What to do with new datasets?

[Dopervigest]

New datasets must be checked for availability and integrity. In order to do this, we have download tests that check if each URL is reachable, and the files they provide correspond to the ones that we intend them to.
The checks work great but consume quite a bit of time. The files I added in PR "Added some peices of miccai2015 dataset" each take around a minute to download, so i didn't commit the tests for now.
I don't know how to resolve this problem, the files are already compressed with 7z and there are only 10 models in both datasets. If you have any suggestions on what i should do, i'd be very glad to hear them

[KonstantinUshenin]

@Dopervigest , @maksim-dzhigil You should make another folders for long test, like test_long. When we separate main and develop, some test can be run on push to develop and some on push to main.

Datasets availability will be checked only on PR to the main, but not to the develop. Also, we will pay for CDN, and files can be loaded faster.