10 @@ backbone_presets = {
     "efficientnet_b0_ra4_e3600_r224_imagenet": {
         "metadata": {
             "description": (
-                "EfficientNet B0 model pre-trained on the ImageNet 1k dataset by"
-                " Ross Wightman. Trained with timm scripts using hyper-parameters"
-                " inspired by the MobileNet-V4 small, mixed with go-to hparams "
-                'from timm and "ResNet Strikes Back".'
+                "EfficientNet B0 model pre-trained on the ImageNet 1k dataset "
+                "by Ross Wightman. Trained with timm scripts using "
+                "hyper-parameters inspired by the MobileNet-V4 small, mixed "
+                "with go-to hparams from timm and 'ResNet Strikes Back'."
             ),
             "params": 5288548,
             "path": "efficientnet",
@@ -38,10 +38,10 @@ backbone_presets = {
     "efficientnet_b1_ra4_e3600_r240_imagenet": {
         "metadata": {
             "description": (
-                "EfficientNet B1 model pre-trained on the ImageNet 1k dataset by"
-                " Ross Wightman. Trained with timm scripts using hyper-parameters"
-                " inspired by the MobileNet-V4 small, mixed with go-to hparams "
-                'from timm and "ResNet Strikes Back".'
+                "EfficientNet B1 model pre-trained on the ImageNet 1k dataset "
+                "by Ross Wightman. Trained with timm scripts using "
+                "hyper-parameters inspired by the MobileNet-V4 small, mixed "
+                "with go-to hparams from timm and 'ResNet Strikes Back'."
             ),
             "params": 7794184,
             "path": "efficientnet",
@@ -86,7 +86,8 @@ backbone_presets = {
             "description": (
                 "EfficientNet B5 model pre-trained on the ImageNet 12k dataset "
                 "by Ross Wightman. Based on Swin Transformer train / pretrain "
-                "recipe with modifications (related to both DeiT and ConvNeXt recipes)."
+                "recipe with modifications (related to both DeiT and ConvNeXt "
+                "recipes)."
             ),
             "params": 30389784,
             "path": "efficientnet",
@@ -181,8 +182,8 @@ backbone_presets = {
     "efficientnet_lite0_ra_imagenet": {
         "metadata": {
             "description": (
-                "EfficientNet-Lite model fine-trained on the ImageNet 1k dataset "
-                "with RandAugment recipe."
+                "EfficientNet-Lite model fine-trained on the ImageNet 1k "
+                "dataset with RandAugment recipe."
             ),
             "params": 4652008,
             "path": "efficientnet",

keras_hub/src/models/efficientnet/fusedmbconv.py

@@ -7,10 +7,8 @@ class FusedMBConvBlock(keras.layers.Layer):
     """Implementation of the FusedMBConv block
 
     Also known as a Fused Mobile Inverted Residual Bottleneck block from:
-        [EfficientNet-EdgeTPU: Creating Accelerator-Optimized Neural Networks with AutoML]
-        (https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html)
-        [EfficientNetV2: Smaller Models and Faster Training]
-        (https://arxiv.org/abs/2104.00298v3).
+    [EfficientNet-EdgeTPU](https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html)
+    [EfficientNetV2: Smaller Models and Faster Training](https://arxiv.org/abs/2104.00298v3).
 
     FusedMBConv blocks are based on MBConv blocks, and replace the depthwise and
     1x1 output convolution blocks with a single 3x3 convolution block, fusing
@@ -78,7 +76,7 @@ class FusedMBConvBlock(keras.layers.Layer):
         dropout=0.2,
         nores=False,
         projection_kernel_size=1,
-        **kwargs
+        **kwargs,
     ):
         super().__init__(**kwargs)
         self.input_filters = input_filters

keras_hub/src/models/efficientnet/mbconv.py

@@ -18,7 +18,7 @@ class MBConvBlock(keras.layers.Layer):
         activation="swish",
         dropout=0.2,
         nores=False,
-        **kwargs
+        **kwargs,
     ):
         """Implementation of the MBConv block
 

