DisabledDropout

class penzai.nn.dropout.DisabledDropout

Bases: Identity

A no-op layer taking the place of a disabled StochasticDropout layer.

This layer can be used to mark a location in a model where dropout could be applied, or where dropout was originally applied before disabling it. Its attributes are unused, except that they can be used to rebuild a StochasticDropout layer.

Variables:

• drop_rate (float) – Drop rate for the enabled (StochasticDropout) version of this layer.

• share_across_axes (tuple[str, ...]) – Shared axis names for the enabled (StochasticDropout) version of this layer.

• random_stream_input_name (str) – The side input key for the random stream used by the model at runtime. (Ignored unless the dropout layer is reenabled.)

Methods

__init__(drop_rate, share_across_axes[, ...])
enable()	Returns a stochastic, enabled version of this layer.

Attributes

random_stream_input_name
drop_rate
share_across_axes

Inherited Methods

(expand to view inherited methods)

attributes_dict()	Constructs a dictionary with all of the fields in the class.
bind_variables(variables[, allow_unused])	Convenience function to bind variables to a layer.
from_attributes(**field_values)	Directly instantiates a struct given all of its fields.
key_for_field(field_name)	Generates a JAX PyTree key for a given field name.
select()	Wraps this struct in a selection, enabling functional-style mutations.
stateless_call(variable_values, argument, /, ...)	Calls a layer with temporary variables, without modifying its state.
tree_flatten()	Flattens this tree node.
tree_flatten_with_keys()	Flattens this tree node with keys.
tree_unflatten(aux_data, children)	Unflattens this tree node.
treescope_color()
__call__(value, **_unused_side_inputs)	Returns the input unchanged.

enable() → StochasticDropout

Returns a stochastic, enabled version of this layer.