Higher-Level Interface of 'torch' Package to Auto-Train Neural Networks

Provides a higher-level interface to the 'torch' package for defining, training, and fine-tuning neural networks through code generation. The package supports several architectures, including feedforward (multi-layer perceptron) and recurrent neural networks (RNN, LSTM, GRU), while reducing boilerplate 'torch' code. Model training methods also bridge to machine learning frameworks in R, particularly the 'tidymodels' ecosystem, including 'parsnip' model specifications, workflows, recipes, and tuning tools.

kindling

Package overview

Title: Higher-Level Interface of ‘torch’ Package to Auto-Train Neural Networks

Whether you’re generating neural network architecture expressions or directly fitting/training models, {kindling} minimizes boilerplate code while preserving {torch}. Since this package uses {torch} as its backend, GPU acceleration is supported.

{kindling} also bridges the gap between {torch} and {tidymodels}. It works seamlessly with {parsnip}, {recipes}, and {workflows} to bring deep learning into your existing {tidymodels} modeling pipeline. This enables a streamlined interface for building, training, and tuning deep learning models within the familiar {tidymodels} ecosystem.

Main Features

• Code generation of {torch} expression

• Multiple architectures available

  • Base models interface: feedforward networks (MLP/DNN/FFNN) and recurrent variants (RNN, LSTM, GRU)
  • Generalized neural network trainer that has the same topology as MLPs

• Native support for R ML workflows and pipelines (currently {tidymodels}; {mlr3} planned)

• Fine-grained control over network depth, layer sizes, and activation functions

• GPU acceleration support via {torch} tensors

Installation

You can install {kindling} on CRAN:

install.packages('kindling')

Or install the development version from GitHub:

# install.packages("pak")
pak::pak("joshuamarie/kindling")
## devtools::install_github("joshuamarie/kindling")

Learn more

• Getting Started with kindling
• Tuning Capabilities
• Custom Activation Function
• Special Cases: Linear and Logistic Regression
• Similar Packages and Comparison

References

Falbel D, Luraschi J (2023). torch: Tensors and Neural Networks with ‘GPU’ Acceleration. R package version 0.13.0, https://torch.mlverse.org, https://github.com/mlverse/torch.

Wickham H (2019). Advanced R, 2nd edition. Chapman and Hall/CRC. ISBN 978-0815384571, https://adv-r.hadley.nz/.

Goodfellow I, Bengio Y, Courville A (2016). Deep Learning. MIT Press. https://www.deeplearningbook.org/.

Citation

If you use {kindling} in a publication, please cite it. Run citation("kindling") in R to get the current citation, or see the CITATION file.

License

MIT + file LICENSE

Code of Conduct

Please note that the kindling project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.