added the rfs transforms for cifarfs

[brando90]

added the rfs transforms for cifarfs

Description

Fixes #[ISSUE NUMBER]

Added the data transform according to rfs paper for cifarfs, fc100 original rfs code: https://github.com/WangYueFt/rfs/blob/f8c837ba93c62dd0ac68a2f4019c619aa86b8421/dataset/cifar.py#L26

Contribution Checklist

If your contribution modifies code in the core library (not docs, tests, or examples), please fill the following checklist.

• [ ] My contribution is listed in CHANGELOG.md with attribution.
• [x] My contribution modifies code in the main library.
• [ ] My modifications are tested.
• [ ] My modifications are documented.

[brando90]

realted PR: https://github.com/learnables/learn2learn/pull/305

[brando90]

Thanks for the PR @brando90, I've added a few comments. In addition, do you want to add your contribution to CHANGELOG.md?

More than happy to. I've not done many PRs before, so I am unsure how to do it though. Do I create another file in the PR itself or edit that file in this PR or create a new PR with the CHANGELOG.md?

[brando90]

is there something missing? :)

[brando90]

wouldn't it be better to have as an option for data sets to receive the transforms e.g.

def cifarfs_tasksets(
    train_ways=5,
    train_samples=10,
    test_ways=5,
    test_samples=10,
    root='~/data',
    device=None,
    **kwargs,
):

adding the option there?

[seba-1511]

@brando90

is there something missing? :)

I believe the proposed changes haven't been pushed? You'd need to clone the branch, incorporate them, and push.

As for the transforms, I'd rather not include them in the benchmarks too.

[brando90]

@brando90

is there something missing? :)

I believe the proposed changes haven't been pushed? You'd need to clone the branch, incorporate them, and push.

As for the transforms, I'd rather not include them in the benchmarks too.

not sure which ones you referring too but I can see the changes here: https://github.com/learnables/learn2learn/pull/303/files

[seba-1511]

I meant the changes requested before merging (removing contented docs, swapping train augmentation for test, etc). As soon as these ones are pushed, we can run the tests.