RomeoV

FYI I like the white boxes as they are. I don't know what you fixed about the multimedia, usually the things I need work fine (?)

I also ran into the issue with faceting. It's not a great solution, but you can add your own Label like so: ```julia plt = draw(...) # suppose this has...

Hey everyone, this looks awesome. Is anyone still working on this? Otherwise I would suggest trying to merge this, even if it's not "perfect" with regards to extra dependencies or...

I'm still getting this error fyi. Has anyone tried adapting the lines of code from the PROPACK reference above to the KSVD.jl implementation? EDIT: Actually I found in my case...

I wonder if this resolves the following issue: ```julia using LinearAlgebra function measure() data = rand(3,3) mat = [ data ones(size(data, 1)) ones(size(data, 2))' 0 ] det(mat) end @code_warntype measure()...

Could use a rebase onto master. That makes the errors go away for me.

It seems there's also other dead links. E.g. on the [Introduction page](https://fluxml.ai/FastAI.jl/dev/FastAI@dev/doc/docs/introduction.md.html) the "callbacks" link points to https://fluxml.ai/FluxTraining.jl/dev/docs/callbacks/reference.md.html, which is dead.

Some issues I'm running into: - `datarecipes` not available at precompile time. Also, I suppose we don't have access to any data that needs to be downloaded. @lorenzoh does `FastVision`...

Hmm this approach brings the TTF-epoch from 77s to about 65s, which is a speed up for sure, but I was kind of hoping for even more. I will have...

Another update - I ran a training similar to the code above, but without any FastAI.jl/FluxTraining.jl, i.e. just Flux.jl and Metalhead.jl (see code below). With using the precompile approach from...