Merge pull request #1504 from rstudio/retether-v3-9-2

Retether-v3-9-2

Author: t-kalinowski

.tether/man/InputLayer.txt

@@ -1,7 +1,7 @@
 Help on class InputLayer in module keras.src.layers.core.input_layer:
 
 class InputLayer(keras.src.layers.layer.Layer)
- |  InputLayer(shape=None, batch_size=None, dtype=None, sparse=None, batch_shape=None, input_tensor=None, optional=False, name=None, **kwargs)
+ |  InputLayer(shape=None, batch_size=None, dtype=None, sparse=None, ragged=None, batch_shape=None, input_tensor=None, optional=False, name=None, **kwargs)
  |
  |  Method resolution order:
  |      InputLayer
@@ -22,6 +22,7 @@ class InputLayer(keras.src.layers.layer.Layer)
  |    batch_size=None,
  |    dtype=None,
  |    sparse=None,
+ |    ragged=None,
  |    batch_shape=None,
  |    input_tensor=None,
  |    optional=False,

.tether/man/Layer.txt

@@ -26,12 +26,10 @@ class Layer(keras.src.backend.tensorflow.layer.TFLayer, keras.src.ops.operation.
  |      trainable: Boolean, whether the layer's variables should be trainable.
  |      name: String name of the layer.
  |      dtype: The dtype of the layer's computations and weights. Can also be a
- |          `keras.DTypePolicy`,
- |          which allows the computation and
- |          weight dtype to differ. Defaults to `None`. `None` means to use
- |          `keras.config.dtype_policy()`,
- |          which is a `float32` policy unless set to different value
- |          (via `keras.config.set_dtype_policy()`).
+ |          `keras.DTypePolicy`, which allows the computation and weight dtype
+ |          to differ. Defaults to `None`. `None` means to use
+ |          `keras.config.dtype_policy()`, which is a `float32` policy unless
+ |          set to different value (via `keras.config.set_dtype_policy()`).
  |
  |  Attributes:
  |      name: The name of the layer (string).
@@ -373,6 +371,20 @@ class Layer(keras.src.backend.tensorflow.layer.TFLayer, keras.src.ops.operation.
  |    **kwargs
  |  )
  |
+ |  rematerialized_call(
+ |    self,
+ |    layer_call,
+ |    *args,
+ |    **kwargs
+ |  )
+ |      Enable rematerialization dynamically for layer's call method.
+ |
+ |      Args:
+ |          layer_call: The original `call` method of a layer.
+ |
+ |      Returns:
+ |          Rematerialized layer's `call` method.
+ |
  |  save_own_variables(self, store)
  |      Saves the state of the layer.
  |

.tether/man/dataset_california_housing.txt

@@ -0,0 +1,65 @@
+__signature__
+keras.datasets.california_housing.load_data(
+  version='large',
+  path='california_housing.npz',
+  test_split=0.2,
+  seed=113
+)
+__doc__
+Loads the California Housing dataset.
+
+This dataset was obtained from the [StatLib repository](
+https://www.dcc.fc.up.pt/~ltorgo/Regression/cal_housing.html).
+
+It's a continuous regression dataset with 20,640 samples with
+8 features each.
+
+The target variable is a scalar: the median house value
+for California districts, in dollars.
+
+The 8 input features are the following:
+
+- MedInc: median income in block group
+- HouseAge: median house age in block group
+- AveRooms: average number of rooms per household
+- AveBedrms: average number of bedrooms per household
+- Population: block group population
+- AveOccup: average number of household members
+- Latitude: block group latitude
+- Longitude: block group longitude
+
+This dataset was derived from the 1990 U.S. census, using one row
+per census block group. A block group is the smallest geographical
+unit for which the U.S. Census Bureau publishes sample data
+(a block group typically has a population of 600 to 3,000 people).
+
+A household is a group of people residing within a home.
+Since the average number of rooms and bedrooms in this dataset are
+provided per household, these columns may take surprisingly large
+values for block groups with few households and many empty houses,
+such as vacation resorts.
+
+Args:
+    version: `"small"` or `"large"`. The small version
+        contains 600 samples, the large version contains
+        20,640 samples. The purpose of the small version is
+        to serve as an approximate replacement for the
+        deprecated `boston_housing` dataset.
+    path: path where to cache the dataset locally
+        (relative to `~/.keras/datasets`).
+    test_split: fraction of the data to reserve as test set.
+    seed: Random seed for shuffling the data
+        before computing the test split.
+
+Returns:
+    Tuple of Numpy arrays: `(x_train, y_train), (x_test, y_test)`.
+
+**`x_train`, `x_test`**: numpy arrays with shape `(num_samples, 8)`
+  containing either the training samples (for `x_train`),
+  or test samples (for `y_train`).
+
+**`y_train`, `y_test`**: numpy arrays of shape `(num_samples,)`
+    containing the target scalars. The targets are float scalars
+    typically between 25,000 and 500,000 that represent
+    the home prices in dollars.
+

