Changes requested

benchmarks/vectorized_random_sharpness.py

@@ -18,7 +18,7 @@ class OldRandomSharpness(BaseImageAugmentationLayer):
     an image less sharp than they were in the original image.
 
     References:
-        - [PIL](https://pillow.readthedocs.io/en/stable/reference/ImageEnhance.html)  # noqa: E501
+        - [PIL](https://pillow.readthedocs.io/en/stable/reference/ImageEnhance.html)
 
     Args:
         factor: A tuple of two floats, a single float or
@@ -36,7 +36,7 @@ class OldRandomSharpness(BaseImageAugmentationLayer):
             Represented as a two number tuple written [low, high].
             This is typically either `[0, 1]` or `[0, 255]` depending
             on how your preprocessing pipeline is set up.
-    """
+    """  # noqa: E501
 
     def __init__(
         self,

keras_cv/bounding_box/iou.py

@@ -93,8 +93,8 @@ def compute_iou(
         unbatched.
       bounding_box_format: a case-insensitive string which is one of `"xyxy"`,
         `"rel_xyxy"`, `"xyWH"`, `"center_xyWH"`, `"yxyx"`, `"rel_yxyx"`.
-        For detailed information on the supported format, see the [KerasCV
-        bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
+        For detailed information on the supported format, see the
+        [KerasCV bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
       use_masking: whether masking will be applied. This will mask all `boxes1`
         or `boxes2` that have values less than 0 in all its 4 dimensions.
         Default to `False`.

keras_cv/datasets/pascal_voc/load.py

@@ -61,9 +61,9 @@ def load(
         split: the split string passed to the `tensorflow_datasets.load()` call.
             Should be one of "train", "test", or "validation."
         bounding_box_format: the keras_cv bounding box format to load the boxes
-            into. For a list of supported formats, please refer [to the keras.io
-            docs](https://keras.io/api/keras_cv/bounding_box/formats/) for more
-            details on supported bounding box formats.
+            into. For a list of supported formats, please refer
+            [to the keras.io docs](https://keras.io/api/keras_cv/bounding_box/formats/)
+            for more details on supported bounding box formats.
         batch_size: how many instances to include in batches after loading
         shuffle_buffer: the size of the buffer to use in shuffling.
         shuffle_files: (Optional) whether to shuffle files, defaults to
@@ -76,7 +76,7 @@ def load(
         containing keys {"images": images, "bounding_boxes": bounding_boxes}
         where images is a Tensor of shape [batch, H, W, 3] and bounding_boxes is
         a `tf.RaggedTensor` of shape [batch, None, 5].
-    """
+    """  # noqa: E501
     if dataset not in ["voc/2007", "voc/2012"]:
         raise ValueError(
             "keras_cv.datasets.pascal_voc.load() expects the `dataset` "

keras_cv/layers/fusedmbconv.py

@@ -34,11 +34,8 @@ class FusedMBConvBlock(layers.Layer):
     """
     Implementation of the FusedMBConv block (Fused Mobile Inverted Residual
     Bottleneck) 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: 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).
 
     FusedMBConv blocks are based on MBConv blocks, and replace the depthwise and
     1x1 output convolution blocks with a single 3x3 convolution block, fusing
@@ -87,7 +84,7 @@ class FusedMBConvBlock(layers.Layer):
     output = layer(inputs)
     output.shape # TensorShape([1, 224, 224, 48])
     ```
-    """
+    """  # noqa: E501
 
     def __init__(
         self,

keras_cv/layers/mbconv.py

@@ -47,8 +47,7 @@ def __init__(
         """
         Implementation of the MBConv block (Mobile Inverted Residual Bottleneck)
         from:
-            (MobileNetV2: Inverted Residuals and Linear Bottlenecks)
-            [https://arxiv.org/abs/1801.04381v4].
+            [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381v4).
 