keras_hub/src/models/electra/electra_backbone.py

@@ -186,8 +186,8 @@ class ElectraBackbone(Backbone):
         # Index of classification token in the vocabulary
         cls_token_index = 0
         sequence_output = x
-        # Construct the two ELECTRA outputs. The pooled output is a dense layer on
-        # top of the [CLS] token.
+        # Construct the two ELECTRA outputs. The pooled output is a dense layer
+        # on top of the [CLS] token.
         pooled_output = self.pooled_dense(x[:, cls_token_index, :])
         super().__init__(
             inputs={

keras_hub/src/models/f_net/f_net_text_classifier.py

@@ -34,9 +34,9 @@ class FNetTextClassifier(TextClassifier):
     Args:
         backbone: A `keras_hub.models.FNetBackbone` instance.
         num_classes: int. Number of classes to predict.
-        preprocessor: A `keras_hub.models.FNetTextClassifierPreprocessor` or `None`. If
-            `None`, this model will not apply preprocessing, and inputs should
-            be preprocessed before calling the model.
+        preprocessor: A `keras_hub.models.FNetTextClassifierPreprocessor` 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 on the model outputs. Set
             `activation="softmax"` to return output probabilities.

keras_hub/src/models/f_net/f_net_text_classifier_preprocessor.py

@@ -22,9 +22,9 @@ class FNetTextClassifierPreprocessor(TextClassifierPreprocessor):
 
      1. Tokenize any number of input segments using the `tokenizer`.
      2. Pack the inputs together using a `keras_hub.layers.MultiSegmentPacker`.
-       with the appropriate `"[CLS]"`, `"[SEP]"` and `"<pad>"` tokens.
-     3. Construct a dictionary with keys `"token_ids"`, and `"segment_ids"`  that
-       can be passed directly to `keras_hub.models.FNetBackbone`.
+        with the appropriate `"[CLS]"`, `"[SEP]"` and `"<pad>"` tokens.
+     3. Construct a dictionary with keys `"token_ids"`, and `"segment_ids"`
+        that can be passed directly to `keras_hub.models.FNetBackbone`.
 
     This layer can be used directly with `tf.data.Dataset.map` to preprocess
     string data in the `(x, y, sample_weight)` format used by

keras_hub/src/models/falcon/falcon_backbone.py

@@ -20,15 +20,17 @@ class FalconBackbone(Backbone):
     Args:
         vocabulary_size: int. The size of the token vocabulary.
         num_layers: int. The number of transformer layers.
-        num_attention_heads: int. The number of attention heads for each transformer.
-            The hidden size must be divisible by the number of attention heads.
+        num_attention_heads: int. The number of attention heads for each
+            transformer. The hidden size must be divisible by the number of
+            attention heads.
         hidden_dim: int. The dimensionality of the embeddings and hidden states.
         intermediate_dim: int. The output dimension of the first Dense layer in
             the MLP network of each transformer.
         layer_norm_epsilon: float. Epsilon for the layer normalization layers in
             the transformer decoder.
         attention_dropout_rate: float. Dropout probability for the attention.
-        feedforward_dropout_rate: flaot. Dropout probability for the feedforward.
+        feedforward_dropout_rate: flaot. Dropout probability for the
+            feedforward.
         dtype: string or `keras.mixed_precision.DTypePolicy`. The dtype to use
             for model computations and weights. Note that some computations,
             such as softmax and layer normalization, will always be done at

keras_hub/src/models/falcon/falcon_causal_lm.py

@@ -40,7 +40,9 @@ class FalconCausalLM(CausalLM):
 
     Use `generate()` to do text generation.
     ```python
-    falcon_lm = keras_hub.models.FalconCausalLM.from_preset("falcon_refinedweb_1b_en")
+    falcon_lm = keras_hub.models.FalconCausalLM.from_preset(
+        "falcon_refinedweb_1b_en"
+    )
     falcon_lm.generate("I want to say", max_length=30)
 
