The Keras project isn't limited to the core Keras API for building and training neural networks. It spans a wide range of related initiatives that cover every step of the machine learning workflow.
KerasHub Documentation - KerasHub GitHub repository
KerasHub is a pretrained modeling library that aims to be simple, flexible, and fast. The library provides Keras implementations of popular model architectures, paired with a collection of pretrained checkpoints. Models can be used with text, image, and audio data for generation, classification, and many other built-in tasks.
KerasRS Documentation - KerasRS GitHub repository
Keras Recommenders is a library for building recommender systems on top of Keras. It provides a collection of building blocks which help with the full workflow of creating a recommender system.
KerasTuner Documentation - KerasTuner GitHub repository
KerasTuner is an easy-to-use, scalable hyperparameter optimization framework that solves the pain points of hyperparameter search. Easily configure your search space with a define-by-run syntax, then leverage one of the available search algorithms to find the best hyperparameter values for your models. KerasTuner comes with Bayesian Optimization, Hyperband, and Random Search algorithms built-in, and is also designed to be easy for researchers to extend in order to experiment with new search algorithms.
AutoKeras Documentation - AutoKeras GitHub repository
AutoKeras is an AutoML system based on Keras. It is developed by DATA Lab at Texas A&M University.
The goal of AutoKeras is to make machine learning accessible for everyone. It provides high-level end-to-end APIs
such as ImageClassifier or
TextClassifier to solve machine learning problems in a few lines,
as well as flexible building blocks to perform architecture search.
import autokeras as ak
clf = ak.ImageClassifier()
clf.fit(x_train, y_train)
results = clf.predict(x_test)
BayesFlow documentation - BayesFlow
A Python library for amortized Bayesian workflows using generative neural networks, built on Keras 3, featuring:
NEAT documentation - NEAT GitHub repository
NEAT (Nash-Equilibrium Adaptive Training) is a Keras-first optimizer library for conflict-aware neural network training.
It provides a Keras 3 optimizer usable with model.compile(...), plus a NumPy reference engine for validating the update rule.
Features include:
import keras
from neat_optim import NEAT
model.compile(
optimizer=NEAT(learning_rate=1e-3),
loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),
)