reference.rst

API reference

This section details the modules, classes, and functions available in NengoDL. It is divided into two sections. The first section describes the objects relevant to :ref:`NengoDL users <user-api>`. More information on these objects can also be found in the :doc:`user-guide`. The second section describes objects that only :ref:`NengoDL developers <developer-api>` need to worry about.

Users

These objects are the main access points for the user-facing features of NengoDL.

Simulator

.. automodule:: nengo_dl.simulator
    :no-members:

    .. autoautosummary:: nengo_dl.Simulator
        :nosignatures:

        nengo_dl.Simulator

    .. autosummary::
        :nosignatures:

        nengo_dl.simulator.SimulationData

    .. autoclass:: nengo_dl.Simulator
        :exclude-members: unsupported, dt

    .. autoclass:: nengo_dl.simulator.SimulationData
        :special-members:
        :exclude-members: __init__, __weakref__

TensorNodes

.. automodule:: nengo_dl.tensor_node
    :no-members:

    .. autosummary::
        :nosignatures:

        nengo_dl.TensorNode
        nengo_dl.Layer

    .. autoclass:: nengo_dl.TensorNode

    .. autoclass:: nengo_dl.Layer
        :special-members:
        :exclude-members: __init__, __weakref__

Converter

nengo_dl.Converter can be used to automatically convert a Keras model to a native Nengo Network. This can be useful if, for example, you want to run a model in different Nengo Simulator backends (which will only support the core Nengo objects).

See the documentation for more details.

.. autosummary::
    :nosignatures:

    nengo_dl.Converter

.. autoclass:: nengo_dl.Converter

Configuration system

.. automodule:: nengo_dl.config

    .. autoautosummary:: nengo_dl.config
        :nosignatures:

Neuron types

.. automodule:: nengo_dl.neurons

    .. autoautosummary:: nengo_dl.neurons
        :nosignatures:

Distributions

.. automodule:: nengo_dl.dists

    .. autoautosummary:: nengo_dl.dists
        :nosignatures:

Loss functions

.. automodule:: nengo_dl.losses

    .. autoautosummary:: nengo_dl.losses
        :nosignatures:

Callbacks

.. automodule:: nengo_dl.callbacks

    .. autoautosummary:: nengo_dl.callbacks
        :nosignatures:

Developers

These objects are only relevant to people interested in modifying the implementation of NengoDL (e.g., adding a new neuron type).

Builder

.. automodule:: nengo_dl.builder

    .. autoautosummary:: nengo_dl.builder
        :nosignatures:

Operator builders

These objects are used to convert Nengo operators into TensorFlow graph elements.

Basic operators

.. automodule:: nengo_dl.op_builders

    .. autoautosummary:: nengo_dl.op_builders
        :nosignatures:

Neuron types

.. automodule:: nengo_dl.neuron_builders

    .. autoautosummary:: nengo_dl.neuron_builders
        :nosignatures:

Learning rules

.. automodule:: nengo_dl.learning_rule_builders

    .. autoautosummary:: nengo_dl.learning_rule_builders
        :nosignatures:

Processes

.. automodule:: nengo_dl.process_builders

    .. autoautosummary:: nengo_dl.process_builders
        :nosignatures:

Transforms

.. automodule:: nengo_dl.transform_builders

    .. autoautosummary:: nengo_dl.transform_builders
        :nosignatures:

TensorNodes

To build .TensorNode objects we need to define a new Nengo operator (.tensor_node.SimTensorNode), a build function that adds that operator into a Nengo graph (.tensor_node.build_tensor_node), and a NengoDL build class that maps that new Nengo operator to TensorFlow operations (.tensor_node.SimTensorNodeBuilder).

.. autosummary::
    :nosignatures:

    nengo_dl.tensor_node.SimTensorNode
    nengo_dl.tensor_node.build_tensor_node
    nengo_dl.tensor_node.SimTensorNodeBuilder

.. autoclass:: nengo_dl.tensor_node.SimTensorNode

.. autofunction:: nengo_dl.tensor_node.build_tensor_node

.. autoclass:: nengo_dl.tensor_node.SimTensorNodeBuilder

Graph construction

.. automodule:: nengo_dl.tensor_graph

    .. autoautosummary:: nengo_dl.tensor_graph
        :nosignatures:

Signals

.. automodule:: nengo_dl.signals
    :exclude-members: TensorSignal

    .. autoautosummary:: nengo_dl.signals
        :nosignatures:

    .. autoclass:: nengo_dl.signals.TensorSignal
        :special-members: __getitem__

Graph optimization

.. automodule:: nengo_dl.graph_optimizer

    .. autoautosummary:: nengo_dl.graph_optimizer
        :nosignatures:

Layer converters

.. automodule:: nengo_dl.converter
    :exclude-members: Converter

    .. autoautosummary:: nengo_dl.converter
        :nosignatures:
        :exclude-members: Converter

Utilities

.. automodule:: nengo_dl.utils
    :exclude-members: MessageBar

    .. autoautosummary:: nengo_dl.utils
        :nosignatures:
        :exclude-members: MessageBar

Benchmarks

.. automodule:: nengo_dl.benchmarks

    .. autoautosummary:: nengo_dl.benchmarks
        :nosignatures:

Interface

The benchmark module also includes a command-line interface for building and running the benchmarks:

.. click:: nengo_dl.benchmarks:main
    :prog: benchmarks
    :show-nested: