RKeOps: kernel operations on GPU, with autodiff, without memory overflows in R
Details
RKeOps is the R package interfacing the cpp/cuda library KeOps. It provides standard R functions that can be used in any R (>=3) codes.
The KeOps library provides seamless kernel operations on GPU, with auto-differentiation and without memory overflows.
With RKeOps, you can compute generic reductions of very large arrays whose entries are given by a mathematical formula. It combines a tiled reduction scheme with an automatic differentiation engine. It is perfectly suited to the computation of Kernel dot products and the associated gradients, even when the full kernel matrix does not fit into the GPU memory.
For more information, please read the vignettes
(browseVignettes("rkeops")) and visit
https://www.kernel-operations.io/.
References
Charlier B, Feydy J, Glaunès JA, Collin F, Durif G (2021). “Kernel Operations on the GPU, with Autodiff, without Memory Overflows.” Journal of Machine Learning Research, 22(74), 1-6. https://jmlr.org/papers/v22/20-275.html.