.tether/man/export_savedmodel.keras.src.models.model.Model.txt

@@ -3,7 +3,7 @@ keras.Model.export(
   self,
   filepath,
   format='tf_saved_model',
-  verbose=True,
+  verbose=None,
   input_signature=None,
   **kwargs
 )
@@ -17,7 +17,8 @@ Args:
         `"tf_saved_model"` and `"onnx"`.  Defaults to
         `"tf_saved_model"`.
     verbose: `bool`. Whether to print a message during export. Defaults
-        to `True`.
+        to `None`, which uses the default value set by different
+        backends and formats.
     input_signature: Optional. Specifies the shape and dtype of the
         model inputs. Can be a structure of `keras.InputSpec`,
         `tf.TensorSpec`, `backend.KerasTensor`, or backend tensor. If
@@ -38,6 +39,10 @@ Args:
 **Note:** This feature is currently supported only with TensorFlow, JAX
 and Torch backends.
 
+**Note:** Be aware that the exported artifact may contain information
+from the local file system when using `format="onnx"`, `verbose=True`
+and Torch backend.
+
 Examples:
 
 Here's how to export a TensorFlow SavedModel for inference.

.tether/man/keras.layers.txt

@@ -24,6 +24,18 @@ Attention(
   seed=None,
   **kwargs
 )
+AugMix(
+  value_range=(0, 255),
+  num_chains=3,
+  chain_depth=3,
+  factor=0.3,
+  alpha=1.0,
+  all_ops=True,
+  interpolation='bilinear',
+  seed=None,
+  data_format=None,
+  **kwargs
+)
 AutoContrast(value_range=(0, 255), **kwargs)
 average(inputs, **kwargs)
 Average(**kwargs)
@@ -440,6 +452,12 @@ Cropping3D(
   data_format=None,
   **kwargs
 )
+CutMix(
+  factor=1.0,
+  seed=None,
+  data_format=None,
+  **kwargs
+)
 Dense(
   units,
   activation=None,
@@ -725,6 +743,7 @@ Input(
   batch_size=None,
   dtype=None,
   sparse=None,
+  ragged=None,
   batch_shape=None,
   name=None,
   tensor=None,
@@ -735,6 +754,7 @@ InputLayer(
   batch_size=None,
   dtype=None,
   sparse=None,
+  ragged=None,
   batch_shape=None,
   input_tensor=None,
   optional=False,
@@ -1008,12 +1028,30 @@ RandomCrop(
   name=None,
   **kwargs
 )
+RandomErasing(
+  factor=1.0,
+  scale=(0.02, 0.33),
+  fill_value=None,
+  value_range=(0, 255),
+  seed=None,
+  data_format=None,
+  **kwargs
+)
 RandomFlip(
   mode='horizontal_and_vertical',
   seed=None,
   data_format=None,
   **kwargs
 )
+RandomGaussianBlur(
+  factor=1.0,
+  kernel_size=3,
+  sigma=1.0,
+  value_range=(0, 255),
+  data_format=None,
+  seed=None,
+  **kwargs
+)
 RandomGrayscale(
   factor=0.5,
   data_format=None,
@@ -1027,6 +1065,22 @@ RandomHue(
   seed=None,
   **kwargs
 )
+RandomInvert(
+  factor=1.0,
+  value_range=(0, 255),
+  seed=None,
+  data_format=None,
+  **kwargs
+)
+RandomPerspective(
+  factor=1.0,
+  scale=1.0,
+  interpolation='bilinear',
+  fill_value=0.0,
+  seed=None,
+  data_format=None,
+  **kwargs
+)
 RandomPosterization(
   factor,
   value_range=(0, 255),
@@ -1108,9 +1162,15 @@ Resizing(
   pad_to_aspect_ratio=False,
   fill_mode='constant',
   fill_value=0.0,
+  antialias=False,
   data_format=None,
   **kwargs
 )
+RMSNormalization(
+  axis=-1,
+  epsilon=1e-06,
+  **kwargs
+)
 RNN(
   cell,
   return_sequences=False,