     # Generate with batched prompts.
@@ -49,7 +51,9 @@ class FalconCausalLM(CausalLM):
 
     Compile the `generate()` function with a custom sampler.
     ```python
-    falcon_lm = keras_hub.models.FalconCausalLM.from_preset("falcon_refinedweb_1b_en")
+    falcon_lm = keras_hub.models.FalconCausalLM.from_preset(
+        "falcon_refinedweb_1b_en"
+    )
     falcon_lm.compile(sampler="top_k")
     falcon_lm.generate("I want to say", max_length=30)
 
@@ -60,7 +64,8 @@ class FalconCausalLM(CausalLM):
     Use `generate()` without preprocessing.
     ```python
     prompt = {
-        # Token ids for "<|endoftext|> Keras is".
+        # Token ids for
+        # "<|endoftext|> Keras is".
         "token_ids": np.array([[50256, 17337,   292,   318]] * 2),
         # Use `"padding_mask"` to indicate values that should not be overridden.
         "padding_mask": np.array([[1, 1, 1, 1]] * 2),
@@ -76,15 +81,20 @@ class FalconCausalLM(CausalLM):
     Call `fit()` on a single batch.
     ```python
     features = ["The quick brown fox jumped.", "I forgot my homework."]
-    falcon_lm = keras_hub.models.FalconCausalLM.from_preset("falcon_refinedweb_1b_en")
+    falcon_lm = keras_hub.models.FalconCausalLM.from_preset(
+        "falcon_refinedweb_1b_en"
+    )
     falcon_lm.fit(x=features, batch_size=2)
     ```
 
     Call `fit()` without preprocessing.
     ```python
     x = {
-        # Token ids for "<|endoftext|> Keras is deep learning library<|endoftext|>"
-        "token_ids": np.array([[50256, 17337,   292,   318,  2769,  4673,  5888, 50256, 0]] * 2),
+        # Token ids for
+        # "<|endoftext|> Keras is deep learning library<|endoftext|>"
+        "token_ids": np.array(
+            [[50256, 17337, 292, 318, 2769,4673,5888, 50256, 0]] * 2
+        ),
         "padding_mask": np.array([[1, 1, 1, 1, 1, 1, 1, 1, 0]] * 2),
     }
     y = np.array([[17337,   292,   318,  2769,  4673,  5888, 50256, 0, 0]] * 2)
@@ -164,8 +174,8 @@ class FalconCausalLM(CausalLM):
         Args:
             token_ids: a dense int Tensor with shape `(batch_size, max_length)`.
             cache: a dense float Tensor, the cache of key and value.
-            cache_update_index: int, or int Tensor. The index of current inputs in the
-                whole sequence.
+            cache_update_index: int, or int Tensor. The index of current inputs
+                in the whole sequence.
 
         Returns:
             A (logits, hidden_states, cache) tuple. Where `logits` is the

keras_hub/src/models/falcon/falcon_tokenizer.py

@@ -36,7 +36,9 @@ class FalconTokenizer(BytePairTokenizer):
 
     ```python
     # Unbatched input.
-    tokenizer = keras_hub.models.FalconTokenizer.from_preset("falcon_refinedweb_1b_en")
+    tokenizer = keras_hub.models.FalconTokenizer.from_preset(
+        "falcon_refinedweb_1b_en"
+    )
     tokenizer("The quick brown fox jumped.")
 
     # Batched input.
@@ -49,7 +51,10 @@ class FalconTokenizer(BytePairTokenizer):
     vocab = {"<|endoftext|>": 0, "a": 4, "Ġquick": 5, "Ġfox": 6}
     merges = ["Ġ q", "u i", "c k", "ui ck", "Ġq uick"]
     merges += ["Ġ f", "o x", "Ġf ox"]
-    tokenizer = keras_hub.models.FalconTokenizer(vocabulary=vocab, merges=merges)
+    tokenizer = keras_hub.models.FalconTokenizer(
+        vocabulary=vocab,
+        merges=merges,
+    )
     tokenizer("a quick fox.")
     ```
     """