InvertibleNetworks.jl - Memory efficient deep learning in Julia
| Title | InvertibleNetworks.jl - Memory efficient deep learning in Julia |
| Publication Type | Conference |
| Year of Publication | 2021 |
| Authors | Philipp A. Witte, Mathias Louboutin, Ali Siahkoohi, Gabrio Rizzuti, Bas Peters, Felix J. Herrmann |
| Conference Name | JuliaCon |
| Month | 07 |
| Keywords | deep learning, Invertible network, Julia, Normalizing flows, segmentation |
| Abstract | We present InvertibleNetworks.jl, an open-source package for invertible neural networks and normalizing flows using memory-efficient backpropagation. InvertibleNetworks.jl uses manually implement gradients to take advantage of the invertibility of building blocks, which allows for scaling to large-scale problem sizes. We present the architecture and features of the library and demonstrate its application to a variety of problems ranging from loop unrolling to uncertainty quantification. |
| Notes | (JuliaCon, virtual) |
| URL | https://slim.gatech.edu/Publications/Public/Conferences/JuliaCon/2021/witte2021JULIACONmedlj/witte2021JULIACONmedlj.pdf |
| URL2 | |
| Citation Key | witte2021JULIACONmedlj |