GitLab

Clean up Conda build folder before every run
Enable artifact upload to GitHub for every workflow run, but upload to Conda/S3 only on nightly pushes

Test plan: `conda install -c pytorch-nightly torchvision; python -c "import torchvision;print(torchvision.io.read_image('hummingbird.jpg').shape)"`

* Add M1 testing job

* libjpeg -> jpeg<=9b in test-m1.yml

* Added export PATH=~/miniconda3/bin... from 6122

* Tests were OK, let's see if we can remove the pinning

Explicitly set PATH to point to `conda` binary, otherwise libjpeg detection logic does not work
Pin libjpeg to the same version on x86 and m1
Add simple tests that jpeg can be decoded by a generated wheel

* [BE] Unify version computation

Instead of hardcoding dev version in various script, use one from
`version.txt` if `setup_build_version` is called without arguments

Also, pass `--pre` option to M1 build/test pip install commands to build
TorchVision against nightly pytorch

* Pin torchvision dependency to a specific pytorch version

This would package libpng and libjpeg.dylib into wheel files
Add a very simple test step, copied from https://github.com/pytorch/pytorch.github.io/blob/1eaa33a3d3f1b83b64c5031c6dd04dbb238d6105/scripts/test_install.py#L78
Cherry-picked from https://github.com/pytorch/builder/blob/d0bc74cc363a9da5a8b6a40e883d40d25d050036/build_m1_domains.sh#L22