Add AWS region fallback to functional-gpu workflow (backport #668)

[mergify]

We're getting out of capacity issues with these smaller instances as well, so add the region fallback logic from our larger e2e jobs to this workflow.

---

This is an automatic backport of pull request #668 done by Mergify.

[mergify]

Cherry-pick of 8ee175e has failed:

On branch mergify/bp/release-v0.8/pr-668
Your branch is up to date with 'origin/release-v0.8'.

You are currently cherry-picking commit 8ee175e.
  (fix conflicts and run "git cherry-pick --continue")
  (use "git cherry-pick --skip" to skip this patch)
  (use "git cherry-pick --abort" to cancel the cherry-pick operation)

Unmerged paths:
  (use "git add <file>..." to mark resolution)
	both modified:   .github/workflows/functional-gpu-nvidia-t4-x1.yml

no changes added to commit (use "git add" and/or "git commit -a")

To fix up this pull request, you can check it out locally. See documentation: https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/reviewing-changes-in-pull-requests/checking-out-pull-requests-locally

[bbrowning]

I fixed the failed cherry-pick here as well as ensured the fix from #669 made its way into this backport as well.

[booxter]

Assuming it passes the job.

[booxter]

Since it's a pull_request_target job, it should first pass https://github.com/instructlab/sdg/actions/runs/15570480126 to confirm it works, the automatic CI triggered won't validate the new workflow.