add untrained alexnet

[dnabanita7]

I have added untrained AlexNet along with a master file for co-ordinating all the vision models. It is linked to issue.

[darsnack]

Good job! There are a couple changes I suggested to make the code more "Julian." For future models, you should use Metalhead as a reference to avoid replicating code, but you don't need to use their style of wrapping layers in a struct.

[dnabanita7]

I have made all the requested changes.

[darsnack]

Perfect! Just some quick stylistic changes.

[darsnack]

Awesome!

[dnabanita7]

Can you please merge it so that I will delete this branch and start working for resnet?

[darsnack]

I can't merge changes into MLH-Fellowship/FluxModels.jl. I don't know how your mentors want you to do stuff, but eventually you need to make a PR to darsnack/FluxModels.jl with these commits which I will merge.

Probably a good idea to sync up with your mentor and figure out exactly what the process they want you to follow is.

---

EDIT: Actually, looking through the Discord, you are supposed to fork darsnack/FluxModels.jl to MLH-Fellowship/FluxModels.jl. You make commits to MLH-Fellowship/FluxModels.jl directly (in some dev branch) then make PRs to darsnack/FluxModels.jl from MLH-Fellowship/FluxModels.jl

[darsnack]

I would get the MLH staff to give you write access to this repo. Then merge these changes into MLH-Fellowship. Create a new PR to darsnack/FluxModels.jl from MLH-Fellowship/FluxModels.jl. In the future, you should make commits to a dev branch of MLH-Fellowship/FluxModels.jl directly.

[dnabanita7]

I have merge access now. So I am merging and creating the PR. I will follow the guidelines next time as I wasn't clear.