Add yolo v3/4

[williamFalcon]

eriklindernoren/PyTorch-YOLOv3#487
https://github.com/Tianxiaomo/pytorch-YOLOv4

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[OrthoDex]

Hello! I'd like to take this up!

[ananyahjha93]

@OrthoDex Have a go at it! We would welcome this contribution from your side. Do go through our contributing guidelines before you submit a PR. (https://pytorch-lightning.readthedocs.io/en/latest/CONTRIBUTING.html)

[threefoldo]

I'm also interested. @OrthoDex Will you publish your code while it's incomplete?

Beside yolo, I also want to rewrite OCR models in lightning.

[ananyahjha93]

@OrthoDex any updates on this?

[OrthoDex]

Hey @ananyahjha93, I'll aim to put up a PR sometime this weekend!

[romesco]

@OrthoDex A couple students in my lab are about to start doing this. Would love to have your PR as a starting point instead of duplicating efforts =].

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[senarvi]

I created a pull request of my YOLO implementation, which my employer Groke Technologies authorized to contribute to PyTorch Lightning. It was quite difficult to get it right, since many details can only be found from the different versions of the Darknet C implementation (and many details are incorrect in the various PyTorch implementations). So if any of you know the YOLO architecture well, it would be great if you could check if you think that I've made any mistakes.