Add binary segmentation example using cpu

[omidvarnia]

Hi @qubvel. I have added a slightly altered version of your binary_segmentation_intro notebook, but forcing torch to use CPU without using a pl trainer, as PyTorch Lightning did not work on our cluster, likely due to a memory issue. Regards, Amir

[qubvel]

Thanks for adding a script example! I added a few comments to improve it and would be glad if you could work on it. Thanks a lot 🤗

[omidvarnia]

Hi @qubvel! I have tried to address you comments and also, tested the code using GUP for an exemplary case of binary segmentation of buildings in the CamVid dataset. Your feedback would be appreciated.

[qubvel]

Thanks for iterating! Looks much better, I left a few more comments. Also, let's not commit images into the repo

[qubvel]

Just a few nits, but other than that looks nice! Thanks a lot 🤗

[omidvarnia]

Many thanks @qubvel. I have updated the script following your latest comments. Also, I have already tested the code on a GPU-enabled server for 100 epochs, and it works just fine. Have a good weekend!

[qubvel]

The one thing left - to apply make fixup to make style test pass, thank you!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

[qubvel]

@omidvarnia Thanks for contribution 🤗

[omidvarnia]

You're welcome, @qubvel! Really appreciate your valuable review and feedback :-)