[feat] Enable NVFP4 for model weight loading daemon

[xuantengh]

Motivation

#27139 enables the daemon mode for model weight loading from CUDA IPC. This PR extends its supporting range further by including NVFP4 models (e.g., nvidia/Qwen3-30B-A3B-FP4).

Modifications

The design of this PR delegates the post-process procedure of NVFP4 weight loading to daemon. The daemon process captures some of the important attributes like weights_padding_cols and intermediate_size_per_partition after loading the raw weights, and handle over them to the client then. At the client side, it just rebinds them into the related quantized parameters after receiving the CUDA IPC handle.

Accuracy Tests

Speed Tests and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

Review and Merge Process

1. Ping Merge Oncalls to start the process. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments or contact authorized users to do so.
   • Common commands include /tag-and-rerun-ci, /tag-run-ci-label, /rerun-failed-ci
4. After green CI and required approvals, ask Merge Oncalls or people with Write permission to merge the PR.

---

CI States

Latest PR Test (Base): ❌ Run #30679818820
Latest PR Test (Extra): ❌ Run #30679818652

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[liusy58]

I will review this PR as soon as possible

[liusy58]

why this is needed?

[xuantengh]

It was sent to the client side to ensure both sides load the weights using the same backend:

https://github.com/xuantengh/sglang/blob/8baf41f1c151f9da02243ec0db8ac083ea1a854f/python/sglang/srt/weight_cache/ipc_loader.py#L388-L401

As different MoE backends post-process the weight tensor (e.g., permute, reorder) into different physical layout, mixing backends will lead to shape mismatch or numeric error.