         MBConv blocks are common blocks used in mobile-oriented and efficient
         architectures, present in architectures such as MobileNet, EfficientNet,
@@ -96,7 +95,8 @@ def __init__(
         output = layer(inputs)
         output.shape # TensorShape([1, 64, 64, 32])
         ```
-        """
+        """  # noqa: E501
+
         super().__init__(**kwargs)
         self.input_filters = input_filters
         self.output_filters = output_filters

keras_cv/layers/object_detection/anchor_generator.py

@@ -31,8 +31,9 @@ class AnchorGenerator(keras.layers.Layer):
     levels), or a dictionary from `{'level_{number}': [parameters at scale...]}`
 
     Args:
-      bounding_box_format: The format of bounding boxes to generate. Refer [to
-        the keras.io docs](https://keras.io/api/keras_cv/bounding_box/formats/) for more details on supported bounding box formats.
+      bounding_box_format: The format of bounding boxes to generate. Refer
+        [to the keras.io docs](https://keras.io/api/keras_cv/bounding_box/formats/)
+        for more details on supported bounding box formats.
       sizes: A list of integers that represent the anchor sizes for each level,
         or a dictionary of integer lists with each key representing a level.
         For each anchor size, anchor height will be

keras_cv/layers/object_detection/roi_generator.py

@@ -45,8 +45,8 @@ class ROIGenerator(keras.layers.Layer):
 
     Args:
         bounding_box_format: a case-insensitive string.
-            For detailed information on the supported format, see the [KerasCV
-            bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
+            For detailed information on the supported format, see the
+            [KerasCV bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
         pre_nms_topk_train: int. number of top k scoring proposals to keep
             before applying NMS in training mode. When RPN is run on multiple
             feature maps / levels (as in FPN) this number is per

keras_cv/layers/object_detection/roi_pool.py

@@ -33,8 +33,8 @@ class ROIPooler(keras.layers.Layer):
 
     Args:
         bounding_box_format: a case-insensitive string.
-            For detailed information on the supported format, see the [KerasCV
-            bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
+            For detailed information on the supported format, see the
+            [KerasCV bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
         target_size: List or Tuple of 2 integers of the pooled shape
         image_shape: List of Tuple of 3 integers, or `TensorShape` of the input
             image shape.

keras_cv/layers/object_detection/roi_sampler.py

@@ -42,8 +42,9 @@ class _ROISampler(keras.layers.Layer):
     if its range is [0, num_classes).
 
     Args:
-      bounding_box_format: The format of bounding boxes to generate. Refer [to
-        the keras.io docs](https://keras.io/api/keras_cv/bounding_box/formats/) for more details on supported bounding box formats.
+      bounding_box_format: The format of bounding boxes to generate. Refer
+        [to the keras.io docs](https://keras.io/api/keras_cv/bounding_box/formats/)
+        for more details on supported bounding box formats.
       roi_matcher: a `BoxMatcher` object that matches proposals with ground
         truth boxes. The positive match must be 1 and negative match must be -1.
         Such assumption is not being validated here.

keras_cv/layers/preprocessing/mosaic.py

@@ -59,8 +59,8 @@ class Mosaic(VectorizedBaseImageAugmentationLayer):
             bounding box is defined by at least these 4 values. The inputs may
             contain additional information such as classes and confidence after
             these 4 values but these values will be ignored and returned as is.
-            For detailed information on the supported formats, see the [KerasCV
-            bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
+            For detailed information on the supported formats, see the
+            [KerasCV bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
             Defaults to None.
         seed: integer, used to create a random seed.
 

keras_cv/layers/preprocessing/posterization.py

@@ -27,8 +27,7 @@ class Posterization(BaseImageAugmentationLayer):
 
     References:
     - [AutoAugment: Learning Augmentation Policies from Data](https://arxiv.org/abs/1805.09501)
-    - [RandAugment: Practical automated data augmentation with a reduced search
-        space](https://arxiv.org/abs/1909.13719)
+    - [RandAugment: Practical automated data augmentation with a reduced search space](https://arxiv.org/abs/1909.13719)
 
     Args:
         value_range: a tuple or a list of two elements. The first value

keras_cv/layers/preprocessing/randomly_zoomed_crop.py

@@ -17,9 +17,6 @@
 from tensorflow import keras
 
 from keras_cv import core
-from keras_cv.layers.preprocessing.vectorized_base_image_augmentation_layer import (  # noqa: E501
-    IMAGES,
-)
 from keras_cv.layers.preprocessing.vectorized_base_image_augmentation_layer import (  # noqa: E501
     VectorizedBaseImageAugmentationLayer,
 )

keras_cv/layers/regularization/stochastic_depth.py

@@ -25,7 +25,7 @@ class StochasticDepth(keras.layers.Layer):
 
     Reference:
         - [Deep Networks with Stochastic Depth](https://arxiv.org/abs/1603.09382)
-        - Docstring taken from [stochastic_depth.py](https://tinyurl.com/mr3y2af6)
+        - [Docstring taken from [stochastic_depth.py](https://tinyurl.com/mr3y2af6)
 
     Args:
         rate: float, the probability of the residual branch being dropped.

keras_cv/layers/spatial_pyramid.py

@@ -25,10 +25,8 @@ class SpatialPyramidPooling(keras.layers.Layer):
     """Implements the Atrous Spatial Pyramid Pooling.
 
