Build the layer at `Layer.__init__`. We can only safely mark the layer as `built=True` in `Layer.__init__` if `build` is not overridd

Update traininable variables according to provided gradient values. `grads` should be a list of gradient tensors with 1:1 mapping to

Computes the mean and variance of values in a dataset. Calling `adapt()` on a `Normalization` layer is an alternative to passing in `

DEPRECATED.

Fits the data generator to some sample data. This computes the internal data stats related to the data-dependent transformations, bas

Sets the values of `layer.weights` from a list of NumPy arrays.

Computes the scale and zero-point for quantizing weight tensors. This function calculates the scale and zero-point required for quantizing

Get the H5 entry group. This method is only available in read mode. Args: path: `str`. The variable path.

Returns the current vocabulary of the layer. Args: include_special_tokens: If `True`, the returned vocabulary wil

Squeeze/expand last dim if ranks differ from expected by exactly 1.

Prints a string summary of the network. Args: line_length: Total length of printed lines (e.g. set this to adapt

Register a new serving endpoint. Args: name: `str`. The name of the endpoint. fn: A callable. It should only leverage

Assert that two arrays are equal. Given two array-like objects, check that their shapes and all elements are equal.

Creates a generator to extract data from the queue. Skip the data if it is `None`. This method is called from the main thread.

Sets the default dtype policy globally. Example: >>> keras.config.set_dtype_policy("mixed_float16")

Write the corresponding SavedModel to disk. Arguments: filepath: `str` or `pathlib.Path` object. Path where to sa

Convert the non-`None` value to either a tuple or a list.

check penalty number availability, raise ValueError if failed.

Compute small and big window sizes for adaptive pooling.

Converts bounding boxes between formats. Supported formats (case-insensitive): `"xyxy"`: [left, top, right, bottom] `"yxyx"`: [top, left,

Rebuild a dequantized weight matrix from group params. This function uses the provided scale and zero tensors to dequantize the input weights

Exclude variables from weight decay. This method must be called before the optimizer's `build` method is called. You can set specific

Track the variables (of a layer or model) and other assets. By default, all variables used by an endpoint function are automatically

Add a value to a variable. This should be used in optimizers instead of `variable.assign_add(value)` to support backend specific opti

Backward pass computation specification. Args: ctx: Context object. grad_output: Gradient with respect to the output.

Call the layer without any side effects. Args: trainable_variables: List of trainable variables of the model. non_tra

Sum of a tensor over the given axes. Args: x: Input tensor. axis: Axis or axes along which the sum is computed. The default is to

Turns `logs` dict into a list as per key order of `metrics_names`.

Computes a vocabulary of string terms from tokens in a dataset. Calling `adapt()` on a `TextVectorization` layer is an alternative to

Flatten and convert log values to Python-native types. This function attempts to convert dict value by `float(value)` and skips the conversio

Add optimizer variables from the list of trainable model variables. Create an optimizer variable based on the information of the supplied

Subtract a value from a variable. This should be used in optimizers instead of `variable.assign_sub(value)` to support backend specif

Returns the current vocabulary of the layer. Args: include_special_tokens: If `True`, the returned vocabulary wil