[moe training] add llama4 benchmarking script

[danielvegamyhre]

Stacked PRs:

• [moe training] integrate rowwise expert quant kernel #2698
• [moe training] add bench script for fp8 rowwise kernels and update autotune configs #2697
• [moe training] add fp8 rowwise kernels for expert weights #2696
• [moe training] add benchmark script for moe layer #2671
• ->[moe training] add llama4 benchmarking script #2669

---

[moe training] add llama4 benchmarking script

• Add benchmarking script for llama4 for use in moe training development
• Rename torchtitan benchmarking scripts to be model-centric

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/2669

• 📄 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 546d90d with merge base 23b0219 ():
💚 Looks good so far! There are no failures yet. 💚

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