keras3: R Interface to 'Keras'

Interface to 'Keras' <>, a high-level neural networks API. 'Keras' was developed with a focus on enabling fast experimentation, supports both convolution based networks and recurrent networks (as well as combinations of the two), and runs seamlessly on both CPU and GPU devices.

Version: 1.1.0
Depends: R (≥ 4.0)
Imports: generics (≥ 0.0.1), reticulate (≥ 1.36.0), tensorflow (≥ 2.16.0), tfruns (≥ 1.5.2), magrittr, zeallot, fastmap, glue, cli, rlang
Suggests: ggplot2, testthat (≥ 2.1.0), knitr, rmarkdown, callr, tfdatasets, withr, png, jsonlite, purrr, rstudioapi, R6, jpeg
Published: 2024-07-17
DOI: 10.32614/CRAN.package.keras3
Author: Tomasz Kalinowski [aut, cph, cre], Daniel Falbel [ctb, cph], JJ Allaire [aut, cph], Fran├žois Chollet [aut, cph], Posit Software, PBC [cph, fnd], Google [cph, fnd], Yuan Tang ORCID iD [ctb, cph], Wouter Van Der Bijl [ctb, cph], Martin Studer [ctb, cph], Sigrid Keydana [ctb]
Maintainer: Tomasz Kalinowski <tomasz at>
License: MIT + file LICENSE
NeedsCompilation: no
Materials: NEWS
CRAN checks: keras3 results


Reference manual: keras3.pdf
Vignettes: Customizing what happens in 'fit()' with TensorFlow
Multi-GPU distributed training with TensorFlow
Distributed training with Keras 3
The Functional API
Getting Started with Keras
Introduction to Keras for engineers
Introduction to Keras for Researchers
Making new layers and models via subclassing
The Sequential model
Save, serialize, and export models
Training & evaluation with the built-in methods
Transfer learning & fine-tuning
Understanding masking & padding
Writing a training loop from scratch in TensorFlow
Writing your own callbacks


Package source: keras3_1.1.0.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
macOS binaries: r-release (arm64): keras3_1.1.0.tgz, r-oldrel (arm64): keras3_1.1.0.tgz, r-release (x86_64): keras3_1.1.0.tgz, r-oldrel (x86_64): keras3_1.1.0.tgz
Old sources: keras3 archive

Reverse dependencies:

Reverse imports: reservr
Reverse suggests: kerastuneR, tensorflow, tfdatasets


Please use the canonical form to link to this page.