     References:
-        [Rethinking Atrous Convolution for Semantic Image Segmentation](
-          https://arxiv.org/pdf/1706.05587.pdf)
-    [Encoder-Decoder with Atrous Separable Convolution for Semantic Image
-        Segmentation](https://arxiv.org/pdf/1802.02611.pdf)
+        [Rethinking Atrous Convolution for Semantic Image Segmentation](https://arxiv.org/pdf/1706.05587.pdf)
+        [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://arxiv.org/pdf/1802.02611.pdf)
 
     inp = keras.layers.Input((384, 384, 3))
     backbone = keras.applications.EfficientNetB0(
@@ -39,7 +37,7 @@ class SpatialPyramidPooling(keras.layers.Layer):
         dilation_rates=[6, 12, 18])(output)
 
     # output[4].shape = [None, 16, 16, 256]
-    """
+    """  # noqa: E501
 
     def __init__(
         self,

keras_cv/losses/iou_loss.py

@@ -34,8 +34,8 @@ class IoULoss(keras.losses.Loss):
     Args:
         bounding_box_format: a case-insensitive string (for example, "xyxy").
             Each bounding box is defined by these 4 values. For detailed
-            information on the supported formats, see the [KerasCV bounding box
-            documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
+            information on the supported formats, see the
+            [KerasCV bounding box documentation](https://keras.io/api/keras_cv/bounding_box/formats/).
         mode: must be one of
             - `"linear"`. The loss will be calculated as 1 - iou
             - `"quadratic"`. The loss will be calculated as 1 - iou<sup>2</sup>
@@ -66,7 +66,7 @@ class IoULoss(keras.losses.Loss):
     ```python
     model.compile(optimizer='adam', loss=keras_cv.losses.IoULoss())
     ```
-    """
+    """  # noqa: E501
 
     def __init__(self, bounding_box_format, mode="log", axis=-1, **kwargs):
         super().__init__(**kwargs)

keras_cv/models/backbones/resnet_v1/resnet_v1_backbone.py

@@ -226,7 +226,7 @@ class ResNetBackbone(Backbone):
     """Instantiates the ResNet architecture.
 
     Reference:
-        - [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385)  # noqa: E501
+        - [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385)
 
     The difference in ResNetV1 and ResNetV2 rests in the structure of their
     individual building blocks. In ResNetV2, the batch normalization and
@@ -235,7 +235,7 @@ class ResNetBackbone(Backbone):
     convolution layers.
 
     For transfer learning use cases, make sure to read the
-    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).  # noqa: E501
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
 
     Args:
         stackwise_filters: list of ints, number of filters for each stack in
@@ -269,7 +269,7 @@ class ResNetBackbone(Backbone):
     )
     output = model(input_data)
     ```
-    """
+    """  # noqa: E501
 
     def __init__(
         self,
@@ -360,7 +360,7 @@ def presets_with_weights(cls):
 ALIAS_DOCSTRING = """ResNetBackbone (V1) model with {num_layers} layers.
 
     Reference:
-        - [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385)  # noqa: E501
+        - [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385)
 
     The difference in ResNetV1 and ResNetV2 rests in the structure of their
     individual building blocks. In ResNetV2, the batch normalization and
@@ -369,7 +369,7 @@ def presets_with_weights(cls):
     convolution layers.
 
     For transfer learning use cases, make sure to read the
-    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).  # noqa: E501
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
 
     Args:
         include_rescaling: bool, whether to rescale the inputs. If set
@@ -387,7 +387,7 @@ def presets_with_weights(cls):
     model = ResNet{num_layers}Backbone()
     output = model(input_data)
     ```
-"""
+"""  # noqa: E501
 
 
 class ResNet18Backbone(ResNetBackbone):

keras_cv/models/backbones/resnet_v2/resnet_v2_backbone.py

@@ -268,8 +268,8 @@ class ResNetV2Backbone(Backbone):
     the batch normalization and ReLU activation are applied after the
     convolution layers.
 
