[TRANS, IMP] Add new max_size parameter to Resize

[NicolasHug]

Addresses (part of) #2868

This PR adds a new max_size parameter to Resize. If the size of the resized image is greater than max_size, than the image is resized again to fit within max_size. This is similar to Detectron2's https://github.com/facebookresearch/detectron2/blob/911b091c23a29c624606475d6a2f345568d79e52/detectron2/data/transforms/augmentation_impl.py#L122

[codecov]

Codecov Report

Merging #3494 (aa5162c) into master (7d41547) will increase coverage by 0.10%.
The diff coverage is 95.45%.

@@            Coverage Diff             @@
##           master    #3494      +/-   ##
==========================================
+ Coverage   78.65%   78.76%   +0.10%     
==========================================
  Files         105      105              
  Lines        9723     9723              
  Branches     1557     1561       +4     
==========================================
+ Hits         7648     7658      +10     
+ Misses       1583     1575       -8     
+ Partials      492      490       -2     

Impacted Files	Coverage Δ
torchvision/transforms/functional_tensor.py	79.84% <90.47%> (+0.65%)	⬆️
torchvision/transforms/functional.py	81.77% <100.00%> (ø)
torchvision/transforms/functional_pil.py	68.12% <100.00%> (+0.51%)	⬆️
torchvision/transforms/transforms.py	84.30% <100.00%> (+0.02%)	⬆️
torchvision/datasets/utils.py	72.58% <0.00%> (+2.37%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7d41547...164c1ec. Read the comment docs.

[fmassa]

Looks great, thanks!