Invertible Neural Networks

I was looking at doing some modelling work using invertible neural-networks in Julia with Lux.jl.

There aren’t many options, but I found GitHub - slimgroup/InvertibleNetworks.jl: A Julia framework for invertible neural networks · GitHub which seems like it has had a lot of work put into it. Unfortunately it does not seem to be actively maintained anymore. Its documentation is also very sparse, and it only has support for Flux.jl with Zygote.

Any other options or ideas on how to best proceed are also welcome. The best way to go is probably to revive InvertibleNetworks.jl, add the Lux.jl / DifferentiationInterface.jl support and add some more documentation. I am hoping others are also interested in that.

1 Like