[BugFix] A4W4 FMoE run_config weight shuffle

[zihaomu]

Motivation

Fix FMoE run_config to preshuffle FP4 weights when marking them as shuffled, keeping weight and scale layouts consistent during tuning compare.

Technical Details

Test Plan

Test Result

Submission Checklist

• Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

[github-actions]

🏷️ CI Guide

Runs automatically on every PR:

• ✅ Pre-checks (submodule verification, code formatting)
• ✅ Aiter op tests (gfx942 + gfx950)
• ✅ Triton tests on MI35X (only when aiter/ops/triton/** or related paths are changed)

Extended tests (opt-in via labels):

Label	Tests
ci:triton-300x	Run an additional Triton test job on MI300X in PRs; main branch always runs both MI35X and MI300X
ci:sglang	SGLang integration tests: DeepSeek-R1-MXFP4 accuracy, Qwen 3.5 accuracy
ci:atom	ATOM benchmark: DeepSeek-R1-0528, GPT-OSS-120B
ci:atom_full	ATOM accuracy suite for PR and main models from ATOM models_accuracy.json
ci:vllm	vLLM benchmark: GPT-OSS-120B, DeepSeek-R1-0528, Kimi-K2.5
ci:all	All standard extended tests (excludes ci:atom_full)

Only add ci:atom_full for FlyDSL or Triton upgrades.
Add labels via the sidebar or gh pr edit 3110 --add-label <label>

[Copilot]

Pull request overview

Fixes the FMoE tuning run_config path to preshuffle FP4 weights when they are treated as preshuffled (is_shuffled=True), keeping the FP4 weight and scale layouts consistent during tuning comparisons.

Changes:

• Preshuffle w1_qt_fmoe / w2_qt_fmoe with shuffle_weight(..., (16, 16)) in the FP4 fallback branch so is_shuffled=True accurately reflects the tensor layout.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.