.tether/man/keras.ops.image.txt

@@ -24,6 +24,12 @@ extract_patches(
   padding='valid',
   data_format=None
 )
+gaussian_blur(
+  images,
+  kernel_size=(3, 3),
+  sigma=(1.0, 1.0),
+  data_format=None
+)
 hsv_to_rgb(images, data_format=None)
 map_coordinates(
   inputs,
@@ -42,6 +48,14 @@ pad_images(
   target_width=None,
   data_format=None
 )
+perspective_transform(
+  images,
+  start_points,
+  end_points,
+  interpolation='bilinear',
+  fill_value=0,
+  data_format=None
+)
 resize(
   images,
   size,

.tether/man/keras.ops.nn.txt

@@ -122,13 +122,20 @@ one_hot(
   dtype=None,
   sparse=False
 )
+polar(abs_, angle)
 psnr(
   x1,
   x2,
   max_val
 )
 relu(x)
 relu6(x)
+rms_normalization(
+  x,
+  scale=1,
+  axis=-1,
+  epsilon=None
+)
 selu(x)
 separable_conv(
   inputs,

.tether/man/keras.ops.numpy.txt

@@ -287,13 +287,19 @@ roll(
   shift,
   axis=None
 )
+rot90(
+  array,
+  k=1,
+  axes=(0, 1)
+)
 round(x, decimals=0)
 select(
   condlist,
   choicelist,
   default=0
 )
 sign(x)
+signbit(x)
 sin(x)
 sinh(x)
 size(x)

.tether/man/keras.ops.txt

@@ -146,7 +146,8 @@ convert_to_numpy(x)
 convert_to_tensor(
   x,
   dtype=None,
-  sparse=None
+  sparse=None,
+  ragged=None
 )
 copy(x)
 correlate(
@@ -462,6 +463,7 @@ pad(
   mode='constant',
   constant_values=None
 )
+polar(abs_, angle)
 power(x1, x2)
 prod(
   x,
@@ -484,6 +486,11 @@ quantile(
 )
 ravel(x)
 real(x)
+rearrange(
+  tensor,
+  pattern,
+  **axes_lengths
+)
 reciprocal(x)
 relu(x)
 relu6(x)
@@ -495,11 +502,22 @@ repeat(
 reshape(x, newshape)
 rfft(x, fft_length=None)
 right_shift(x, y)
+rms_normalization(
+  x,
+  scale=1,
+  axis=-1,
+  epsilon=None
+)
 roll(
   x,
   shift,
   axis=None
 )
+rot90(
+  array,
+  k=1,
+  axes=(0, 1)
+)
 round(x, decimals=0)
 rsqrt(x)
 saturate_cast(x, dtype)
@@ -556,6 +574,7 @@ separable_conv(
 shape(x)
 sigmoid(x)
 sign(x)
+signbit(x)
 silu(x)
 sin(x)
 sinh(x)