Map a function over leading array axes. Like Python’s builtin map, except inputs and outputs are in the form of stacked arrays. Consider usin

DEPRECATED.

Compile the distiller with proper integration. Arguments: optimizer: Optimizer for training the student model. loss:

Prepares and chunks the calibration dataloader, repeating short datasets. All processing happens on the CPU. Args: tokenizer: Th

Determines if a given filepath indicates a remote location. This function checks if the filepath represents a known remote pattern such

The number of dimensions of the variable.

Transforms non-negative/positive integer/integers into an integer tuple. Args: value: int or iterable of ints. The value to validate and

Stateless version of `apply` that returns modified variables. Args: optimizer_variables: list of tensors containing the current v

check penalty number availability, raise ValueError if failed.

Instantiates the EfficientNet architecture. Args: width_coefficient: float, scaling coefficient for network width. depth_coefficient:

Convert the padding arguments from Keras to the ones used by Torch. Torch starts with an output shape of `(input-1) * stride + kernel_size`, t

Clips bounding boxes to be within the image dimensions. Args: bounding_boxes: A dictionary with 'boxes' shape `(N, 4)` or `(ba

Checks whether NNX specific features are enabled for the JAX backend. Returns: bool: `True` if NNX backend features are enabled, `False`

Rearranges the axes of a Keras tensor according to a specified pattern, einops-style. Args: tensor: Input Keras tensor. patte

Instantiates the EfficientNetV2 architecture using given scaling coefficients. Args: width_coefficient: float, scaling coefficient fo

Returns model metrics as a dict if the keys match with input logs. When the training / evaluation is performed with an asynchronous steps,

Registers call-context args for this layer. If this layer declares a `call()` method that accepts one or more of the given args, thos

DEPRECATED.

Publicly accessible method for determining the current backend. Returns: String, the name of the backend Keras is currently using. One of

Compute the total loss, validate it, and return it. Subclasses can optionally override this method to provide custom loss computation

Retrieves a Keras `DTypePolicy` instance. The `identifier` may be the string name of a `DTypePolicy` class. >>> policy = dtype_policies.get(

Retrieves a layer based on either its name (unique) or index. If `name` and `index` are both provided, `index` will take precedence.

Quantize a float tensor into discrete levels [0, maxq] using per-tensor/per-channel/grouped scaling. Returns `q` (same dtype as inputs/scales

Retrieve a Group within a given File or Group. Args: parent: the parent h5py.File or h5py.Group. name: the name of the Group to r

Check if two shapes are equal. Args: shape1: A list or tuple of integers for first shape to be compared. shape2: A list or tuple

Backwards compatibility wrapper for `compute_loss`. This should be used instead `compute_loss` within `train_step` and `test_step` to

Ensure the current shard is the last one created.

Switch to a different H5 file with the specified mode. This is useful for retrieving information from all shards, such as the total l