Reapply "Add EXECUTORCH_THREADPOOL_SIZE options, default to u… (#14307)

[GregoryComer]

This reverts commit 750cba7.

Re-applying the better threadpool size defaults from #14090 with the fix from #14838. This gives a 2-4x speedup for many models and platforms (I measured 4x speedup on M1 with MobileNet V3 + XNNPACK). On high core count server platforms (doing evals, for example), this can give a 100x speedup out of box.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/14842

• 📄 Preview Python docs built from this PR

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

❌ 2 New Failures, 1 Pending

As of commit 0cea48f with merge base 2672dd3 ():

NEW FAILURES - The following jobs have failed:

• pull / test-samsung-models-linux / linux-job (gh)
  RuntimeError: Command docker exec -t f33338dcce02c26c73f3ab074bd8c8e16b6863e41d0869763df83dbb38132252 /exec failed with exit code 1
• Test CUDA Builds / test-voxtral-cuda-e2e / linux-job (gh)
  RuntimeError: Command docker exec -t 9b53b41f0e428fdf068a3c506b4d5cea1320d8780798fecf3809b615d83c4854 /exec failed with exit code 2

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

[GregoryComer]

Voxtral CI failure is due to running on a fork. Samsung failure is also pre-existing.

[kimishpatel]

assuming this is reapply. stamping

[GregoryComer]

assuming this is reapply. stamping

Yeah, this is a direct re-apply of the previously landed PR. Thanks.