Huy Do

cc @zengk95

> potential race condition when sync jobs with two versions of the conf with the same key are run in parallel I think we could avoid this by not running...

FYI, I have another prime candidate for this script besides CUDA stuffs, https://github.com/pytorch/pytorch/blob/master/.jenkins/pytorch/win-test-helpers/installation-helpers/install_miniconda3.bat#L12. This is definitely going to be useful :)

Btw, you could try to send a HEAD request to S3 to see which metadata is returned. I couldn't remember for sure if there is a way to get hash...

@pytorchbot rebase -s

> LGTM from the functorch side (i.e. skimmed that all the functorch tests look to be running), but I definitely don't know the build stuff as well so might want...

@pytorchbot merge

I'm seeing some data race warning when running under tsan https://github.com/pytorch/pytorch/actions/runs/3137374824/jobs/5095634188. Let me look closer into them to see if they are legit or just noise

@pytorchbot merge

> So, I'm a bit curious: what's the point of adding a config that does not error out on TSAN violations? Oh, this is simply because I haven't had the...