Run slow model E2E CI on periodic only

[kirklandsign]

Summary

Some model E2E are super slow (get idea from https://hud.pytorch.org/pytorch/executorch/commit/6d6630edae0bdc37c8450e55f2fc0ef90cb5f50c). Remove DL3 and EDSR for now.

Test plan

Reduced CPU usage

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/6742

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 0493d9f with merge base ddc8ea6 ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[guangy10]

@kirklandsign On what backend are these models running slow? Is it on xnnpack? I observed extreme slowness today when running previously supported HF models on xnnpack. And it produces garbage outputs (see the CI job on your PR: https://github.com/pytorch/executorch/actions/runs/11750351996/job/32738448869?pr=6742#step:14:10567). I'm wondering the slowness you are seeing is same. If yes, we should not remove/disable the CI/tests.

[kirklandsign]

@kirklandsign On what backend are these models running slow? Is it on xnnpack? I observed extreme slowness today when running previously supported HF models on xnnpack. And it produces garbage outputs (see the CI job on your PR: https://github.com/pytorch/executorch/actions/runs/11750351996/job/32738448869?pr=6742#step:14:10567). I'm wondering the slowness you are seeing is same. If yes, we should not remove/disable the CI/tests.

These are for XNNPACK and portable. EDSR is very slow. Do you think we want to remove HF models on XNNPACK as well?