Introducing Keras 3 for R

We’re thrilled to introduce keras3, the subsequent model of the Keras R
bundle. keras3 is a ground-up rebuild of {keras}, sustaining the
beloved options of the unique whereas refining and simplifying the API
primarily based on precious insights gathered over the previous few years.

Keras gives a whole toolkit for constructing deep studying fashions in
R—it’s by no means been simpler to construct, prepare, consider, and deploy deep
studying fashions.

Set up

To put in Keras 3:

set up.packages("keras3")
library(keras3)
install_keras()

What’s new:

Documentation

Nice documentation is important, and we’ve labored exhausting to ensure
that keras3 has glorious documentation, each now, and sooner or later.

Keras 3 comes with a full refresh of the web site:
https://keras.posit.co. There, you’ll discover guides, tutorials,
reference pages with rendered examples, and a brand new examples gallery. All
the reference pages and guides are additionally out there by way of R’s built-in assist
system.

In a fast-paced ecosystem like deep studying, creating nice
documentation and wrappers as soon as isn’t sufficient. There additionally have to be
workflows that make sure the documentation is up-to-date with upstream
dependencies. To perform this, {keras3} consists of two new maintainer
options that make sure the R documentation and performance wrappers will keep
up-to-date:

• We now take snapshots of the upstream documentation and API floor.
  With every launch, all R documentation is rebased on upstream
  updates. This workflow ensures that each one R documentation (guides,
  examples, vignettes, and reference pages) and R operate signatures
  keep up-to-date with upstream. This snapshot-and-rebase
  performance is carried out in a brand new standalone R bundle,
  {doctether}, which can
  be helpful for R bundle maintainers needing to maintain documentation in
  parity with dependencies.

• All examples and vignettes can now be evaluated and rendered throughout
  a bundle construct. This ensures that no stale or damaged instance code
  makes it right into a launch. It additionally means all person going through instance code
  now moreover serves as an prolonged suite of snapshot unit and
  integration assessments.

  Evaluating code in vignettes and examples continues to be not permitted
  in keeping with CRAN restrictions. We work across the CRAN restriction
  by including extra bundle construct steps that pre-render
  examples
  and
  vignettes.

Mixed, these two options will make it considerably simpler for Keras
in R to keep up function parity and up-to-date documentation with the
Python API to Keras.

Multi-backend help

Quickly after its launch in 2015, Keras featured help for hottest
deep studying frameworks: TensorFlow, Theano, MXNet, and CNTK. Over
time, the panorama shifted; Theano, MXNet, and CNTK had been retired, and
TensorFlow surged in recognition. In 2021, three years in the past, TensorFlow
turned the premier and solely supported Keras backend. Now, the panorama
has shifted once more.

Keras 3 brings the return of multi-backend help. Select a backend by
calling:

use_backend("jax") # or "tensorflow", "torch", "numpy"

The default backend continues to be TensorFlow, which is the only option
for many customers right this moment; for small-to-medium sized fashions that is nonetheless the
quickest backend. Nevertheless, every backend has totally different strengths, and
with the ability to swap simply will allow you to adapt to modifications as your
undertaking, or the frameworks themselves, evolve.

Right this moment, switching to the Jax backend can, for some mannequin varieties, carry
substantial velocity enhancements. Jax can also be the one backend that has
help for a brand new mannequin parallelism distributed coaching API. Switching
to Torch could be useful throughout improvement, usually producing easier
trackbacks whereas debugging.

Keras 3 additionally permits you to incorporate any pre-existing Torch, Jax, or Flax
module as a typical Keras layer by utilizing the suitable wrapper,
letting you construct atop current tasks with Keras. For instance, prepare
a Torch mannequin utilizing the Keras high-level coaching API (compile() +
match()), or embrace a Flax module as a element of a bigger Keras
mannequin. The brand new multi-backend help permits you to use Keras à la carte.

The ‘Ops’ household

{keras3} introduces a brand new “Operations” household of operate. The Ops
household, at present with over 200
features,
gives a complete suite of operations sometimes wanted when
working on nd-arrays for deep studying. The Operation household
supersedes and enormously expands on the previous household of backend features
prefixed with k_ within the {keras} bundle.

The Ops features allow you to write backend-agnostic code. They supply a
uniform API, no matter for those who’re working with TensorFlow Tensors,
Jax Arrays, Torch Tensors, Keras Symbolic Tensors, NumPy arrays, or R
arrays.

The Ops features:

• all begin with prefix op_ (e.g., op_stack())
• all are pure features (they produce no side-effects)
• all use constant 1-based indexing, and coerce doubles to integers
  as wanted
• all are secure to make use of with any backend (tensorflow, jax, torch, numpy)
• all are secure to make use of in each keen and graph/jit/tracing modes

The Ops API consists of:

• Everything of the NumPy API (numpy.*)
• The TensorFlow NN API (tf.nn.*)
• Widespread linear algebra features (A subset of scipy.linalg.*)
• A subfamily of picture transformers
• A complete set of loss features
• And extra!