-    For transfer learning use cases, make sure to read the [guide to transfer
-    learning & fine-tuning](https://keras.io/guides/transfer_learning/).
+    For transfer learning use cases, make sure to read the
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
 
     Args:
         stackwise_filters: list of ints, number of filters for each stack in
@@ -416,8 +416,8 @@ def presets_with_weights(cls):
     the batch normalization and ReLU activation are applied after the
     convolution layers.
 
-    For transfer learning use cases, make sure to read the [guide to transfer
-    learning & fine-tuning](https://keras.io/guides/transfer_learning/).
+    For transfer learning use cases, make sure to read the
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
 
     Args:
         include_rescaling: bool, whether to rescale the inputs. If set

keras_cv/models/convnext.py

@@ -55,8 +55,8 @@
 }
 
 BASE_DOCSTRING = """Instantiates the {name} architecture.
-    - [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545)
-    (CVPR 2022)
+    - [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) (CVPR 2022)
+
     This function returns a Keras {name} model.
     Args:
         include_rescaling: bool, whether to rescale the inputs. If set

keras_cv/models/csp_darknet.py

@@ -59,8 +59,8 @@
         - [CSPNet Paper](https://arxiv.org/pdf/1911.11929)
         - [YoloX Paper](https://arxiv.org/abs/2107.08430)
         - [YoloX implementation](https://github.com/ultralytics/yolov3)
-    For transfer learning use cases, make sure to read the [guide to transfer
-    learning & fine-tuning](https://keras.io/guides/transfer_learning/).
+    For transfer learning use cases, make sure to read the
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
     Args:
         include_rescaling: bool, whether to rescale the inputs. If set to
             True, inputs will be passed through a `Rescaling(1/255.0)` layer.
@@ -93,7 +93,7 @@
         name: string, optional name to pass to the model, defaults to "{name}".
     Returns:
         A `keras.Model` instance.
-"""
+"""  # noqa: E501
 
 
 @keras.utils.register_keras_serializable(package="keras_cv.models")
@@ -107,8 +107,8 @@ class CSPDarkNet(keras.Model):
         - [CSPNet Paper](https://arxiv.org/pdf/1911.11929)
         - [YoloX Paper](https://arxiv.org/abs/2107.08430)
         - [YoloX implementation](https://github.com/ultralytics/yolov3)
-    For transfer learning use cases, make sure to read the [guide to transfer
-    learning & fine-tuning](https://keras.io/guides/transfer_learning/).
+    For transfer learning use cases, make sure to read the
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
     Args:
         depth_multiplier: A float value used to calculate the base depth of the
             model this changes based the detection model being used.
@@ -145,7 +145,7 @@ class CSPDarkNet(keras.Model):
         name: (Optional) name to pass to the model, defaults to "CSPDarkNet".
     Returns:
         A `keras.Model` instance.
-    """
+    """  # noqa: E501
 
     def __init__(
         self,

keras_cv/models/darknet.py

@@ -41,8 +41,7 @@
         - [YoloV3 implementation](https://github.com/ultralytics/yolov3)
 
     For transfer learning use cases, make sure to read the
-    [guide to transfer learning & fine-tuning](
-        https://keras.io/guides/transfer_learning/).
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
 
     Args:
         include_rescaling: bool, whether to rescale the inputs. If set to
@@ -68,7 +67,7 @@
 
     Returns:
         A `keras.Model` instance.
-"""
+"""  # noqa: E501
 
 
 @keras.utils.register_keras_serializable(package="keras_cv.models")
@@ -83,8 +82,8 @@ class DarkNet(keras.Model):
     Reference:
         - [YoloV3 Paper](https://arxiv.org/abs/1804.02767)
         - [YoloV3 implementation](https://github.com/ultralytics/yolov3)
-    For transfer learning use cases, make sure to read the [guide to transfer
-    learning & fine-tuning](https://keras.io/guides/transfer_learning/).
+    For transfer learning use cases, make sure to read the
+    [guide to transfer learning & fine-tuning](https://keras.io/guides/transfer_learning/).
 
     Args:
         blocks: integer, numbers of building blocks from the layer dark2 to
@@ -116,7 +115,7 @@ class DarkNet(keras.Model):
 
     Returns:
         A `keras.Model` instance.
-    """
+    """  # noqa: E501
 
     def __init__(
         self,