Ingest tabular knowledge with layer_feature_space()

keras3 gives a brand new set of features for constructing fashions that ingest
tabular knowledge: layer_feature_space() and a household of function
transformer features (prefix, feature_) for constructing keras fashions
that may work with tabular knowledge, both as inputs to a keras mannequin, or
as preprocessing steps in a knowledge loading pipeline (e.g., a
tfdatasets::dataset_map()).

See the reference
web page and an
instance utilization in a full end-to-end
instance
to be taught extra.

New Subclassing API

The subclassing API has been refined and prolonged to extra Keras
varieties.
Outline subclasses just by calling: Layer(), Loss(), Metric(),
Callback(), Constraint(), Mannequin(), and LearningRateSchedule().
Defining {R6} proxy courses is now not vital.

Moreover the documentation web page for every of the subclassing
features now incorporates a complete itemizing of all of the out there
attributes and strategies for that sort. Take a look at
?Layer to see what’s
doable.

Saving and Export

Keras 3 brings a brand new mannequin serialization and export API. It’s now a lot
easier to avoid wasting and restore fashions, and likewise, to export them for
serving.

• save_model()/load_model():
  A brand new high-level file format (extension: .keras) for saving and
  restoring a full mannequin.

  The file format is backend-agnostic. This implies that you would be able to convert
  educated fashions between backends, just by saving with one backend,
  after which loading with one other. For instance, prepare a mannequin utilizing Jax,
  after which convert to Tensorflow for export.

• export_savedmodel():
  Export simply the ahead cross of a mannequin as a compiled artifact for
  inference with TF
  Serving or (quickly)
  Posit Join. This
  is the simplest option to deploy a Keras mannequin for environment friendly and
  concurrent inference serving, all with none R or Python runtime
  dependency.

• Decrease degree entry factors:

  • save_model_weights() / load_model_weights():
    save simply the weights as .h5 information.
  • save_model_config() / load_model_config():
    save simply the mannequin structure as a json file.

• register_keras_serializable():
  Register customized objects to allow them to be serialized and
  deserialized.

• serialize_keras_object() / deserialize_keras_object():
  Convert any Keras object to an R record of straightforward varieties that’s secure
  to transform to JSON or rds.

• See the brand new Serialization and Saving
  vignette
  for extra particulars and examples.

New random household

A brand new household of random tensor
turbines.
Just like the Ops household, these work with all backends. Moreover, all of the
RNG-using strategies have help for stateless utilization if you cross in a
seed generator. This allows tracing and compilation by frameworks that
have particular help for stateless, pure, features, like Jax. See
?random_seed_generator()
for instance utilization.

Different additions:

• New form()
  operate, one-stop utility for working with tensor shapes in all
  contexts.

• New and improved print(mannequin) and plot(mannequin) technique. See some
  examples of output within the Purposeful API
  information

• All new match() progress bar and stay metrics viewer output,
  together with new dark-mode help within the RStudio IDE.

• New config
  household,
  a curated set of features for getting and setting Keras world
  configurations.

• The entire different operate households have expanded with new members:

Migrating from {keras} to {keras3}

{keras3} supersedes the {keras} bundle.

In case you’re writing new code right this moment, you can begin utilizing {keras3} proper
away.

In case you have legacy code that makes use of {keras}, you’re inspired to
replace the code for {keras3}. For a lot of high-level API features, such
as layer_dense(), match(), and keras_model(), minimal to no modifications
are required. Nevertheless there’s a lengthy tail of small modifications that you simply
would possibly must make when updating code that made use of the lower-level
Keras API. A few of these are documented right here:
https://keras.io/guides/migrating_to_keras_3/.

In case you’re operating into points or have questions on updating, don’t
hesitate to ask on or
https://github.com/rstudio/keras/discussions.

The {keras} and {keras3} packages will coexist whereas the group
transitions. Through the transition, {keras} will proceed to obtain
patch updates for compatibility with Keras v2, which continues to be
printed to PyPi beneath the bundle title tf-keras. After tf-keras is
now not maintained, the {keras} bundle might be archived.

Abstract

In abstract, {keras3} is a strong replace to the Keras R bundle,
incorporating new options whereas preserving the convenience of use and
performance of the unique. The brand new multi-backend help,
complete suite of Ops features, refined mannequin serialization API,
and up to date documentation workflows allow customers to simply take
benefit of the most recent developments within the deep studying group.

Whether or not you’re a seasoned Keras person or simply beginning your deep
studying journey, Keras 3 gives the instruments and adaptability to construct,
prepare, and deploy fashions with ease and confidence. As we transition from
Keras 2 to Keras 3, we’re dedicated to supporting the group and
making certain a clean migration. We invite you to discover the brand new options,
take a look at the up to date documentation, and be part of the dialog on our
GitHub discussions web page. Welcome to the subsequent chapter of deep studying in
R with Keras 3!